Abstracts

Identification of reciprocal hybrids in citrus by the broadness of the leaf petiole wing

Rosa M.L. Ballve^1,2, Herculano Penna Medina-Filho² and Rita Bordignon²

¹ Fundação IAC, Campinas, SP, Brasil.

² Seção de Genética, Instituto Agronômico de Campinas, Caixa Postal 28, 13001-970

Campinas, SP, Brasil. Send correspondence to H.P.M.-F.

ABSTRACT

Broadness of leaf petiole wing (WB) was investigated as a morphological marker for screening hybrids of the very narrow-winged species Citrus limonia and C. sunki with broad-winged species C. aurantium and C. sinensis. Controlled polinizations produced over 500 reciprocal hybrids with potential in the ongoing rootstock breeding program identified by the Pgi-1 and Prx_a-1 isozyme loci.

Measurement ratios WB/leaf length, WB/leaf broadness and WB/petiole length identified 86 to 91% of the reciprocal hybrids produced. However, visual classification of WB was an equally efficient but much easier and faster method. It can be very useful in breeding programs when large number of plants have to be screened or when isozyme, RFLP or RAPD laboratories are not available.

INTRODUCTION

Reproduction of most Citrus species is characterized by the occurrence of nucellar polyembryony in the seeds, a feature that leads to mixed offsprings containing nucellar clones (true-to-type maternal genotype) and zygotic (sexual) seedlings (Cameron and Frost, 1968). In breeding programs where hybridizations are involved, it is very desirable to identify the type of each seedling before the trials are set up in the field because in this way only the hybrids are evaluated. Early identification of hybrids from seed parents highly polyembryonic results in increased money and time savings because most of the plants produced are unwanted nucellar clones (Soost and Cameron, 1975).

Dependable methods to separate nucellar from hybrid seedlings in progenies rely on the evaluation of genetic markers. Certainly, in order to screen a large number of plants, the ideal markers are morphological traits, in view of the ease and quickness of the evaluations. For the identification of interspecific hybrids within the genus Citrus, these markers are unfortunately scarce. Apparently the only well-studied case that has a known genetic basis is the presence of anthocyanine pigment in some red colored Citrus limon cultivars (Toxopeus, 1962). Because this trait is dominant, it is useful only when C. limon is the male parent. Otherwise, as the seed genitor, both hybrid and nucellar plants have red shoots, being undistinguishable. This situation is similar to the intergeneric crosses involving the trifoliate leaf trait of Poncirus trifoliata (Toxopeus, 1962) and the red colored young shoots of Severinia buxifolia (Medina-Filho et al., in press).

An attempt to use a morphological trait to identify hybrids was made by Teich and Spiegel-Roy (1972) who investigated the ratio of leaf length/width (L/W) in some C. reticulata cultivars and hybrids. Hybrids of cultivars having significantly different L/W ratio supposedly could be discriminated if the value was out of the range of the seed parent. On the other hand, the values in the mean range of the two parents were considered hybrids. However, the reliability of this method was not evaluated, due to the lack at that time of a bona fide marker to be used as a control.

Although RFLP and more recently RAPD analysis are powerful techniques for hybrid identification (Roose et al., 1992; Waugh and Powell, 1992), isozyme alleles are well known and still recognized as accurate markers in citrus. Since the study of Torres et al. (1978), they have been extensively used to distinguish nucellar from zygotic plants (Torres, 1983; Anderson et al., 1991; Fatta Del Bosco et al., 1994). The usefulness of isozymes has been demonstrated in comparison with other criteria, like trifoliate leaves (Manzocchi et al., 1981), the polyphenol-oxidase browning and coagulation of young shoot homogenates (Geraci et al., 1981), or visual observations of plant phenotype (Moore and Castle, 1988). A key attribute of the isozymes as markers is that they are coded by codominant genes which allow the unequivocal identification of nucellar and reciprocal hybrids, in most cases. A through discussion of the use of isozymes and their efficiency as markers in citrus is presented by Ballvé et al. (1991).

Toxopeus (1962) studied the broadness of leaf petioles as a marker. However, the understanding and usefulness of it was impaired by a complicated cross and hindered by very unexpected results. A cross of C. maxima with C. hystrix, two monofoliate broad-winged species, produced mostly plants with trifoliate leaves and, among the monofoliate ones, the great majority had narrow-winged petioles.

This paper reports investigations on the broadness of the leaf petiole wing (WB) as a marker for screening reciprocal hybrids of the very narrow-winged species C. limonia and C. sunki with broad-winged species C. aurantium and C. sinensis. These hybrids were produced aiming at combinations of several agronomic attributes present in different rootstocks (Pompeu Jr., 1990). Plants were classified to the broadness of leaf petiole using several methods, with their efficiency determined by isozymes.

MATERIAL AND METHODS

Seedlings three to twelve months old were obtained from crosses of C. limonia Osbeck ‘Limeira‘ x C. aurantium L. ‘São Paulo‘, C. sunki Hort. ex. Tanaka ‘200‘ x C. aurantium, their reciprocals, and C. sunki x C. sinensis Osbeck ‘Natal‘,‘ Valencia‘, ‘Pera‘ and ‘Hamlin‘. Hand pollinations were performed according to the procedures of Bordignon et al. (1990). Three criteria were used to classify the plants of each progeny as nucellar or hybrid: visual observation of petioles, measurements of petiole wings, and isozyme analysis of the plants.

Visual observation

The predominant type of leaf petiole of each seedling at the nursery was determined by direct observation. Among offspring of C. sunki or C. limonia as the mother parent in crosses with sour or sweet orange as the pollen parent, narrow-winged seedlings, as the mother plants, were classified as nucellars and those which showed broad-winged petioles, in various degrees, were considered hybrids. Conversely, when sour orange was the female genitor, seedlings with similar broad wings were considered nucellars and those with narrow-wings in various degrees, hybrids.

Measurements

C. sunki, C. limonia and C. aurantium parental trees were characterized in a sample of 100 leaves. Measurements of the petiole wing broadness (WB), leaf broadness (LB), petiole length (PL), and leaf length (LL) were taken (Figure 1), and the ratios WB/PL, WB/LB, and WB/LL were calculated. For WB and the above ratios, the range of variation, mean, standard deviation and confidence limits were determined. The confidence limits at the 5% level were considered the typical values of parental clones. Statistical analyses were performed using the SANEST computer program.

This same procedure was applied to 10 leaves of each seedling screened, and the confidence limit was calculated. When it fell out of the range of the seed parent, the seedling was classified as a hybrid. Otherwise, it was considered nucellar. Each measure was studied independently so that the same seedling could be indicated as nucellar by one measure but hybrid by another. Sweet orange was not included in this study, as explained later.

Isozyme analysis

Segregation and recombination of the alleles of phosphoglucoisomerase-1 (Pgi-1) and anodic peroxidase-1 (Prx_a-1) loci (Table I) allowed unequivocal identification of all nucellar and hybrid seedlings from crosses between C. sunki or C. limonia with C. aurantium. For the crosses of C. sunki with C. sinensis, only Prx_a-1 was useful since its genotype FF theoretically identifies 50% of the hybrids produced.

Table I - Genotypes of Pgi-1 and Prx_a-1 isozymic loci of Citrus aurantium, C. limonia, C. sunki and C. sinensis.

Electrophoresis was accomplished according to Ballvé et al. (1995) and Tanksley (1979), with equipment for horizontal starch gels with the following specifications:

Samples: a leaf disk of 1 cm in diameter was crushed with five drops of distilled water;

Gel: hydrolized potato starch: 120 g/l;

Buffer: 25 ml/l Tris 0.73 M (0.018 M) and 25 ml/l citric acid 0.15 M (0.36 M), pH 8.2 adjusted with Tris (0.73 M);

Electrophoresis: electrodes buffer: 0.3 M boric acid, pH 8.3 adjusted with 4 N NaOH;

Electrical conditions: for inserting the extracts into the gel a current of 25 mA (not exceeding 150V) was applied for 25 min. After removal of wicks, the same electrical conditions were maintained for 50-60 min. Then, the amperage was adjusted to 30 mA, not exceeding 300 V, and gels ran for 30-40 min until they reached 300 V. After that, the gels were run for 2 h within the limits of 300 V and 30 mA.

RESULTS AND DISCUSSION

In C. sunki and C. limonia the petioles of all leaves are invariably very narrow-winged. Contrarily, C. aurantium and C. sinensis display leaves with broad wings (Figure 2). In these broad-winged clones, however, a few leaves with very reduced wings can be observed. This variation is represented in the measurements by the range of the histograms of distribution shown in Figure 3. C. sunki and C. limonia, in accordance with the visual observations, displayed very little variation. In C. aurantium, WB and its ratios showed wide variation. For all measurements, C. aurantium showed a certain proportion of leaves within the range of values of C. sunki and C. limonia (WB 9%; WB/PL 29%; WB/LB 34%; WB/LL 18%) indicating, in numbers, the aforementioned occurrence of narrow-winged leaves. However, as depicted by the modes, means, and confidence limits of the measurements and their ratios (Table II), the great majority of leaves are winged and appreciably different from C. sunki and C. limonia.

Table II - Values of modes, means, standard deviations (SD) and confidence limits (CL) of the wing broadness (WB) and its ratio to petiole length (WB/PL), leaf length (WB/LL), and leaf broadness (WB/LB) of 100 leaves of parental trees of Citrus limonia, C. sunki and C. aurantium.

The isozyme analyses indicated (Table III) that from 1024 seedlings analyzed 561 hybrids were identifyed. As mentioned before, with exception of the C. sunki x C. sinensis cross, the seedlings classified as nucellar were, surely, of nucellar origin and, as expected, all of them showed exactly the same type of petiole of the female parental plant from which they ontogenetically originated.

The percentage of hybrids from the crosses shown in Table III were quite high: 29% for C. limonia, 22% for C. aurantium and 74% for C. sunki. This suggests a problem for the use of these species as rootstocks. However, the present data resulted from artificial cross-pollinations with different clones, a situation not observed in commercial seed orchards. C. sunki bearing usually one to three embryos per seed produces 44-77% of hybrids when artificially cross-pollinated with compatible clones, but when open pollinated or selfed it yields over 98% of nucellar plants (Carvalho et al., in press).

The data of Table III indicate that all criteria for evaluation of leaf petiole wing are quite satisfactory to identify hybrids, showing some variation when each cross is analyzed separately. Considering the crosses altogether, it is noteworthy that the visual observation criterion recognized 90% of hybrids. Only the WB measurement provided a better efficiency (91%). The 10% of hybrids not identified, thus confounded with nucellars, belong to a class of seedlings whose wing broadness overlap the limit range of the seed parents.

The frequency distribution observed in the F1 plants (Figure 4) suggests that petiole wing is controlled by polygenes and that parental clones must be heterozygous in some loci and/or some kind of epistasis is involved. Furthermore, broad-winged genotypes do not express the trait uniformly in all leaves of the plant. In fact, considerable variation was observed in different leaves of single plants as seen in the case of C. aurantium (Figure 3). The genetic basis of the trait must await a forthcoming study of specific F₂ progenies. This, however, does not preclude the herein suggested use of the broadness of petiole wings in the identification of hybrids.

Indeed, the efficiency of the visual evaluation was comparable to the laborious identification by measurements. Since it is the simplest of all methods, it should be preferred. In order to increase its efficiency, an alternative strategy would be to associate it with electrophoresis, checking by isozymes only those plants that are visually classified as nucellars. Such a procedure would identify the remaining 10% of hybrids in the progenies, if necessary.

An important feature of the petiole wing as a marker is that it is useful to identify hybrids from both crossing directions with comparable efficiency, as observed in the reciprocal crosses between C. sunki and C. aurantium. A somewhat different proportion of hybrids were recognized in reciprocal crosses of C. aurantium x C. limonia. However, these results are not conclusive since a small number of seedlings were evaluated.

Concluding, the broadness of leaf petiole wing is a good marker for the identification of hybrids of sour (C. aurantium) and sweet orange (C. sinensis) in crosses with very narrow-winged species such as C. sunki and C. limonia. Although 10% of the hybrids cannot be identified on the basis of this easily screened trait, all seedlings visually classified as hybrids were in fact hybrids, as checked by isozyme analysis. This trait therefore can be very useful in extensive breeding programs if a large number of plants have to be screened, or when an isozyme, RFLP or RAPD laboratory is not available. The hybrids so produced and identified are being evaluated as to their potential as rootstocks for citrus in Brazil.

ACKNOWLEDGMENTS

The authors are indebted to Dr. Violeta Nagai for assistance in the statistical analysis and to the Centro de Citricultura Sylvio Moreira of IAC for providing parental plants and growing progenies of crosses.

Herculano Penna Medina Filho and Rita Bordignon were recipient of CNPq and CAPES fellowships, respectively. Publication supported by FAPESP.

RESUMO

Investigou-se a largura da asa do pecíolo da folha (WB) como um marcador morfológico para identificar híbridos recíprocos de Citrus limonia e C. sunki, que possuem a asa do pecíolo bastante reduzida, com C. aurantium e C. sinensis, cujas asas dos pecíolos são bem desenvolvidas. Polinizações controladas produziram mais de 500 híbridos identificados pelos loci isoenzímicos Pgi-1 e Prx_a-1. WB e a razão desta pelo comprimento da folha, pela largura da folha e pelo comprimento do pecíolo identificaram, em média, 86 a 91% dos híbridos recíprocos produzidos. A simples observação subjetiva da WB é, entretanto, um processo igualmente eficiente porém muito mais fácil e rápido. Este poderá ser bastante útil e econômico em programas de melhoramento nos casos em que se necessite a seleção de grande número de plantas ou quando não se dispõe de laboratórios de eletroforese de isoenzimas RFLP ou RAPD. Os híbridos produzidos e identificados têm sido avaliados quanto ao seu potencial como porta-enxertos de citros no Brasil.

REFERENCES

Anderson, C.M., Castle, W.S. and Moore, G.A. (1991). Isozymic identification of zygotic seedlings in Swingle Citrumelo Citrus paradisi x Poncirus trifoliata nursery and field populations. J. Am. Soc. Hort. Sci. 116: 322-326.

Ballvé, R.M.L., Bordignon, R., Medina Filho, H.P., Siqueira, W.J., Teófilo Sobrinho, J. and Pompeu Júnior, J. (1991). Isoenzimas na identificação precoce de híbridos e clones nucelares no melhoramento de citros. Bragantia 50: 57-76.

Ballvé, R.M.L., Medina Filho, H.P., Bordignon, R. and Lima, M.M.A. (1995). Methodology for starch gel electrophoresis and protocols for isozymes of 32 plant genera. Rev. Bras. Genet. 18: 491-502.

Bordignon, R., Medina Filho, H.P. and Ballvé, R.M.L. (1990). Melhoramento genético de citros no Instituto Agronômico. Laranja, 11: 167-176.

Cameron, J.W. and Frost, H.B. (1968). Genetics, breeding and nucellar embryony. In: The Citrus Industry (Reuther, W., Batcherlor, L.D. and Webber, H.J., eds.). Vol. II. Univ. Calif. Press, Berkeley, pp. 325-370.

Carvalho, M.R.T., Bordignon, R., Ballvé, R.M.L., Pinto Maglio, C.A.F. and Medina Filho, H.P. Aspectos biológicos do reduzido número de sementes da tangerina Sunki (Citrus sunki). Bragantia (in press).

Fatta Del Bosco, S., Matranga, G. and Geraci, G. (1994). Isozyme analysis of Citrus rootstock populations to identify zygotic seedlings. Adv. Hort. Sci. 8: 71-74.

Geraci, G., Manzocchi, L.A., Tusa, N., Occorso, G., Radogna, L. and De Pasquale, F. (1981). Comparison of different methods for identifying zygotic and nucellar seedling in citrus. Proceedings of the International Society of Citriculture, 5. Tokyo, Japan, 1982, Vol. 1, pp. 1-4.

Manzocchi, L.A., Tusa, N. and Geraci, G. (1981). Correlation between phenotypic and biochemical genetic markers in offsprings of sour orange x P. trifoliata. Genet. Agr. 35: 367-376.

Medina Filho, H.P., Bordignon, R. and Ballvé, R.M.L. Sunkifolias and Buxisunkis: sexually obtained hybrids of Citrus sunki X Severinia buxifolia. Braz. J. Genet. (in press).

Moore, G.A. and Castle, W.S. (1988). Morphological and isozymic analysis of open-pollinated citrus rootstock populations. J. Hered. 79: 59-63.

Pompeu Jr., J. (1990). Situação do uso de porta-enxertos no Brasil. In: Seminário Internacional de Citros. Bebedouro, Brasil. Anais, pp. 1-10.

Roose, M.L., Jarrell, D.C. and Kupper, R.S. (1992). Genetic mapping in a Citrus x Poncirus F₂ population. Proceedings of the International Society of Citriculture, 7. Acireale, Italy, 1992. Vol. 1, pp. 210-213.

Soost, R.K. and Cameron, J.W. (1975). Citrus. In: Advances in Fruit Breeding (Janick, J. and Moore, J.N., eds). Purdue Univ. Press, West Lafayette, Indiana.

Tanksley, S.D. (1979). An efficient and economical design for starch gel electrophoresis. Report of the Tomato Genetic Cooperative. U.C. Davis, CA. 29: 37-38.

Teich, A.H. and Spiegel-Roy, P. (1972). Differentiation between nucellar and zygotic citrus seedlings by leaf shape. Theor. Appl. Genet. 42: 314-315.

Torres, A.M. (1983). Fruit trees. In: Isozyme in Plant Genetics and Breeding. Part B (Tanksley, S.D. and Orton, T.J., eds.). Elsevier Sci. Publ., Amsterdam, pp. 401-421.

Torres, A.M., Soost, R.K. and Diedenhofen, U. (1978). Leaf isozymes as genetic markers in citrus. Am. J. Bot. 65: 869-881.

Toxopeus, H. (1962). Notes on the genetics of a few leaf characters in the genus Citrus. Euphitica 11: 19-25.

Waugh, R. and Powell, W. (1992) . Using RAPD markers for crop improvement. Trends Biotechnol. 10: 186-191.

(Received July 11, 1996)

Figure 2 - Different typical leaf petioles of plants from a progeny of Citrus limonia (L) x C. aurantium (A) as male parent. Except for the rightmost leaf, which belongs to a nucellar plant, all of them are from hybrids identified by isozymes.

Figure 4 - Broadness of leaf petiole wing in Citrus species of Figure 3 and in 10 leaves of each reciprocal hybrid of Table III. L, C. limonia; A, C. aurantium; S, C. sunki.

Publication Dates

History