"Somatic Hybridization of Citrus with Wild Relatives for Germplasm

FEATURE

Somatic Hybridization of Citrus with Wild Relatives for Germplasm Enhancement and Cultivar Development Jude W. Grosser1 and Frederick G. Gmitter, Jr.2 Citrus Research and Education Center, IFAS, University of Florida, 700 Experiment Station Road, Lake Alfred, FL 33850

Publicity regarding the loss of existing ge- tergeneric somatic hybridization of plants has Genetic perspectives on intergeneric netic resources because of human activity in been accomplished in very few families. It somatic hybridization of plants the centers of plant, origin and diversity has has been a powerful research tool, ‘but thus A wide range of mechanisms prevents stimulated efforts to collect and maintain far there has been little direct application of successful hybridization between related germplasm in international genebank net- the technique to plant breeding and cultivar genera of plants. Barriers to hybridization works (Poopathy, 1986). The problem is development. The propensity for interspe- include genetically controlled gametophytic particularly acute for Citrus and related gen- cific hybridization within the genus Citrus and sporophytic incompatibility systems era because government-sponsored land de- and among the genera of the “True Citrus (Taylor, 1980), prezygotic mechanisms (e.g., velopment programs are destroying the natural Fruit Tree” subtribal group (Table 1) is re- inhibition of pollen germination, inadequate habitat of many Citrus spp., and relatives. markable and rare among woody plants or arrested pollen tube growth), postzygotic The International Board of Plant Genetic Re- (Barrett, 1977; Swingle and Reece, 1967). mechanisms (e.g., endosperm failure, ploidy sources (IBPGR) [with liaison to the United However, factors such as long reproductive imbalance), and somatic incompatibilities Nation’s Food and Agricultural Organization cycle, influence of apomixis and heterozy- (Harms, 1983). It is often difficult to identify (FAO)] is negotiating with the Univ. of Ma- gosity on selection, difficulty with sexual the mechanism(s) preventing sexual hybrid- laysia to set up a genebank network for Cit- hybrid survival, and complexity of detecting ization in specific cases. However, if this rus (Poopathy, 1986). Justification and genetic traits have precluded a full under- can be achieved, techniques are available to impetus for this program may be provided standing of taxonomic relationships and lim- facilitate sexual hybridization (Collins et al., by the abundance of economically important its to gene transfer within and among Citrus 1984; Sharma and Shivanna, 1986; Williams traits (e.g., tolerance to biotic and abiotic and related genera (Barrett, 1977). The work and DeLautour, 1980; Yamashita, 1987; stress, horticultural! characteristics and per- described herein is a good example of the Yamashita and Tanimoto, 1985). Hybrids formance, etc.) that can be found among Cit- direct application of biotechnological meth- produced from such wide combinations are rus spp. and related genera. For such programs ods to the attempted solution of agricultural often infertile. In some cases, doubling the to be successful, however, technology must problems as well as to the elucidation of crit- chromosome number (sometimes a difficult be available to maximize the use of the col- ical genetic questions for perennial plants. task) can restore fertility by providing ap- lected germplasm. A major goal of traditional plant breeding propriate chromosome partners for normal The use of somatic hybridization to obtain is the transfer of genes conferring disease meiotic pairing. The probability of produc- wild germplasm for woody fruit tree im- resistance or stress tolerance from related wild ing fertile hybrids in this fashion between provement is discussed in this article, using species into modern high-yielding crop cul- distant relatives is relatively low. Citrus as a model. Recovery of intergeneric tivars (Cocking, 1985). This approach is dif- In contrast, somatic hybridization via pro- somatic hybrid plants produced between sex- ficult to apply to Citrus breeding (and most toplasm fusion bypasses the sexual process to ually incompatible Citrus sinensis L. Osb. other crops) because the majority of related CV. Hamlin (an early sweet orange) and the wild species are either difficult or impossible wild relative Citropsis gilletiana Swing. and to hybridize with Citrus using traditional Table 1. Genera of tribe Citrae, subtribe II: Ci- Ken. (Gillet’s cherry orange), a genus native crossing methods. The Reed for gene transfer trinae (Citrus fruit trees, three subtribal groups, to Africa, is described. Methodology used to is increasing because urban development of 13 genera), according to Swingle and Reece produce these hybrids is outlined. Hybridity prime agricultural land is forcing agriculture (1967) and compatibility with Citrus (Sykes, was confirmed by analysis of morphology, onto less-desirable land, where environmen- 1988). somatic cell chromosome number, and phos- tal stress is a greater factor. Moreover, the Sexual Graft phohexose isomerase (PHI; EC 5.3.1.9) iso- use of chemicals to control pests and path- compatibility compatibility zyme banding patterns obtained by ogens is being limited or eliminated in many A) Primitive citrus fruit trees electrophoresis on horizontal starch gels. This cases to minimize contamination of the en- Severiniaz — + hybrid and other intergeneric somatic hy- vironment. Most of the world Citrus industry Pleiospermium – — brids that we have produced are being field- is a perennial monocultural system com- Burkillanthus — — evaluated for citrus rootstock potential. In- posed of a small number of closely related Limnocitrus — — genotypes cultivated in relatively small areas Hesperathusa — + of dense plantings, creating the potential for B) Near citrus fruit trees Citropsisz — + disaster from a build up of pathogens (Bar- — Received for publication 18 Nov. 1988. Florida Atalantia + rett and Rhodes, 1976). The narrow, vul- C) True citrus fruit trees Agricultural Experiment Station Journal Series no. nerable germplasm base of cultivated Citrus 9365. We thank J.L. Chandler and Pan-Chi Lieu Fortunella + + necessitates the successful integration of im- Eremocitrus + + for technical assistance. The cost of publishing z portant traits from distant relatives into su- Poncirus + + this paper was defrayed in part by the payment of — page charges. Under postal regulations, this paper perior new Citrus cultivars. Further, Citrus Clymenia + is amenable to contemporary techniques of Microcitrus + + therefore must be hereby marked advertisement z solely to indicate this fact. cell culture and gene transfer, thereby pro- Citrus + + 1Associate Professor. viding the opportunity to bypass barriers to zIndicates successful somatic hybridization with 2Assistant Professor. sexual hybridization. Citrus.

HORTSCIENCE, VOL. 25(2), FEBRUARY 1990 147 produce allotetraploid (amphidiploid) hy- bility is generally difficult or impossible 1971; Button and Kochba, 1977; Chaturvedi brids. It is an additive process that combines (Barrett, 1977; Swingle and Reece, 1967). and Mitra, 1974; Kobayashi et al., 1983; the complete nuclear genomes of both par- The subtribe Citrinae of the orange subfam- Mitra and Chaturvedi, 1972; Moore, 1985; ents; however, in extremely wide combina- ily Aurantiodeae (family Rutaceae) has been Ohgawara et al., 1985; Rangan et al., 1969; tions, chromosome elimination can occur split into three subtribal groups (Table 1), Starrantino and Russo, 1980; Vardi et al., (Kao, 1977). The potential heterozygosity of and there are no existing sexual hybrids among 1982, 1986), the emphasis of our research somatic hybrids is great, depending on cu- them. For example, Citropsis gilletiana has been to develop and implement a routine mulative allelic differences between the con- Swing. and M. Ken., a distant relative of somatic hybridization system capable of tributing parents. Also, in contrast to sexual Citrus and a member of the “Near Citrus generating somatic hybrids among Citrus and hybridization, where cytoplasmic (chloro- Fruit Tree” subtribal group (Table 1), was sexually incompatible genera (Grosser et al., plasts and mitochondria) genome inheritance identified as an excellent candidate species 1988a, 1988b). is predominantly maternal, somatic hybrid- for Citrus germplasm enhancement. This Protoplasm research involving Citrus spe- ization allows for cytoplasmic genome con- species is immune to Phytophthora ci- cies has been limited (as with all woody spe- tributions from both parents. Additional trophthora (Sm. and Sm.), a severe type of cies), but there are several examples of whole- genetic variability may arise from cyto- foot rot (Swingle and Reece, 1967), and re- plant regeneration via somatic embryogene- plasmic and nuclear gene recombination or sistant to Radopholus citrophilus Huettle, sis from Citrus protoplasts derived from nu- somaclonal variation (Cocking, 1985). Dixon and Kaplan, the burrowing nematode cellar callus (Kobayashi et al., 1983; Vardi Successful somatic hybridization has been (Ford and Feder, 1960). Citropsis gilletiana et al., 1982, 1986). The first example of largely limited to the Solanaceae, the Cru- and several other potentially important’ Cit- somatic hybridization in Citrus was reported ciferae, and, more recently, the Legumino- rus relatives (e. g., Severinia, Feronia, by Ohgawara et al. (1985). This intergeneric sae (Evans, 1983; Sundburg and Glimelius, Swinglea, etc.) are graft-compatible with somatic hybrid, produced between sexually 1986; Wright et al., 1987). The lack of ef- Citrus, but sexually incompatible, and they compatible members of the “True Citrus Fruit ficient protoplasm-to-plant regeneration sys- are not horticulturally acceptable when used Tree” subtribal group, was generated from tems outside these families and somatic directly as rootstock for Citrus (Swingle and the fusion of leaf protoplasts of Poncirus tri- incompatibilities between distantly related Reece, 1967). Successful hybridization with foliata (L.) Raf. with embryogenic callus species (Harms, 1983) have been the pre- Citrus may minimize or eliminate problems protoplasts of Citrus sinemsis CV. Trovita. We dominant limiting factors. Somatic hybrids of inadequate horticultural performance. have produced a similar hybrid between P. in the aforementioned families have not been Fertile intergeneric hybrids would provide new trifoliata CV. Flying Dragon and C. sinensis directly useful in the field because they carry sources of breeding material with unique ge- CV. Hamlin (Grosser et al., 1988a). This was negative traits from the wild relative species netic potential for scion and rootstock im- accomplished by fusing ‘Hamlin’ protoplasts that reduce agronomic or horticultural per- provement programs (Grosser et al., 1988a; isolated from a habituated embryogenic sus- formance and harvestable product quality to 1988b). Intergeneric somatic hybrids may pension culture with ‘Flying Dragon’ pro- unacceptable levels. Extensive backcrossing have direct potential as rootstock; however, toplasts isolated from juvenile leaf tissue programs are required to eliminate such traits, if backcrossing is required to improve hor- (Grosser and Chandler, 1987; Grosser et al., but frequently such efforts are hampered by ticultural performance, one or two genera- 1988a). low fertility of the somatic hybrids (Cock- tions may be sufficient. Therefore, successful The hybrid selection scheme used was ing, 1985). Another negative factor is the somatic hybridization of woody species may based on complementation of the regenera- increased ploidy level, which is unsuitable provide horticulturally useful, economically tive capability of the ‘Hamlin’ cell line and for most breeding programs. important, new combinations of character- the subsequent expression of the trifoliate leaf istics. Somatic hybridization offers other ad- character of ‘Flying Dragon’. Unfused The role of somatic hybridization in vantages over traditional breeding methods ‘Hamlin’ protoplasts acted as nurse culture Citrus germplasm enhancement and for Citrus. Most cultivated Citrus types are cells and provided the appropriate cell den- cultivar development highly heterozygous hybrids that reproduce sity necessary for heterokaryon survival. The Citrus scion cultivars, like many fruit and apomictically from seed and produce few, if ‘Hamlin’ line was in culture for more than nut trees, are seldom grown on their own any, zygotic embryos or seedlings because 4 years. Although protoplasts isolated from roots. Rather, they are almost universally of nucellar (maternal) embryony. Male and/ this line were amenable to culture and fu- budded onto related, but genetically distinct, or female sterility is also sometimes ob- sion, the capacity for whole-plant recovery genotypes (rootstock) that are better-adapted served. Zygotic seedlings, when produced, was lost (attributed to culture age). The pro- to a particular environment, resulting in stress are often weak and do not survive. This is toplasm culture medium used did not support tolerance, improved fruit quality, and in- perhaps due to the expression of deleterious cell wall resynthesis or mitosis in unfused creased yield. This method of propagation recessive genes that are masked in the par- ‘Flying Dragon’ leaf protoplasts. Hybrid offers a unique advantage for employing new ents (Barrett and Rhodes, 1976; Swingle and plants were regenerated via somatic embryogene transfer methods because it allows ge- Reece, 1967). Such deleterious recessive genesis and multiplied organogenically. Hy- netic improvement of the rootstock without genes should remain masked and not be ex- brid morphology was intermediate to that of changing the genetic integrity of the scion or pressed in somatic hybrids. Moreover, traits the parents and chromosome counts indi- marketable fruit characteristics. The major- for which expression is controlled by dom- cated that the hybrids were allotetraploids ity of chronic problems in citriculture, inant gene systems have a high probability (2n = 36). Inheritance of malate dehydrog- including susceptibility to biotic and envi- of being expressed in amphidiploid somatic enase (MDH; EC 1.1.1.37) isozymes sepa- ronmental stresses, are influenced or con- hybrids. For Citrus, these traits include cold rated by starch gel electrophoresis was used trolled by rootstock. hardiness, polyembryony, and resistance to to verify hybrid identity further. More than Adequate resistance to citrus “blight”, a citrus tristeza virus, foot rot (Phytophthora 200 plants have been produced, and many rootstock-mediated malady of unknown cause parasitic Dast.), and citrus nematode [Ty- of these have been budded and entered into that afflicts commercial Citrus plantings, has lenchulus semipenetrans (Cobb)] (Hutchi- field trials to evaluate the potential of this not been identified within the genus Citrus. son, 1985). hybrid as a rootstock for Citrus. This is a crucial problem for Florida and Using a modification of this system, we Brazilian citriculture, as up to 1 million pro- Previous accomplishments have also produced > 150 somatic hybrid ductive trees are lost annually to blight in Considering the complexity of breeding plants between the sexually incompatible Florida alone. Successful hybridization of systems, the lack of knowledge regarding the genera Severinia disticha (Blanco) Swing. Citrus with related genera offers potential to mechanisms that prevent sexual wide hy- (the Philippine box-orange) and C. sinensis generate improved blight-resistant material. bridization, and the advancements made in CV. Hamlin (Grosser et al., 1988b). These The genetic diversity found among Citrus tissue and protoplasm culture in Citrus (Bar- genera are distantly related and belong to and related genera is profound, but accessi- lass and Skene, 1982; Button and Bornman, different subtribal groups (Table 1). Suc-

148 HORTSCIENCE, VOL. 25(2), FEBRUARY 1990 cessful hybridization of these genera was ac- tosis in growth regulator-free protoplasm plants (Fig. 1). Poncirus develops trifoliate complished by fusing embryogenic suspension culture media. Following fusion, unfused leaves; this is a dominant trait that is ex- culture-derived protoplasts of ‘Hamlin’ with protoplasts of the Citrus relative die or dis- pressed in both sexual and somatic Citrus × seedling epicotyl callus-derived protoplasts integrate; however, when the protoplasts of Poncirus hybrids (Grosser et al., 1988a; of S. disticha. Seedling epicotyl callus is the Citrus relatives are fused with habituated Ohgawara et al., 1985). Severinia exhibits nonembryogenic, and, in contrast with Cit- C. sinensis protoplasts, complementation red pigmentation in new growth, and Citrus rus embryogenic callus, is grown on a high- occurs, thereby restoring the capacity for has a winged petiole; both of these traits were auxin medium. Therefore, only somatic hy- whole-plant recovery from resulting hetero- expressed in the somatic hybrid (Grosser et brid fusion products were capable of pro- karyons. al., 1988 b). Likewise, Citropsis gilletiana ducing viable embryos in this culture system. We have already described two proven has a hi-winged pentafoliate leaf that facili- These hybrids also exhibited morphology in- sources of such protoplasts—leaves (Grosser tated identification of Citrus × Citropsis so- termediate to that of the parents and were and Chandler, 1987; Grosser et al., 1988a) matic hybrids. allotretraploid. Along with the expression of or callus tissue initiated from seedling leaf, morphological characters from both parents, epicotyl, or hypocotyl explants (Grosser et Production and characterization of further verification of hybridity was obtained al., 1988b) and grown on a high-auxin me- Citrus sinensis × Citropsis gilletiana from electrophoretic analyses of phospho- dium. Leaves provide a high-yielding source somatic hybrid plants glucomutase (PGM; EC 2.7.5.1) isozymes. of protoplasts (Grosser and Chandler, 1987); The only source of Citropsis gilletiana plant To the best of our knowledge, this was the but, in some cases, they may contain sec- material available for somatic hybridization first example of somatic hybrid plants pro- ondary metabolizes that could interfere with with C. sinensis was a juvenile seedling tree, duced among sexually incompatible genera successful protoplasm culture. Stem or leaf »3 m high. Citropsis gilletiana protoplasts of woody plants and between the “Primitive callus, consisting of undifferentiated cells were isolated either directly from sterilized Citrus” and “True Citrus” subtribal groups. committed to mitosis, are an alternative juvenile leaves or from nonembryogenic cal- Production of this hybrid indicates that so- source. The production of secondary metab- Ius tissue initiated from juvenile stem sec- matic hybridization may provide a means of olizes by such cells appears to be reduced. tions using methods already described bypassing barriers to sexual hybridization to Also, there may be a carry-over of endoge- (Grosser and Chandler, 1987). Citropsis access previously unavailable, but desirable, nous auxin in protoplasts isolated from such protoplasts from both sources were chemi- Rutaceous germplasm. callus that could possibly enhance plating ef- cally fused with C. sinensis CV. Hamlin pro- ficiency following fusions. toplasts isolated from a habituated Somatic hybrid selection Adequate yields of protoplasts that are embryogenic suspension culture and cultured The key to successful somatic hybridiza- amenable to fusion can be isolated from auxin- as described by Grosser et al. (1988a, 1988b). tion is having an efficient selection scheme grown callus of many Citrus relatives (un- Numerous somatic hybrid embryos (> 250) to identify and isolate the desired fusion published data), including Citropsis gille- of normal morphology were recovered from products. Following the chemical fusion of’ tiana. It is also beneficial if the explant tissue fusions involving Citropsis leaf protoplasts. protoplasts, the freqtiency of heterokaryon has the capacity to undergo adventitious bud Some embryos were produced directly on the formation is generally low (in the range of induction and shoot formation (organoge- liquid medium, but most arose from calli that 1% to 10%). Hybrid cells must be able to nesis) when placed on media containing high had been transferred to solid medium. About compete with unfused parental cells, and there levels of the synthetic cytokinin benzyl- 50% of these embryos produced shoots when must be a system to regenerate, identify, and adenine. This characteristic increases the cultured on a medium containing 1.0 mg GA3/ separate hybrid plants. Many methods have possibility of regenerating whole plants fol- liter (gibberellic acid) and 15 mg coumarin/ been devised to accomplish this in various lowing fusions (Grosser et al., 1988 b), and liter. A few embryos that produced both roots systems, and excellent reviews are available it may provide a means to multiply resulting and shoots were transferred directly to pots (Collins et al., 1984; Evans,, 1983). Based hybrid plants for field evaluation (Grosser et containing a commercial soil mixture. The on the production of the previously discussed al., 1988a, 1988b). remaining majority of shoots formed func- somatic hybrids, we have devised a flexible Numerous distinct morphological differ- tional roots when cultured on a half-strength system that restricts whole-plant recovery to ences among Citrus and its wild relatives MT basal medium (Murashige and Tucker, somatic hybrid material. The system takes simplify the identification of somatic hybrid 1969) containing 0.5 g neutralized activated advantage of differences in metabolic requirements and in morphogenetic ability of cells from different explant sources (somatic embryogenesis vs. organogenesis). Habituated C. sinensis callus with embryogenic potential can be grown on growth regulator-free liquid or solid medium. If maintained in culture for a long period of time (usually 2 or more years), the capacity for embryogenesis is maintained, but the capability to recover whole plants from recovered embryos can be either greatly diminished or eliminated. Suspension cultures of such tissue (on a 2-week subculture cycle) are a superior, high-yielding source of protoplasts already adapted to liquid culture. As demonstrated (Grosser et al., 1988a, 1988b), unfused habituated C. sinensis protoplasts act as nurse cells to provide the cell densities necessary for adequate plating ef- ficiency. To make the selection scheme work, the source of Citrus relative protoplasts must come from an explant that has growth regulator requirements for successful culture. Protoplasts isolated from such explants can- Fig. 1. Leaf morphology (from left to right) of Citropsis gilletiana, somatic hybrid of Citropsis not resynthesize cell walls nor undergo mi- gilletiana × Citrus sinensis CV. Hamlin, and C. sinensis CV. Hamlin.

HORTSCIENCE, VOL. 25(2), FEBRUARY 1990 149 charcoal/liter. As of Aug. 1988, > 135 plants a more complex, complementary pattern that the second somatic hybrid produced between were growing vigorously in soil and were showed the presence and activity of alleles sexually incompatible woody species. ready for [he evaluation of this hybrid as a unique to each of the contributing parents potential Citrus rootstock. (Fig. 2). To the best of our knowledge, these Prospects Recovery of somatic hybrid embryos was C. sinensis × Citropsis gilletiana somatic hy- As discussed above, it is possible that the much less efficient from fusions involving brid plants are the first example of viable intergeneric somatic hybrids of Citrus with Citropsis callus protoplasts than with leaf hybrids produced” between members of the Citropsis, Severinia, and Poncirus may have protoplasts. No embryos were recovered di- “Near Citrus Fruit Tree” and the “True Cit- direct use as improved rootstock for the world rectly on liquid medium, but 13 embryos were rus Fruit Tree” subtribal groups, and only citrus industry (Fig. 3). These selections re- recovered from calli that had been transferred to solid medium. Four of these embryos produced shoots when cultured on the germination medium, and they rooted read- ily when cultured on rooting medium. These results contrast with the generation of the C. sinensis × Severinia disticha somatic hybrid, in which plants were recovered only from fusions involving S. disticha callus protoplasts. Fusions of C. sinensis with S. disti- da leaf protoplasts did not result in the recovery of hybrid plants. Although the morphology of the first leaves produced by putative regenerated hybrid plants was quite variable (Fig. 1), it eventually sta- bilized to an intermediate trifoliate form with an elongated winged petiole. All of the regenerated plants appeared uniform. Hybrid plants were vigorous and produced well-de- veloped root systems. Cytological evaluation of 10 randomly selected somatic hybrid plants, using the method described by Grosser et al. (1988a), revealed the allotetraploid chromosome number of 36. Additional verification of somatic hybrid identity was obtained from PHI isozyme analysis. PHI isozymes were separated using horizontal starch gel electrophoresis with constant current of 60 mA for 3 hr at 6C and the pH 5.7 histidine– citrate buffer system of Cardy et al. (1981). The enzyme activity stain recipe was from Vallejos (1983). Both Citrus sinensis and Citropsis gilletiana are heterozygous at the PHI locus and display a three-banded pattern typical of a dimeric enzyme. However, these two parents can be distinguished easily because the protein products specified by the Citropsis gilletiana alleles migrate faster than those of C. sinensis. The zymogram produced by the somatic hybrid plants revealed

Fig. 2. Starch gel electrophoresis of PHI isozymes. The photo of the zymogram demon- Fig. 3. Intergencne somatic hybrid trees; ungrafted, and as rootstock budded with ‘Valencla’ sweet strates a complementary banding “pattern in orange scion: Citrus sinensis CV. Hamlin + Poncirus trifoliata CV. Flying Dragon somatic hybrid somatic hybrid plants (H + Cg) produced be- (A) ungrafted and (B) grafted; Citrus sinensis CV. Hamlin × Severinia disticha somatic hybrid (C) tween ‘Hamlin’ sweet orange (H) and Citropsis ungrafted and (D) grafted; and Citrus sinensis CV. Hamlin × Citropsis gilletiana somatic hybrid (E) gilletiana (Cg). ungrafted and (F) grafted.

150 HORTSCIENCE, VOL. 25(2), FEBRUARY 1990 quire the same evaluation process that sexual Literature Cited Bot. Club 99:184-189. hybrids would undergo to determine their Barlass, M. and K.G.M. Skene. 1982. [n vitro Moore, G.A. 1985. Factors affecting in vitro cm- suitability as rootstock. Specifically, resis- plantlet formation from Citrus species and hy- bryogenesis from underdeveloped ovules of ma- tance/tolerance of biotic and environmental brids. Scientia Hort. 17:333–341. ture citrus fruits. J. Amer. Soc. Hort. Sci. stresses and horticultural characteristics Barrett, H.C. 1977. Intergeneric hybridization of 110:66-70. Citrus and other genera in citrus cultivar im- Murashigc, T. and D.P.H. Tucker. 1969. Growth (nursery and field performance) must be factor requirements of citrus tissue culture. Proc. evaluated and compared with current stan- provement. Proc. Intl. Soc. Citricult. 2:586- 589. First Intl. Citrus Symp. 3:1155-1161. dard cultivars. This is an unavoidable, time- Ohgawara, T., S. Kobayashi, E. Ohgawara, H. consuming requirement that is necessary be- Barrett, H.C. and A.M. Rhodes. 1976. A numer- ical taxonomic study of affinity relationships in Uchimaya, and S. Ishii. 1985. Somatic hybrid fore these hybrids can be released or rec- cultivated citrus and its close relatives. Syst. plants obtained by protoplasm fusion between ommended to the industry. Bet. 1:105-136. Citrus sinensis and Poncirus trifoliata. Theor. Although the hybrids are tetraploid, they Button, J. and C.H. Bornman. 1971. Develop- Applied Genet. 71:1-4. possess. great potential as breeding parents ment of nucellar plants from unpollinated and Poopathy, V. 1986. Oranges from the wilds. News that may serve as “genetic bridges” between unfertilized ovules of the Washington naval or- Straits Times (Malaysia). 13 Nov. ange in vitro. J. S. Afr. Bet. 37:127–134. Rangan, T. S., T. Murashige, and W.P. Bitters. Citrus and related genera. Further hybridi- 1969. In vitro studies of zygotic and nucellar zations could be undertaken using some of Button, J. and J. Kochba. 1977. Tissue culture in the citrus industry, p. 70–92. In: J. Reinert and embryogenesis in Citrus. Proc. First Ind. Citrus the uncommon, naturally occurring Citrus V.P.S. Bajaj (eds.). Applied and fundamental Symp. 1:225-229. tetraploids as suitable parents. The produc- aspects in plant, cell, tissue and organ culture. Sharma, N. and K.R. Shivanna. 1986. Treatment tion of monoploid (n = 2x = 18) plants Springer-Verlag, Berlin. of the stigma with an extract of a compatible from anther or microspore cultures originat- Cardy, B. J., C.W. Stuber, and M.M. Goodman. pistil overcomes self-inwmpatibility in Petu- ing from somatic hybrids is another alter- 1981. Techniques for starch gel electrophoresis nia. New Phytol. 102:443-447. native approach to using these unique genetic of enzymes from maize (Zea mays L.). North Starrantino, A. and F. Russo. 1980. Seedlings from lines in Citrus breeding at the diploid level. Carolina State Univ., Raleigh. Inst. Stat. Mi- underdeveloped ovules of ripe fruits of polyem- meo. Serv. 1317. bryonic citrus cultivars. HortScience 131:253- Research on methods to accomplish this on 258. a routine basis is necessary; however, the Chaturvedi, H.C. and G.C. Mitra. 1974. Clonal propagation of Citrus from somatic callus cul- Sundburg, E. and K. Glimelius. 1986. A method probability of successful monoploid produc- tures. HortScience 9:1 18–120. for production of interspecific hybrids within tion seems great, considering the amenabil- Cocking, E.C. 1985. Somatic hybridization: Im- Brassicae via somatic hybridization, using re- ity of Citrus tissue. Any use of these hybrids plications for agriculture, p. 101-113. In: M. synthesis of Brassica napus as a model. Theor. in sexual breeding schemes depends, of Zaidin (ed.). Biotechnology in plant science. Applied Genet. 43:155-162. course, on their capability to reach maturity Academic, Orlando, Fla. Swingle, W.T. and P.C. Reece. 1967. The botany and produce flowers that possess some de- Collins, G. B., N.L. Taylor, and J.W. Deverna. of citrus and its wild relatives, p. 190430. In: gree of fertility. 1984. In vitro approaches to interspecific hy- W. Reuther, L.D. Batchelor, and H.J. Webber bridization and chromosome manipulation in crop (eds.). The citrus industry. vol. 1. Univ. of Cal- The difficulties that limit Citrus breeding plants, p. 323-383. In: J.P. Gustafson (ed.). ifornia Press, Berkeley. by standard methods (i. e., sterility, apo- Gene manipulation in plant improvement. 16th Sykes, S.R. 1988. Overview of the family Ruta- mixis, prolonged juvenility, and possible in- Stadler Genet. Symp. Plenum, New York. ccac, p. 93–100. In: R.R. Walker (ed.). Citrus breeding depression) also limit meaningful Evans, D.A. 1983. Protoplasm fusion, p. 291-321. breeding workshop. CSIRO, Melbourne, Aus- genetic studies with Citrus. However, so- In: D.A. Evans, W.R. Sharp, P.V. Ammirato, tralia. matic hybrids may provide unique opportu- and Y. Yamada (eds. ). Handbook of plant cell Taylor, N.L. 1980. Clovers, p. 261-272. [n: W.R. nities to study genetic control of traits of culture. Macmillan, New York. Fehr and H.H. Halley (eds.). Hybridization of economic significance. By creating specific Ford, H.W. and W.A. Feder. 1960. Citropsis gil- crop plants. Amer. Soc. Agron. and Crop Sci. combinations of tolerant and susceptible pa- letiana, a citrus relative resistant to the burrow- Amer., Madison, Wis. ing nematode in laboratory tests. Proc. Fla. State Vallejos, C.E. 1983. Enzyme staining activity, p. rental types and studying the behavior of so- Hort. Soc. 73:60-64. 469-516. In: S.O. Tanksley and T.J. Orton matic hybrids when challenged, the additivity, Grosser, J.W. and J.L. Chandler. 1987. Aseptic (eds.). Isozymes in plant genetics and breeding. or dominance/recessivenessj of these genetic isolation of leaf protoplasts from Citrus, Pon- part A. Elsevier, Amsterdam. traits could be determined. Studies of meiosis cirus, Poncirus × Cirrus hybrids and Severinia Vardi, A., D.J. Hutchison, and E. Galun. 1986. and relative fertility of flowering somatic hy- for use in somatic hybridization experiments. A protoplasm-to-tree system in Microcitrus based brids can yield valuable information regard- Scientia Hort. 31:253-257. on protoplasts derived from a sustained embry- ing chromosome homology among Citrus and Grosser, J. W., F.G. Gmitter, and J.L. Chandler. ogenic callus. Plant Cell Rpts. 5:412–414. related genera that may prove useful in at- 1988a. Intergeneric somatic hybridization of Vardi, A., P. Spiegel-Roy, and E. Galun. 1982. tempts to describe valid taxonomic relation- Citrus sinensis CV. Hamlin and Poncirus tnfol- Plant regeneration from citrus protoplasts: Var- iata CV. Flying Dragon. Plant Cell Rpts. 7:5– iability in methodological requirements among ships of members of the Rutaceae. 8. cultivars and species. Theor. Applied Genet. Somatic hybridization is an additive process Grosser, J. W., F.G. Gmitter, and J.L. Chandler. 62:171-176. and does not involve recombination and seg- 1988b. Intergeneric somatic hybridization of Williams, E.G. and G. DeLautour. 1980. The use regation of traits in progeny. Each hybridi- sexually incompatible woody species: Citrus si- of embryo culture with transplanted nurse en- zation results presumably in only one hybrid nensis and Severinia disticha. Theor. Applied dosperm for the production of interspecific hy- genotype. The number and genetic diversity Genet. 75:397-401. brids in pasture legumes. Bet. Gaz. 141:252- of hybrids available for evaluation and se- Harms, C.T. 1983. Somatic incompatibility in the 257. development of higher plant somatic hybrids. Wright, R. L., D.A. Somers, and R.L. McGraw. lection should be increased to maximize the Quart. Rev. Biol. 58:325-353. 1987. Somatic hybridization between birdsfoot probability of creating and selecting geneti- Hutchison, D.J. 1985. Rootstock development treefoil (Lotus corniculatus L.) and L. coim- cally and horticulturally superior hybrids with screening and selection for disease tolerance and brensis Wind. Theor. Applied Genet. 75:151- this technique. This can be accomplished by horticultural characteristics. Fruit Var. J. 39:21- 156. increasing the number and diversity of pa- 25. Yamashita, K. 1987. Studies on self-incompati- rental combinations used to create somatic Kao, K.N. 1977. Chromosomal behavior in so- bility of Hassaku (Citrus hassaku Hort. ex Tan- hybrids. Future work should include efforts matic hybrids of soybean-Nicomo-ra glauca. Mol. aka): 1. Some fundamental experiments focusing to create somatic hybrids using other Citrus Gen. Genet. 150:225-230. on the significance of stigmas in incompatible relatives (including other species of Citrop- Kobayashi, S., H. Uchimaya, and I. Ikeda. 1983. reactions. J. Jpn. Soc. Hort. Sci. 56:268–273. sis and Severinia) and Citrus species other Plant regeneration from ‘Trovita’ orange pro- Yamashita, K. and S. Tanimoto. 1985. Studies on toplasts. Jpn. J. Breeding 33: 119–122. self-incompatibility of Hassaku (Citrus hassaku than C. sinensis as parents. In this manner, Mitral, G.C. and H.C. Chaturvedi. 1972. Em- Hort. ex Tanaka): H. Glycoprotein composition the untapped genetic potential that resides bryoids and complete plants from unpollmated in pistils soon after self-pollination, and several within genera related to Citrus may be ex- ovaries and from ovules of in vitro grown emas- trials to overcome self-incompatibility. J. Jpn. ploited. culated flower buds of Citrus spp. Bul. Torrey Soc. Hort. Sci. 54:178-183.

HORTSCIENCE, VOL. 25(2), FEBRUARY 1990 151