Soybean genetic resources and crop improvement

Transcription

1 605 Soybean genetic resources and crop improvement R.J. Singh and T. Hymowitz Abstract: The soybean (Glycine max (L.) Merr.) is an economically important leguminous crop for feed, oil, and soyfood products. It contains about 40% protein and 20% oil in the seed and, in the international trade markets, is ranked number one in oil production (48%) among the major oil seed crops. Despite its economic importance, the genetic base of soybean cultivars is extremely narrow. The indigenous cultivars and landraces in East Asia are on the verge of extinction, because farmers are now growing high yielding soybean cultivars. The exotic germplasm, enriched with genes for abiotic and biotic stresses, has not been fully exploited by soybean breeders. Mutation breeding has improved the fatty acids of the soybeans and has produced soybeans tolerant to herbicides. By using recombinant DNA technology, Monsanto has produced stable glyphosate tolerant soybean lines known as Round Up Ready soybean. DuPont is producing transgenic soybean lines with improved fatty acids content. The feasibility of developing hybrid soybeans is still an open question. Key words: soybean, Glycine spp., exotic germplasm, mutation, interspecific hybridization, biotechnology, hybrid soybeans. Résumé : Le soya (Glycine max (L.) Merr.) est une légumineuse cultivée importante sur le plan économique en matière d alimenation animale, d huile et de produits transformés de soya. Il contient environ 40% de protéine et 20% d huile dans la graine et se classe bon premier pour la production d huile (48%) parmi les principales oléagineuses transigées sur les marchés internationaux. En dépit de son importance économique, les bases génétiques des cultivars de soya est extrémement étroite. Les cultivars indigènes et les races locales en Asie orientale sont à veille de disparaître parce que les agriculteurs cultivent maintenant des cultivars à haut rendement. Le germoplasme exotique, riche en gènes conférant la résistance à des stress biotiques et abiotiques, n a pas été pleinement exploité par les améliorateurs. L amélioration par mutation a permis d améliorer les acides gras du soya et a produit des soyas résistants à des herbicides. En faisant appel aux technologies de l ADN recombinant, Monsanto a produit des lignées stables de soya résistantes au glyphosate connues sous le nom de soya «Round Up Ready». DePont est en train de produire des lignées transgéniques présentant une composition en acides gras améliorée. La faisabilité du développement de soyas hybrides reste à vérifier. Mots clés : soya, Glycine spp., germoplasme exotique, mutation, hybridation interspécifique, biotechnologie, hybrides de soya. [Traduit par la Rédaction] Singh and Hymowitz 616 Introduction The soybean (Glycine max (L.) Merr.), rich in seed protein (range 30 48%, average 40%) and oil (range 13 22%, average 20%), is an economically important leguminous seed crop for feed, oil, and soyfood products. Soybean is ranked number one in world oil production (48%) in the international trade markets among the major crops, viz., as cottonseed, peanut (groundnut), sunflower seed, rapeseed, coconut, and palm kernel (Fig. 1). Nowadays, the soybean is a world crop, cultivated widely in the United States, Brazil, Corresponding Editor: R.S. Singh. Received November 18, Accepted April 8, R. J. Singh 1 and T. Hymowitz. Department of Crop Sciences, University of Illinois, Urbana, Illinois, U.S.A. 1 Author to whom all correspondence should be addressed ( r-singh@uiuc.edu). Genome 42: (1999) Argentina, China, and India. The United States is the leader in soybean production (Fig. 2). Soybean was domesticated from the wild soybean, Glycine soja Sieb. & Zucc. (formerly Glycine ussuriensis Regal & Maack) (Fukuda 1933). Wild soybean is an annual weedy-form climber, whose pods contain black seeds that shatter at maturity. The plant grows wild in China and adjacent regions of Russia, Korea, Taiwan, and Japan (Hymowitz 1970). The wild soybean is rich in seed protein ( %) but poor in oil ( %) composition (Hymowitz et al. 1972). The cultivated soybean and its progenitor G. soja belong to the subgenus Soja (Moench.) F.J. Herm., and both are cross compatible, contain 2n = 40 chromosomes, and produce vigorous fertile intermediate F 1 hybrids. The F 1 plants are similar to Glycine gracilis Skvortzow (Skvortzow 1927; Fukuda 1933; Karasawa 1952; Hadley and Hymowitz 1973; Palmer et al. 1987; Singh and Hymowitz 1988, 1989).

2 606 Genome Vol. 42, 1999 The subgenus Glycine Willd. comprises 16 wild perennial species. These species are indigenous to Australia, diverse in morphological feature and genomes, and harbor invaluable genetic resources of economic importance, such as resistance to biotic and abiotic stresses. However, the untapped valuable reservoir of genetic diversity of the wild annual soybean and the 16 wild perennial species is largely unexplored by soybean breeders for improvement of the cultivated soybean. Thus, the genetic base of soybean is extremely narrow (Hymowitz et al. 1977; Delannay et al. 1983; Gizlice et al. 1993, 1994, 1996; Salado-Navarro et al. 1993; Sneller 1994; Burton 1997). Domestication and dissemination of soybean: a chronology (1) Hymowitz (1970) proposed, based on linguistic, geographical, and historical literatures, that the soybean was domesticated in the eastern half of north China (primary center of origin) during the Shang dynasty (ca B.C.) or probably earlier. Soybean was considered one of the five sacred grains (rice, wheat, barley, soybean, and millet) essential for the existence of Chinese civilization (Morse et al. 1949). (2) The soybean was disseminated to central and south China and peninsular Korea by the first century A.D., owing to degeneration of dynasties, migration of populations, and consolidation of territories (Hymowitz 1990). (3) Soybean was introduced into Japan, Indonesia, Philippines, Vietnam, Thailand, Burma, north India, and Nepal from the 1st to 16th centuries. Landraces were developed and these regions are the secondary center of origin of the soybean (Hymowitz 1990). (4) In the late 16th and throughout the 17th centuries, missionaries and sailors brought the soybean to Europe from China and Japan. Soybean was grown in 1740 in botanic gardens in France and in 1790 in the Royal Botanic Garden, Kew, England. Soybean was cultivated in several European countries but failed to attain a place in European agriculture (Morse et al. 1949). (5) Soybean was introduced into North America from China by Samuel Bowen in 1765 and was planted in Greenwich, located at Thunderbolt, a few miles east of Savannah, Georgia (Hymowitz and Harlan 1983). (6) Since 1765, soybean has been introduced into United States from China several times by scientists, seed dealers, merchants, military expeditions, and various individuals. (7) In 1770, Benjamin Franklin sent soybean seeds from London to botanist John Bartram, who planted them in his garden, which was located on the west bank of the Schuykill River below Philadelphia. (8) In 1829, Professor Thomas Nuttall grew soybean in the botanic gardens at Cambridge, Massachusetts. (9) During the years , soybean was introduced to Alton, Illinois, by Benjamin Franklin Edwards. Mr. John H. Lea of Alton planted the soybeans in his garden in the summer of In 1852, the multiplied seed was grown in Davenport, Iowa, by Mr. J.R. Jackson and in Cincinnati, Ohio, by Mr. A.H. Ernst. In 1853, Mr. Ernst Fig. 1. A diagrammatic chart showing production of the major vegetable oils of the world. Numbers in parentheses are millions of tonnes (source, Soy Stats 1996). Sunflower seed 10% (25.91) Peanut 10% (25.93) Argentina 10% (12.64) Rapeseed 14% (34.49) Paraguay 2% (2.30) India 4% (4.47) Copra 2% (5.01) Cottonseed 14% (34.59) Palm Kernel 2% (4.70) Other 7% (8.04) Soybean 48% (123.65) Fig. 2. A diagrammatic chart showing world soybean production. Numbers in parentheses are millions of tonnes (source, Soy Stats 1996). China 11% (13.50) EU 1% (0.94) Brazil 19% (23.20) U.S.A. 46% (58.56) distributed seeds to the New York State Agricultural Society, the Massachusetts Horticultural Society, and the Commissioner of Patents (Hymowitz 1987). (10) During the period , the introduced soybeans were grown in every agricultural research station in the United States. Research on soybean was conducted throughout the country in small plantings and (or) on a commercial scale, mainly for pastures, silage and hay, and animal feeds (Probst and Judd 1973). The Office of

3 Singh and Hymowitz 607 Foreign Seed and Plant Introduction was established in 1898 within the United States Department of Agriculture (USDA) and initiated the introduction of a large number of soybeans from Asiatic countries (Morse et al. 1949). This facilitated centralized plant introduction activities. Introduced plants were assigned a permanent plant introduction (PI) number. (11) The enormous economic value of the soybean was realized in the early two decades of the 20th century. Osborn and Mendel (1917) demonstrated experimentally that heated soybean meal promoted growth in rat at a normal rate, which was in contrast with raw soybean meal. This study resulted in the establishment of the soybean processing industries in the United States. Mr. A.E. Staley Sr. laid the foundation of operational soybean processing facilities in Decatur, Illinois, in He provided incentives to Illinois farmers to grow soybeans. His company processed and made more than 1000 different items of soybean food, feeds, and other products that have revolutionized the agriculture of the Midwest, and Decatur, Illinois, is now called the soybean capital of the world (Windish 1981). To enhance soybean germplasm resources, plant exploration trips to China, Japan, India, and Korea were conducted by William J. Morse and P. H. Dorsett. They collected more than introductions, representing more than 2500 distinct types (Morse et al. 1949). Prior to 1945, many accessions were used to improve vegetable-type soybeans but, during World War II, the emphasis of soybean production and utilization shifted to commercial purposes. Soybeans became one of the major crops in the United States. Soybean research exploded from trial and error to several disciplines, such as plant breeding, plant pathology, plant physiology, soil science, weed science, agricultural engineering, food processing, extension, and marketing. Thus, that little orient immigrant which used to be called a curiosity became the miracle gold bean from the soil (Windish 1981). Since soybeans were introduced into the United States from several geographical regions of East Asia, the response to the adopted country was extremely variable. Morse et al. (1949) realized the problem and developed the concept of relative maturity groups based on critical day length. They identified 9 soybean maturity groups (0 through VIII) in the soybean growing regions of the United States. Maturity group 0 and I cultivars were those adapted to the northern part of the country. Succeeding maturity groups contained cultivars adapted to areas farther south, with those of group VIII being suited for the Gulf coast region. To date, 13 maturity groups (000 through X) for soybeans have been identified. Cultivars are chosen for maturity groups for the appropriate latitude for maximum yield. Another important discovery during this period was the isolation and identification of antinutritional factors. Kunitz (1945) isolated, identified, and crystallized the protein that inhibited the proteolytic action of trypsin, which is commonly known as Kunitz trypsin inhibitor (SBTI-A2). Kunitz trypsin inhibitor protein is one of the major antinutritional elements present in raw mature soybeans. Antinutritional factors are destroyed by treating the moist seed with heat. The isolation of a mutant without SBTI-A2 would be extremely important economically, because the raw meal could Table 1. The primary gene pool of the soybean collection data base (Hill and Nelson 1997). Soybean Glycine max (L.) Merr Introduced soybeans Germplasm releases 141 Modern cultivars 392 Old cultivars 208 Private cultivars 35 Williams isolines 100 Clark isolines 295 Harosoy isolines 134 Other isolines 35 Genetic types 150 G. soja Sieb. & Zucc Total be fed to monogastric animals. Singh et al. (1969) reported a variant of SBTI-A2 among soybean cultivars using polyacrylamide disc electrophoresis. They demonstrated that both proteins (slow and fast migrating) were controlled by a single gene with codominant alleles. They concluded, The methods outlined here should be useful in studies on the inheritance of SBTI and screening to detect varieties which possess little or none of this nutritionally deleterious protein. Hymowitz et al. (1978) identified a line, PI , lacking Kunitz trypsin inhibitor, following the screening of USDA germplasm. Orf and Hymowitz (1979) studied the genetics of STBI-A2. Bernard and Hymowitz (1986) released three soybean germplasm isolines (L , L , and L ) and Bernard et al. (1991) registered L as a Kunitz soybean (Registration No. 271, PI ) cultivar for commercial production. Gene pools of the soybean No. of accessions Harlan and de Wet (1971) proposed the concept of three gene pools, primary (GP-1), secondary (GP-2), and tertiary (GP-3), based on the success rate of hybridization among species. Soybean GP-1 GP-1 consists of biological species, and crossing within this gene pool is easy hybrids are vigorous, exhibit normal meiotic chromosome pairing, and possess total seed fertility; gene segregation is normal and gene exchange is generally easy. GP-1was further subdivided into subspecies A, which includes cultivated races, and subspecies B, which includes spontaneous races. Soybean (G. max) cultivars and landraces and their wild annual progenitor, G. soja, are included in GP- 1. Diversity in soybean germplasm has been elegantly elaborated by Hymowitz and Bernard (1991) and Palmer et al. (1995). The soybean germplasm is a rich reservoir, with more than G. max accessions, probably less than G. soja accessions, and about 3500 accessions of wild perennial Glycine species in germplasm banks throughout the world (Palmer et al. 1995). However, Hill and Nelson (1997) reported a total of only accessions of soybeans (as of 31 December 1996) in the USDA soybean

4 608 Genome Vol. 42, 1999 Table 2. The tertiary gene pool (wild perennial species of the subgenus Glycine Willd.) of the soybean collection data base (Burridge and Hymowitz 1997). Species 2n Genome No. of accessions Distribution G. albicans Tind. & Craven 40 II 2 Australia G. arenaria Tind. 40 HH 5 Australia G. argyrea Tind. 40 A 2 A 2 13 Australia G. canescens F.J. Herm. 40 AA 125 Australia G. clandestina Wendl. 40 A 1,A Australia G. curvata Tind. 40 C 1,C 1 9 Australia G. cyrtoloba Tind. 40 CC 49 Australia G. falcata Benth. 40 FF 28 Australia G. hirticaulis Tind. & Craven 40, (80) H 1 H 1 1, (1) Australia G. lactovirens Tind. & Craven 40 I 1 I 1 2 Australia G. latifolia (Benth.) Newell & Hymowitz 40 B 1,B 1 46 Australia G. latrobeana (Meissn.) Benth. 40 A 3 A 3 12 Australia G. microphylla (Benth.) Tind. 40 BB 32 Australia G. pindanica Tind. & Craven 40 H 2 H 2 5 Australia G. tabacina (Labill.) Benth. 40 B 2 B 2 14 Australia 80 Complex 131 Australia a?? 91 Australia a G. tomentella Hayata 38 EE 22 Australia 40 Complex 57 Australia a 78 Complex 55 Australia b 80 Complex 54 Australia c?? 133 Australia Total 1025 a Also West Central and South Pacific Islands. b Also Papua New Guinea. c Also Papua New Guinea, Philippines, Taiwan. data base (Table 1). There are introduced accessions of soybean and G. soja includes 1102 accessions. The USDA germplasm bank has 858 soybean accessions from eight provinces of China (the center of origin of soybeans): 88 from Fujian, 43 from Jiangxi, 60 from Guangxi, 183 from Anhui, 198 from Sichuan, 57 from Hunan, 118 from Yunan, and 111 from Guizhou. Hymowitz (1984) described the fate of 4451 soybean accessions collected by P.H. Dorsett and W.J. Morse during their plant exploration to East Asia. Today, only 495 accessions are available in the USDA germplasm collection. These accessions are an extremely valuable treasure to the soybean breeding programs in North America. The loss of a large portion of the Dorsett and Morse soybean collection is a disgrace and, because farmers in East Asia now grow high yielding soybean cultivars and have stopped growing the old landraces, also irreplaceable. Thus, indigenous cultivars and landraces are on the verge of extinction. Soybean GP-2 Harlan and de Wet (1971) defined GP-2 as, All species that can be crossed with GP-1 with at least some fertility in F 1. This suggests that the soybean (G. max) does not have a GP-2. Soybean GP-3 GP-3 is the extreme outer limit of potential genetic resource. Hybrids between GP-1 and GP-3 are anomalous, lethal, or completely sterile, and gene transfer is not possible or requires radical techniques (Harlan and de Wet 1971). Based on this definition, GP-3 includes 16 wild perennial species of the subgenus Glycine. These species are indigenous to Australia and are geographically isolated from G. max and G. soja (native of China). Extensive plant exploration trips by T. Hymowitz and A.H.D. Brown in Australia and to the South and West Central Pacific Islands have enriched the collection of wild perennial Glycine species germplasm. A total of 1025 accessions are being maintained at the University of Illinois (Table 2). Intersubgeneric hybrids have been produced and fertile modified diploid lines have been derived. This suggests that soybean has a GP-3, and indications are that gene transfer from wild perennial species to the soybean is feasible (Riggs et al. 1998). Soybean germplasm enhancement Conventional plant breeding Soybean is a self-pollinated legume; natural crossing varies from <0.5 to about 1% (Carlson and Lersten 1987). Thus, the soybean is a highly inbreeding cultigen. The first soybean cultivars grown in the United States were direct introductions from Asia and selections from the introduced germplasm. During the last five decades, soybean breeding has been limited to hybridization within GP-1, and selection was focused on high yielding cultivars with high seed protein and oil. Delannay et al. (1983) traced the pedigrees for 158 U.S. and Canadian soybean cultivars and observed that the ancestors of the North American soybean gene pool were 50 plant introductions. Gizlice et al. (1993) examined pedi-

5 Singh and Hymowitz 609 Fig. 3. A graphic representation of soybean yield (t/ha) in the United States, China, and India in (source, Food and Agricultural Organization soybean production data base). grees of North American public soybean cultivars released from 1947 through They concluded that fewer than 15 progenitors constituted the major portion of the genetic base for U.S. soybean production. In another analysis, Gizlice et al. (1994) observed that only six ancestors constituted more than half the genetic base of North American soybeans. Sneller (1994) performed CP (coefficient of parentage) analysis with 122 lines from the northern and southern regions of the United States and found that the average CP was The CP values were and among northern and southern lines, respectively. Soybean breeders usually hybridize parents from within a maturity group rather than using parents that differ by even one maturity group (Gizlice et al. 1996). Salado-Navarro et al. (1993) noted, In fact, some of the oldest released cultivars were found to have the highest yields in the Argentina environments. These observations reveal that North American soybean cultivars have an extremely narrow genetic base. The narrow genetic base of public soybean cultivars may be attributed to breeding methods. Pedigree, bulk, massselection, single seed descent, and early-generation testing methods have been followed in breeding cultivars for high seed yield, pest resistance, maturity group, lodging resistance, seed size, seed quality, seed protein and oil quantity and quality, shattering resistance, resistance to mineral deficiencies and toxicities, and resistance to herbicide injury. Soybean breeders have been confined within GP-1, and have seldom used widely different maturity groups. Thus, the limited genetic diversity may be one reason why North American public and proprietary soybean cultivars have made only modest advances in yield for the past 20 years (Fig. 3). A similar situation prevails for soybean yields (0.98 t/ha in 1976 and 0.87 t/ha in 1996) in India (Fig. 3). The genetic base of soybean cultivars grown in India is extemely narrow. Soybean cultivars are direct introductions from the United States, selections from the introduced germplasm, and single-cross (two-parent crosses) hybrids (Karmakar and Bhatnagar 1996). Major constraints for enhancing the soybean yield in India are (i) poor seed longevity; (ii) lack of appropriate technology; (iii) non-availability of high yielding cultivars carrying tolerance to pod shattering, resistance to biotic and abiotic stresses, and early maturing and photoperiod insensitive; (iv) nonavailability of efficient Bradyrhizobium cultures, proper herbicides, phosphatic fertilizers, good quality seeds in sufficient quantity; (v) limited farm mechanization particularly at the time of harvesting (Bhatnagar 1995). Soybean yields in China are improving slightly, but seemed to reach a plateau by 1993 (Fig. 3). Interspecific and intersubgeneric hybridization Soybean breeders have not yet exploited the wealth of genetic diversity from exotic germplasm, such as the soybean s progenitor G. soja or the 16 wild perennial species of the subgenus Glycine. Hallauer and Miranda (1981) defined exotic germplasm: Exotic germplasm includes all germplasm that does not have immediate usefulness without selection for adaptation for a given area. This definition is not complete without the action of hybridization. Exotic germplasm often harbors single genes of economic importance, such as resistance to pests and pathogens. Glycine soja, the progenitor of G. max, may be an excellent source of genetic variability, although it harbors several undesirable genetic traits, for example, vining, lodging susceptibility, lack of complete leaf abscission, seed shattering, and small black seed coat (Carpenter and Fehr 1986). However, G. soja constitutes unimproved germplasm and, during the course of selection in successive backcross generations, undesirable traits can be discarded. Limited numbers of interspecific crosses between G. max and G. soja have been performed (Fukuda 1933; Palmer et al. 1987; Singh and

6 610 Genome Vol. 42, 1999 Hymowitz 1988). Earlier hybridizations had just been for curiosity, but the hybridizations of Palmer et al. (1987) and Singh and Hymowitz (1988) were conducted for cytogenetic information, with the later study establishing, by pachytene chromosome analysis, the first soybean chromosome map, a prerequisite for the development of the soybean cytogenetic map, which is not yet available. Attempts to broaden the genetic base of soybeans by utilizing G. soja were reported by Hartwig (1973), Ertl and Fehr (1985), and Carpenter and Fehr (1986). Hartwig (1973) reported highly productive and high-protein lines derived from soybean and G. soja hybrids. Ertl and Fehr (1985) concluded that the introgression of G. soja germplasm into the two soybean cultivars was not an effective method for increasing their yield potential. To obtain a relatively high frequency of useful segregates for cultivar development, three backcrosses to the soybean were preferred. However, small seeded (seed of <100 mg) cultivars, such as SS201, SS202 (Fehr et al. 1990a, 1990b), and Pearl (Carter et al. 1995), have been developed where G. soja was used as a nonrecurrent parent. The small seeded cultivars are used for sprouts and the fermented Japanese product natto. Qian et al. (1996) have recorded the accessions of G. soja that are potential sources of additional genes that restrict nodulation of soybean with specific strains of Bradyrhizobium. They concluded that, Introgression of such genes could result in soybean cultivars that exclude some of the indigenous strains and become nodulated with commercial strains that are more efficient at fixing nitrogen. The 16 wild perennial species of the subgenus Glycine have not been exploited in soybean breeding programs. These species are extremely diverse morphologically, cytologically, and genomically, grow in very diverse climatic and soil conditions, and have a wide geographical distribution (Hyomowitz et al. 1998; Kollipara et al. 1997; Singh et al. 1988, 1989, 1992). Wild perennial Glycine species represent the exotic germplasm for soybean improvement. They are a rich source of agronomically useful genes, such as resistance to soybean rust (Phakopsora pachyrhizi Sydow) (Schoen et al. 1992), soybean brown spot (Septoria glycines Hemmi.) (Lim and Hymowitz 1987), powdery mildew (Microsphaera diffusa Cke. & Pk.) (Mignucci et al. 1978), phytophthora root rot (Phytophthora sojae H.J. Kaufmann & J.W. Gerdemann) (Kenworthy 1989), white mold (Sclerotinia sclerotiorum (Lib. de Bary)), sudden death syndrome (Fusarium solani (Mart.) Sacc.)), tobacco ring spot (G.L. Hartman, personal communication), yellow mosaic virus (Singh et al. 1974), alfalfa mosaic virus (Horlock et al. 1997), and soybean cyst nematode (Heterodera glycines Ichinohe) (Riggs et al. 1998) and tolerance to certain herbicides (Hart et al. 1991) and salt (Pantalone et al. 1997). Soybean rust is one of the major soybean diseases in China, Thailand, India, Australia, and Taiwan. A significant reduction (80%) in yield may be caused by the pathogen (Hartman 1996). Soybean rust has been reported in Puerto Rico and Brazil (Bonde and Peterson 1996). Killgore (1996) reported soybean rust on vegetable soybeans grown on the islands of Kauai, Oahu, and Hawaii. This suggests that soybean rust is a great threat to mainland U.S. soybean production. Significant yield loss (>10%) is predicted in nearly all soybean growing areas. However, the greatest loss (up to 50%) could occur in the Mississippi delta and the southeastern coastal areas (Yang 1996). Several researchers have attempted to hybridize wild perennial Glycine spp. with the soybean, but only a few sterile intersubgeneric F 1 hybrid combinations have been reported (see Hymowitz et al. 1998). Thus far, only Singh et al. (1990, 1993) have successfully produced backcross-derived fertile progenies from the soybean and a wild perennial, Glycine tomentella (2n = 78). Currently, individual monosomic alien addition lines (MAALs) and modified diploid (2n = 40) lines are being isolated and identified (Singh et al. 1998a). The modified diploid lines could be screened for pests and pathogens. Riggs et al. (1998) reported the introgression of soybean cyst nematode resistance from G. tomentella into modified derived diploid soybean lines. This study sets the stage for the exploitation of exotic germplasm to broaden the germplasm base of the cultivated soybean. Mutation breeding Mutation breeding in soybeans has lagged behind other economically important crops. Micke et al. (1985) compiled information on cultivars produced using induced mutations. They listed 17 soybean cultivars developed by various mutagens: 10 cultivars from China (Hei Noun Nos. 4, 5, 7, 8, 16, and 26; Mu Shi No. 6 and Tai Nung Nos. 1(R) and 2(R); and Tie Feng 18); 3 cultivars from Japan (Nanbushirome, Raiden, and Raiko); and one cultivar from each of Bulgaria (Boriana), Algeria (Cerag Nr. 1), Korea (KEX-2), and the former U.S.S.R. (Universal I). All cultivars from Japan were high-yielding, because they were resistant to nematodes. Cultivar KEX-2 from Korea was early maturing (11 days) with a high yield (ca. 16%) and large seed. Karmakar and Bhatnagar (1996) listed 43 soybean cultivars released in India from 1969 to Three cultivars (Birsa Soy1, VL Soy1, and NRC2) were developed by mutagenesis, five cultivars (Bragg, Lee, Improved Pelican, Hardee, and Monetta) were direct introductions from the United States, and the remaining cultivars were selected from introductions and single crosses (two parent). Buss (1983) isolated a recessive genetic male sterile (gms) line from a M 3 generation of Essex soybean that had been irradiated with neutrons. Allelism tests revealed that the gms line inherited independently from ms1, ms2, ms3, and ms4. Thus, the newly identified Essex gms gene was assigned the symbol ms5. By using chemical mutagenesis (ethyl methanesulfonate (EMS), N-nitroso-N-methylurea (NMU), or ethyl nitrosourea (ENU)), Sebastian and Chaleff (1987) and Sebastian et al. (1989) isolated soybean lines with increased tolerance for sulfonylurea herbicides. Sebastian and Chaleff (1987) identified four single recessive genes. Allele tests revealed that each mutation resided at one of three loci (hs1, hs2, orhs3). They observed, in biochemical studies, that mutants did not contain an altered form of acetolactate synthase (the site of action of sulfonylurea herbicide). In subsequent studies, Sebastian et al. (1989) identified a monogenic semidominant mutant that was nonallelic to the hs1, hs2, and hs3 genes. They assigned the gene symbol Als1 to the line resistant to the action of sulfonylurea herbicide. Carroll et al. (1985) mutagenized soybean seeds of cv. Bragg with EMS. They isolated 15 independent nitrate-

7 Singh and Hymowitz 611 Table 3. Changes in fatty acid content in soybeans produced through mutagenesis. Fatty acid Content level Mutagen Reference Palmitic High Ethyl methanesulfonate Wilcox and Cavins 1990 Fehr et al N-nitroso-N-methyl-urea Fehr et al Low Ethyl methanesulfonate Wilcox and Cavins 1990 N-nitroso-N-methyl-urea Fehr et al Stearic High Sodium azide Hammond and Fehr 1983b X-rays Rahman et al Oleic High Ethyl methanesulfonate Brossman and Wilcox 1984 X-rays Rahman et al Low Sodium azide Hammond and Fehr 1983b Linolenic High Ethyl methanesulfonate Brossman and Wilcox 1984 X-rays Takagi et al Low Ethyl methanesulfonate Hammond and Fehr 1983a Brossman and Wilcox 1984 Table 4. Fatty acid composition of mutants A5 and A6 and their parents, FA9525 and FA8077 (Hammond and Fehr 1983a, 1983b). Fatty acid (%) Palmitic Stearic Oleic Linoleic Linolenic Mutagen Genotype 16:0 18:0 18:2 18:2 18:3 Arachidic EMS A FA Sodium azide A FA <1.0 tolerant symbiotic (nts) mutants from 2500 M 2 families. Mutant nts382 was studied extensively. In the presence of KNO 3, nts382 produced six times more nodules than those observed in control Bragg grown under identical culture conditions. Song et al. (1995) evaluated yield, N 2 fixation, and the effects on cereal crops grown subsequent to the harvest of intermediate supernodulating (2 times), extreme supernodulating (6 times), and nonnodulating mutants of Bragg, genotypes derived from the mutants, and commercial cultivars. The experiment was conducted for 6 years at two locations. The results were as follows: (i) the supernodulators and Manark were similar, with values 13 21% above Centaur ; (ii) in the plots fertilized with nitrogen, the supernodulators exhibited higher activity than the commercial cultivars, including Manark ; (iii) grain yield of the supernodulators was either the same or up to 25% less than Bragg and Centaur ; and (iv) oats and barley sown immediately after soybean harvest produced significantly greater yields than after commercial soybean cultivars. Soybean seed oil is the major vegetable oil (48%) among oil seed crops produced in the world (Fig. 1). Genetic studies have elucidated that oil synthesis in soybean seed is determined largely by the genotype of the maternal plants, because the oil content of F 1 plants was not significantly different from those of selfed seeds of the female parent (Singh and Hadley 1968). Similarly, fatty acid composition in soybean seed is determined by the maternal parent (Hammond et al. 1972). Breeding efforts to increase soybean oil from a 20% level have been unsuccessful, because oil content and seed yield have a negative relationship (Burton 1985). Soybean breeders initiated programs to improve soybean oil quality (Table 3). The principal fatty acids in soybean oil are palmitic (16:0), stearic (18:0), oleic (18:1), linoleic (18:2), and linolenic (18:3). A common soybean cultivar contains 11% palmitic, 3% stearic, 22% oleic, 56% linoleic, and 8% linolenic acid (Wilcox 1985). The high linolenic acid content (7 9%) is associated with poor flavor (fishy painty grassy melony) stability in soybean oil (Dutton et al. 1951). Mounts et al. (1988) analyzed the fatty acid composition of more than 5000 soybean samples from both northern and southern soybean germplasm collections, and identified one line, PI B, with low linolenic (4.2%) and normal oleic acid content. The low-linolenic acid content of PI B remained stable regardless of environmental conditions, which suggests that soybean germplasm lacks a strain with a linolenic acid content of 3% or less (Hammond and Fehr 1975). Mutagenesis has been an excellent tool for creating variability for fatty acid content in soybeans. From the M 4 generation, following the use of EMS, Hammond and Fehr (1983a) selected a line, designated A5, that contained an average of 4.1% linolenic acid, while the parent (FA9525) contained 6.3%; the content of other fatty acids remained unchanged (Table 4). They also isolated a line with an elevated stearic acid (28.1%) content and a marked reduction in oleic acid (19.8%), designated A6, from an M 2 population of

8 612 Genome Vol. 42, 1999 sodium azide treated seeds of FA8077 (Hammond and Fehr 1983b; Table 3). Wilcox et al. (1984) identified a genetically stable low linolenic acid (3.4%) mutant from ca M 2 plants, where seeds of soybean cv. Century had been treated with EMS. The linolenic acid content of seeds in the M 2 populations ranged from 3.4 to 11.1% and, for Century, it ranged from 6.6 to 9.4%. In contrast, Takagi et al. (1989) developed a line with high linolenic acid content (18.4%) by treating seeds of cv. Bay with x-ray irradiation. Bay contains 9.4% linolenic acid. Linolenic acid is essential in the mammalian diet. In some mammals, lack of linolenic acid causes skin lesions, lowered learning ability, stunted growth, and mental retardation (Coscina et al. 1986). Wilcox and Cavins (1987) assigned gene symbols for linolinic acid content: Fan Fan for high levels (x = 7.2 ± 0.11%), Fan fan for intermediate levels (x = 5.2 ± 0.07%), and fan fan for low levels (x = 3.2 ± 0.13%). They observed that linolenic acid content was controlled by the genotype of the embryo rather than by the genotype of the maternal parent. Rahman et al. (1994) observed no maternal and cytoplasmic effects for linolenic acid content. It has also been demonstrated that low linolenic acid content is a quantitative trait (Fehr et al. 1992). Palmitic (16:0) and stearic (18:0) acids are the two main saturated fatty acids in the soybean. Fehr et al. (1991) produced a mutant containing reduced palmitic acid content by treating soybean cv. A1937 with NMU. Low palmitic acid content was controlled by two different alleles at two different loci. They assigned the genotypes fap1 fap1, fapx fapx. This line contains ~44 g kg 1 palmitic acid, the lowest content known in soybean. Wilcox and Cavins (1990) isolated two mutants, C1726 (registration No. GP-116; PI ) and C1727 (registration No. GP-117; PI ) from cv. Century by EMS treatment. Mutant C1726 contained 8.5% palmitic acid and mutant C1727 contained 17.2% palmitic acid, while Century had 11.2%. Both mutants bred true for low and high palmitic acid content. Genetic studies revealed that alleles from two independent loci segregated for palmitic acid percentage and that the gene action was additive. The gene symbol fap1 was assigned to an allele in C1726 that acts to lower the palmitic acid level in soybean oil and fap2 was assigned to an allele in C1727 that acts to increase the palmitic acid level (Erickson et al. 1988). A reduction in palmitic acid content improves the quality of oil. An elevated palmitic acid content enhances its use in the production of food products, such as shortening and margarine (Schnebly et al. 1994). Rahman et al. (1996) examined the genetics of mutants with high oleic acid content (M11 and M23) produced by x- ray irradiation. Low oleic acid content in cv. Bay was partially dominant to the high oleic acid content in mutant M23, but completely dominant to the high oleic acid content in mutant M11. An inverse relationship between oleic and linolenic acid content in both mutants was recorded. Oil with high levels of oleic acid is less susceptible to oxidative changes during refining, storage, and frying (Miller et al. 1987). Mutagenesis can sometimes be used to break the linkage between two closely linked genes. The grassy beany flavor in soybeans and soybean products is caused by lipoxygenases. Three soybean lipoxygenases, (L1, L2, and L3), have been characterized. L1 and L2 are linked. Hajika et al. (1995) isolated a line without L1, L2, and L3 lipoxygenases by γ-ray irradiation. Soybean plants lacking lipoxygenases showed normal plant growth and yield. The production of soybeans without lipoxygenases is cost-effective, because heat treatment to inactivate these enzymes will not be required in the processing of soybean food products. The above studies elucidate that mutation breeding provides an alternative method to wide hybridization and biotechnology. Biotechnology Conventional plant breeding has failed to revolutionize gains in soybean yield (Fig. 3). Biotechnology is considered a cutting edge science with which to broaden the genetic base of crops by overcoming the genetic barriers in extremely distant crosses. Biotechnological methods include somaclonal variation, cybrids, and recombinant DNA technology. Thus, genetic engineering is one of the alternatives for developing high yielding soybeans with high protein and oil content, resistance to pests and pathogens, and tolerance to herbicides. Soybeans have received attention from tissue culture scientists of both public and proprietary institutes in order to generate normal soybean plants with increased genetic variability. Larkin and Scowcroft (1981) called these variants somaclonal variation. Morphological variants in soybean have been obtained through cell and tissue culturing (Graybosch et al. 1987; Bailey et al. 1993) and, although these research efforts failed to deliver high yielding soybeans, methodologies were developed to regenerate complete soybean plants, a prerequisite for genetic transformation. Foreign genes of economic importance can be delivered into soybeans by Agrobacterium, particle bombardment, and electroporation (Finer et al. 1995; Christou 1997). Padgette et al. (1995) reported a stable glyphosate-tolerant soybean line (known as Round up Ready ) that had been developed using the Agrobacterium-mediated gene transfer method. Kinney (1996) produced a high oleic acid content (84%) soybean through particle-bombardment-mediated transformation. The high oleic acid containing transgenic soybean lines were stable over a number of different environments during a single growing season and were competitive in terms of yield with the parental commercial soybean line. High lysine containing (up to 12%) soybean lines have been produced by transformation. Normal soybeans contain about 6% lysine. The high lysine trait was stable in R 2 and R 3 seeds. Soybean transformant with a lysine content higher than 15% carried wrinkled seed coat and exhibited poor germination (Falco et al. 1995). Soybean transformation methods are not routinely reproducible (Christou 1997). Soybean transformants are often sterile, and sterility is attributed mostly to chromosomal aberrations (Singh et al. 1998b). Frequently, unexpected segregations and low expression or disappearance of foreign genes have been observed. Genes may be physically present but gene activity may be poorly expressed or totally lost in subsequent generations. This is generally attributed to the poorly understood phenomenon of cosuppression or gene silencing (Stam et al. 1997).

9 Singh and Hymowitz 613 Potential to produce hybrid soybeans Attempts to produce a hybrid soybean have not succeeded, because (i) the soybean is autogamous, with flowers opening after pollination and fertilization; (ii) the success rate of obtaining large numbers of hybrid pods is low; (iii) low numbers of seed set per pod is common; (iv) natural crossing is extremely poor (0 1%); and (v) cytoplasmic male sterility (cms) has not been confirmed in soybean. A patent, No (October 8, 1985), has been granted for hybrid soybean production (Davis 1985). The methodology remains on the books but its application in hybrid soybean production has not been realized. Sun et al. (1997) isolated a cytoplasmic nuclear male sterile soybean line (cms A line) and its maintainer (B line) from an interspecific hybrid between G. max (line 167) and G. soja (line 035). Average pollen sterility in all BC 4 plants was about 98% and the parallel crosses showed that the female of the line was normal. Male sterility was stable. Several genic (nuclear) male sterile (gms) soybean lines (ms1, ms2, ms3, ms4, ms5, and ms6) are available (Graybosch and Palmer 1988; Skorupska and Palmer 1989; Palmer and Skorupska 1990). Recently, Jin et al. (1997) identified a gms mutant not allelic to any previously described soybean gms lines. However, they have not been exploited by soybean breeders for developing populations for biotic and abiotic stresses. Monogenic male sterile lines could be used in hybridization where artificial crossing is difficult. More than 99% of the seed set on monogenic ms1 ms1 male-sterile plants is the result of natural crossing (Brim 1973). Distinguishing morphological markers that are visible in seedlings and tightly linked with gms may facilitate early identification of gms plants (Skorupska and Palmer 1989). Skorupska and Palmer (1989) recorded close linkage (2.48 ± 0.1% 3.15 ± 0.1%) between the w1 locus (white flower and green hypocotyl) and the ms6 locus. By utilizing w1 ms6 genetic stock, Lewers et al. (1996) developed the cosegregation method for hybrid soybean production: purple seedlings are removed shortly after germination leaving only male-sterile plants, and escapes and recombinants with purple flower are removed manually at flowering. They used the terms traditional and dilution to describe the methods for hybrid soybean seed production developed by Specht and Graef (1992). The cosegregation method produced higher seed yield, better efficiency, and equal or better seed quality than the traditional and dilution methods. The cosegregation method may be useful for male-sterile-facilitated selection, the cyclic mating system (Specht and Graef 1992), and marker-assisted recurrent selection (Lewers and Palmer 1997) for cultivar development. The degree of heterosis is an important issue in hybrid soybean production. Nelson and Bernard (1984) examined 27 hybrid combinations. Five hybrids yielded 13 19% more than their better parent in at least in one season. Manjarrez- Sandoval et al. (1997) recorded heterosis for yield as high as 11% across locations. This may justify resources for breeding hybrid soybean. However, they suggested that the identification of favorable heterotic combinations may need more field experimentation. The commercialization of hybrid soybean production may be economically feasible in developing countries where labor is abundant. Conclusions The soybean is an economically important leguminous crop for feed, oil, and soyfood products. The seed contains approximately 40% protein and 20% oil. Soybean is ranked number one in oil production (48%) in the international trade markets among the major oil seed crops. The United States is the leader in soybean production, followed by Brazil, China, Argentina, and India. Despite its economic importance, the genetic base of soybean cultivars is extremely narrow. The soybean germplasm collection has many extremely useful traits. By screening USDA soybean germplasm, a line (PI ) without the Kunitz trypsin inhibitor gene has been identified and a soybean cultivar, Kunitz, has been released. Soybean breeders have exploited, to some extent, exotic germplasm such as G. soja (an annual wild progenitor of soybean). Soybean cultivars with high protein and small seed size (<100 mg) have been released. The 16 wild perennial species of the subgenus Glycine remain out of reach to breeders, because their exploitation needs cytogenetic expertise, which is lacking at most soybean research institutions. Recently, modified derived fertile diploid lines from G. max and G. tomentella intersubgeneric hybrids were obtained, and indications are that it is possible to introgress genes for resistance to pest and pathogens into the soybean. Mutation breeding has improved the fatty acid composition of soybean and has facilitated the isolation of a line lacking lipoxygenases. Lipoxygenases are responsible for the grassy beany flavor in soybean and soybean products. By using mutagenesis, soybean lines with increased tolerance for sulfonylurea herbicides have been isolated. Recombinant DNA technology has generated glyphosatetolerant soybean. Although a patent for hybrid soybean has been issued, hybrid soybean production is not cost-effective. References Bailey, M.A., Boerma, H.R., and Parrott, W.A Genotypespecific optimization of plant regeneration from somatic embryos of soybean. Plant Sci. 93: Bernard, R.L., and Hymowitz, T Registration of L , L , and L soybean germplasm lines lacking the Kunitz trypsin inhibitor. Crop Sci. 26: Bernard, R.L., Hymowitz, T., and Cremeens, C.R Registration of Kunitz soybean. Crop Sci. 31: Bhatnagar, P.S Soybean production in India: a success story. In A success story publication series on soybean production in India. Edited by N. Chomchalow and H.V. Henle. RAP publication 1995/36. FAO (Food and Agricultural Organization of the United Nations) Regional Office for Asia and the Pacific, Bangkok, Thailand. pp Bonde, M.R., and Peterson, G.L Research at the USDA, ARS containment facility on soybean rust and its causal agent. In Proceedings of the Soybean Rust Workshop held at Urbana, Ill., 9 11 August Edited by J.B. Sinclair and G.L. Hartman. College of Agricultural, Consumer, and Environmental Sciences, National Soybean Research Laboratory Publication No. 1. Urbana, Ill. pp Brim, C.A Quantitative genetics and breeding. In Soybeans: improvement, production, and uses. Edited by B.E. Caldwell. American Society of Agronomy Publ. No. 16. American Society of Agronomy, Madison, Wis. pp

10 614 Genome Vol. 42, 1999 Brossman, G.D., and Wilcox, J.R Induction of genetic variation for oil properties and agronomic characteristics of soybean. Crop Sci. 24: Burridge, J.A., and Hymowitz, T Management of the USDA wild perennial Glycine collection, Soybean Genet. Newsl. 24: Burton, J.W Breeding soybeans for improved protein quantity and quality. In Proceedings of the 3rd World Soybean Research Conference, held at Ames, Iowa, August Edited by R. Shibles. Westview Press, Boulder and London. pp Burton, J.W Soybean (Glycine max (L.) Merr.). Field Crops Res. 53: Buss, G.R Inheritance of a male-sterile mutant from irradiated Essex soybeans. Soybean Genet. Newsl. 10: Carpenter, J.A., and Fehr, W.R Genetic variability for desirable agronomic traits in populations containing Glycine soja germplasm. Crop Sci. 26: Carlson, J.B., and Lersten, N.R Reproductive morphology. In Soybeans: improvement, production, and uses. Edited by J.R. Wilcox. American Society of Agronomy Publ. No nd ed. American Society of Agronomy, Madison, Wis. pp Carroll, B.J., McNeil, D.L., and Gresshoff, P.M Isolation and properties of soybean [Glycine max (L.) Merr.] mutants that nodulate in the presence of high nitrate concentrations. Proc. Natl. Acad. Sci. U.S.A. 82: Carter, T.E., Jr., Huie, E.B., Burton, J.W., Farmer, F.S., and Gizlice, Z Registration of Pearl soybean. Crop Sci. 35: Christou, P Biotechnology applied to grain legumes. Field Crops Res. 53: Coscina, D.V., Yehuda, S., Dixon, L.M., Kish, S.J., and Greenwood, C.E.L Learning is improved by soybean oil diet in rats. Life Sci. 38: Davis, W.H Route to hybrid soybean production. United States Patent No , October 8, Delannay, X., Rodgers, D.M., and Palmer, R.G Relative genetic contributions among ancestral lines to North American soybean cultivars. Crop Sci. 23: Dutton, H.J., Lancaster, C.R., Evans, C.D., and Cowan, J.C The flavor problem of soybean oil. VIII. Linolenic acid. J. Am. Oil. Chem. Soc. 28: Erickson, E.A., Wilcox, J.R., and Cavins, J.F Inheritance of altered palmitic acid percentage in two soybean mutants. J. Hered. 79: Ertl, D.S., and Fehr, W.R Agronomic performance of soybean genotypes from Glycine max Glycine soja crosses. Crop Sci. 25: Falco, S.C., Guida, T., Locke, M., Mauvais, J., Sanders, C., Ward, R.T., and Webber, P Transgenic canola and soybean seeds with increased lysine. Bio/Technology, 13: Fehr, W.R., Cianzio, S.R., and Welke, G.A. 1990a. Registration of SS202 soybean. Crop Sci. 30: Fehr, W.R., Cianzio, S.R., Welke, G.A., and LeRoy, A.R. 1990b. Registration of SS201 soybean. Crop Sci. 30: Fehr, W.R., Welke, G.A., Hammond, E.G., Duvick, D.N., and Cianzio, S.R Inheritance of reduced palmitic acid content in seed oil of soybean. Crop Sci. 31: Fehr, W.R., Welke, G.A., Hammond, E.G., Duvick, D.N., and Cianzio, S.R Inheritance of reduced linolenic acid content in soybean genotypes A16 and A17. Crop Sci. 32: Finer, J.J., Cheng, T.-S., and Verma, D.P.S Soybean transformation: technologies and progress. In Soybean genetics, molecular biology and biotechnology. Edited by D.P.S. Verma and R.C. Shoemaker. Commonwealth Agricultural Bureaux International, Wallingford, Oxon, U.K. pp Fukuda, Y Cyto-genetical studies on the wild and cultivated Manchurian soy beans (Glycine L.). Jpn. J. Bot. 6: Gizlice, Z., Carter, T.E., Jr., and Burton, J.W Genetic diversity in North American soybean: I. Multivariate analysis of founding stock and relation to coefficient of parentage. Crop Sci. 33: Gizlice, Z., Carter, T.E., Jr., and Burton, J.W Genetic base for North American public soybean cultivars released between 1947 and Crop Sci. 34: Gizlice, Z., Carter, T.E., Jr., Gerig, T.M., and Burton, J.W Genetic diversity patterns in North American public soybean cultivars based on coefficient of parentage. Crop Sci. 36: Graybosch, R.A., and Palmer, R.G Male sterility in soybean an overview. Am. J. Bot. 75: Graybosch, R.A., Edge, M.E., and Delannay, X Somaclonal variation in soybean plants regenerated from the cotyledonary node tissue culture system. Crop Sci. 27: Hadley, H.H., and Hymowitz, T Speciation and cytogenetics. In Soybeans: improvement, production, and uses. Edited by B.E. Caldwell. American Society of Agronomy Publ. No. 16. American Society of Agronomy, Madison, Wis. pp Hajika, M., Igita, K., and Nakazawa, Y Induction of a soybean [Glycine max (L.) Merrill] line lacking all seed lipoxygenase isozymes. Jpn. Agric. Res. Q. 29: Hallauer, A.R., and Miranda, J.B Quantitative genetics in maize breeding. Iowa State University Press, Ames. Hammond, E.G., and Fehr, W.R Oil quality improvement in soybeans Glycine max (L.) Merr. Fette Seifen Anstrichm. 77: Hammond, E.G., and Fehr, W.R. 1983a. Registration of A5 germplasm line of soybean (Reg. No. GP44). Crop Sci. 23: 192. Hammond, E.G., and Fehr, W.R. 1983b. Registration of A6 germplasm line of soybean (Reg. No. GP45). Crop Sci. 23: Hammond, E.G., Fehr, W.R., and Snyder, H.E Improving soybean quality by plant breeding. J. Am. Oil Chem. Soc. 49: Harlan, J.R., and de Wet, J.M.J Toward a rational classification of cultivated plants. Taxon, 20: Hart, S.E., Glenn, D.S., and Kenworthy, W.J Tolerance and the basis of selectivity in perennial Glycine species to 2,4-D. Weed Sci. 39: Hartman, G.L Highlights of soybean rust research at the Asian Vegetable Research and Development Center. In Proceedings of the Soybean Rust Workshop held at Urbana, Ill., 9 11 August Edited by J.B. Sinclair, and G.L. Hartman. College of Agricultural, Consumer, and Environmental Sciences, National Soybean Research Laboratory Publication No. 1. Urbana, Ill. pp Hartwig, E.E Varietal development. In Soybeans: improvement, production, and uses. Edited by B.E. Caldwell. American Society of Agronomy Publ. No. 16. American Society of Agronomy, Madison, Wis. pp Hill, J.L., and Nelson, R.L USDA soybean germplasm collection report. Soybean Genet. Newsl. 24: 7 8. Horlock, C.M., Teakle, D.S., and Jones, R.M Natural infection of the native pasture legume, Glycine latifolia, by alfalfa mosaic virus in Queensland. Australas. Plant Pathol. 26: