Characterization of madarin (CITRUS SPP.) Using morphological and aflp markers.

Ernesto Tapia Campos, María Alejandra Gutiérrez Espinosa, Marilyn L. Warburton, Amalio Santacruz Varela and Ángel Villegas Monter

Ernesto Tapia Campos. M.Sc. in Fruitculture, Instituto de Recursos Genéticos y Productividad, Colegio de Postgraduados (IREGEP/CP), México. Doctoral Student, IREGEP/CP. Address: Km. 36.5 carretera México-Texcoco. Montecillo, Edo. de México 56230, Mexico. e-mail: etapia@colpos.mx

María Alejandra Gutiérrez Espinosa. Ph.D. in Horticultural Sciences, University of Florida, USA. Professor, IREGEP/CP, Mexico. e-mail: alexge@colpos.mx

Marilyn L. Warburton. B.S. and M.S., University of Arizona, USA. PhD. in Molecular Genetics, University of California at Davies, USA. Researcher, International Maize and Wheat Improvement Center (CIMMYT), Mexico. e mail: mwarburton@cigiar.org

Amalio Santacruz Varela. Ph.D. in Plant Breeding, University of Iowa, USA. Professor, IREGEP/CP, Mexico. e-mail: asvarela@colpos.mx

Ángel Villegas Monter. Doctor, Universidad de Córdova, España, Professor, IREGEP/CP, Mexico. e-mail:avillegas@colpos.mx

Resumen

Se evaluaron 63 cultivares de mandarina (Citrus spp.) provenientes de la colección del Campo Citrícola Experimental Francisco Villa, Tamaulipas, México, usando marcadores morfológicos y AFLP (Amplified Fragment Length Polymorphism). Se usaron 20 caracteres cuantitativos y 10 cualitativos de hojas, flores y frutos. Las mejores combinaciones de iniciadores AFLP fueron la Mse +CAG más Eco +ACA, y Mse +CAA más Eco +AGG, dando un total de 109 bandas con un 86% de polimorfismo. Tanto los marcadores morfológicos como los moleculares mostraron un alto grado de variación entre los individuos analizados, lo que indica una importante fuente de diversidad genética que puede ser utilizada en futuros programas de mejoramiento genético. Aunque la comparación de los datos morfológicos y moleculares usando la prueba de Mantel no mostró una correlación significativa (r= 0.31), ambas técnicas parecen ser complementarias para la caracterización de mandarinas.

Summary

Sixty-three mandarin (Citrus spp.) cultivars from the collection of the Campo Citrícola Experimental Francisco Villa, Tamaulipas, Mexico, were evaluated using morphological and Amplified Fragment Length Polymorphism (AFLP) markers. Twenty quantitative and 10 qualitative morphological characters from leaves, flowers and fruits were evaluated. The Mse +CAG plus Eco +ACA, and Mse + CAA plus Eco + AGG AFLP primers were the best combinations and generated 109 bands with 86% polymorphism. Both morphological and molecular analysis showed a high degree of variation among analyzed accessions, indicating an important source of genetic diversity that can be used in future breeding programs. Although comparison of morphological and molecular data using the Mantel test did not indicate a significant correlation (r= 0,31), both techniques appeared to be complementary for mandarin characterization.

Resumo

Avaliaram-se 63 cultivares de mandarina (Citrus spp.) da coleção do Campo Citrícola Experimental Francisco Villa, Tamaulipas, México utilizando marcadores morfológicos e AFLP (Amplified Fragment Length Polymorphism). Avaliaram-se 20 caracteres quantitativos e 10 qualitativos em folhas, flores e frutos, os quais mostraram um alto nível de variação entre os materiais avaliados. As combinações Mse +CAG mais Eco +ACA, y Mse +CAA mais Eco +AGG foram as melhores gerando 109 bandas com um polimorfismo de 86%. Tanto a análise morfológica como a molecular mostraram um alto nível de variação indicando a existência de uma fonte importante de diversidade genética que pode ser usada em futuros programas de melhoramento. Mesmo que a comparação dos dados morfológicos e moleculares utilizando a prova de Mantel indicou uma correlação baixa entre estes (r=0,31), ambas as técnicas parecem ser complementarias para a caracterização de mandarinas.

KEY WORDS / Citrus Germplasm / Cluster Analysis / Genetic Similarity / Molecular Markers /

Received: 04/27/2005. Modified: 08/19/2005. Accepted: 09/01/2005.

Introduction

Mandarin (Citrus spp.), together with the grapefruit (C. grandis L.) and citrons (C. medica L.) are considered the three true Citrus species. Mandarin is the second most important citrus plant worldwide. The mandarin group is comprised of numerous species as well as intergeneric and interspecific hybrids which made them the most phenotypically heterogeneous of the three true Citrus (Moore, 2001). Traditionally, morphological characters have been used to identify Citrus; however, there is a high level of genetic variability which can sometimes make an accurate separation for each variety impossible. Although there is a large amount of variability within the Citrus genus with which the breeder can work, and closely related genera provide an even wider selection of characters, there are several barriers to the full utilization of this variability (sterility, incompatibility, nucellar embryony, juvenility). In fact, this genus includes some of the most difficult species to improve.

The exchange of materials for plant improvement is one of the most important purposes of maintaining germplasm collections. Correct classification and identification of accessions in a germplasm bank allows solving management problems such as to avoid duplication in the exchange and in the conservation of the germplasm within the bank, detect mislabeled accessions, certify propagated material and infer the genetic variability that the collection represents in order to increase or maintain an appropriate range of genetic diversity. Accessions may be classified, levels of diversity measured and relationships between individuals or populations may be established using a variety of methods. Molecular markers may extend and complement characterization based on morphological or biochemical descriptions, providing more accurate and detailed information than classical phenotypic data. Some of these techniques are used routinely in the management of germplasm collections of horticultural species (Karp et al., 1997).

A variety of methods have been used to analyze genetic diversity of Citrus cultivars. Isozyme analysis has been used to identify nucellar and zygotic embryos (Torres et al., 1982) and trifoliate orange cultivars (Khan and Roose, 1988; Fang et al., 1997). Restriction Fragment Length Polymorphism (RFLP) markers have been reported to be highly polymorphic in Citrus (Liou et al., 1996), and used to assay genetic relation among Citrus species (Green et al., 1986). With the development of the polymerase chain reaction (PCR), numerous molecular technologies have been developed, which can be used for the detection, characterization and evaluation of the genetic diversity. Random Amplified Polymorphic DNA analysis (RAPD) has been used in the identification of lemon mutants (Deng et al., 1995), chimeras (Sugarawa and Oowada, 1995), somatic hybrids (Guo et al., 2000) and polyembryony (Ramalho et al., 2000; Andrade-Rodríguez et al., 2004). Simple Sequence Repeat (SSR) analysis has been used in phylogenetic and linkage analysis in Citrus and Poncirus (Kijas et al., 1995), mandarin characterization (Koehler-Santos et al., 2003) and grapefruits variability (Corazza-Nunes et al., 2002). Inter-Simple Sequence Repeat (ISSR) analysis has been used in Poncirus (Fang et al., 1997) and Citrus identification (Fang and Roose, 1997). Inter-Retrotransposon Amplified Polymorphism (IRAPs) and Amplified Fragment Length Polymorphism (AFLP) have been used in Clementina mandarin identification (Breto et al., 2001).

Amplified Fragment Length Polymorphism (AFLP; Vos et al., 1995) is a PCR based fingerprinting technique which has been used extensively for studying genetic diversity in different plant species, since it can detect a large number of polymorphisms in a single reaction; and presents a good repeatability, generating primarily dominant markers that are distributed throughout the genome. In fruit crops like grapevine (Cervera et al., 1998; Martínez et al., 2003), papaya (Kim et al., 2002) and coconut (Perera et al., 1998), AFLP has been used successfully in diversity assays.

The goals of this work were to characterize accessions of mandarin (Citrus spp.) using morphological and AFLP markers, in order to evaluate the genetic diversity of these accessions and to detect redundancies inside the collection.

Materials and Methods

Plant material

Sixty-three mandarin accessions from the collection of the Campo Citrícola Experimental Francisco Villa in Tamaulipas, Mexico, were analyzed (Table I). One to three plants were used to represent each accession, and a total of 152 plants were analyzed. The materials were originally imported from Texas in 1974 by CONAFRUT (National Council for Fruit crops). The entire collection comprises 262 different citrus cultivars or citrus relatives. In the present work only those accessions belonging to the mandarin group were analyzed.

Morphological analysis

Twenty quantitative and 10 qualitative characters were evaluated from 10 leaves, 5 flowers and 5 fruits from each plant (Table II). The selection of morphological characters was made by applying the IPGRI Citrus spp descriptor (IPGRI, 2000). Morphological character data were standardized using the YBAR option of the Stand program from the NTSYS-pc 2.1 software (Rohlf, 2000). Duplicate measurements for each specimen were averaged and used to design a data matrix of pairwise similarities between genotypes. The Simple Matching Coefficient (SM) was used to measure similarity, as it was the coefficient with the best results following a cophenetic test (Martínez et al., 2003). Principal Components Analysis (PCA) was used to depict non-hierarchical relationships among the specimens. Eigenvalues and eigenvectors were calculated by the Eigen program using a correlation matrix as input (calculated using standardized morphological data) and 2Dplot were used to generate the two-dimensional PCA plot from the software NTSYS-pc 2.1 (Rohlf, 2000).

DNA isolation

Very young, healthy leaves from each accession were collected, labeled, packed in ice and stored at -80ºC at the Laboratorio de Biotecnología de Frutales, State of México. DNA was extracted using the Saghai-Maroof et al. (1984) protocol with modifications: 2g of leaf tissue were ground to powder in a mortar with liquid N2. Nine ml of the extraction buffer [100mM Tris pH 7.7; 700mM NaCl; 50mM EDTA pH 8.0; 1% CTAB (mixed alkyltrimethyl-ammonium bromide); 140mM b-mercaptoethanol] were added and the mixture was incubated for 60min at 65ºC. 4.5ml chloroform/octanol (24:1) were added and centrifuged at 1500g for 10min at room temperature. The supernatant was transferred to a new tube and the chloroform/octanol step was repeated. Thirty µl of RNase (10mg/ml) were added and the mixture was incubated at 37ºC for 45min. An equal volume of isopropanol was added, mixed and incubated at -20ºC for 15min. The pellet formed after centrifugation for 10min at 1500g at room temperature was washed with 10µl of 5M NaCl and 800µl of absolute ethanol. The DNA pellet was left to dry and dissolved in 200µl of distilled water for storage at 4ºC.

AFLP analysis

AFLP analysis was based on the protocol of Vos et al. (1995). Briefly, 1µg of genomic DNA was digested with two pairs of restriction endonucleases (MseI + EcoRI or MseI + PstI), respectively, for 4h under conditions recommended by the supplier. Restricted DNA fragments were ligated to previously annealed adaptors and preamplification was carried out using EcoRI+1 / Mse1+1 and PstI+1 / MseI+I primers. Twenty four EcoRI+3 / Mse1+3 and PstI+3 / MseI+3 primers combinations were used for selective amplification, the forward primer in each selective amplification reaction was labeled with digoxigenin.

PCR products produced from selective amplification were analyzed on a 6% acrylamide gel run in a 1X TBE buffer on a Bio-Rad sequencing gel apparatus at 100W for 3h, and then were transferred to a nylon membrane previously soaked in an 0.5X TBE. The membrane was dried for 15min at 65ºC and cross linked at 12000µjoules. Dig-labeled fragments were detected using CSPD (disodium 3-(4-methoxispirol {1,2-dioxetane-3,2'(5'-chloro) tricyclo [3.3.1.1^3.7]decan}-4-yl) phenyl phosphate).

AFLP bands were scored based on the presence (1) or absence (0) of polymorphic fragments for each primer, and used to calculate a genetic similarity matrix using the Jaccard (J) coefficient, which is more appropriate for dominant markers as it does not count 0/0 matches in the calculation. The genetic distance between each pair of genotypes was calculated by SIMGEND analysis of the NYSYS-pc software package version 2.1 (Rohlf, 2000). Cluster analysis was performed on both, morphological and molecular, similarity matrices using the unweighted pair group method using arithmetic means (UPGMA) algorithm, from which dendrograms depicting similarity among varieties were drawn and plotted using NTSYS-pc. The cophenetic correlation was calculated in order to find the degree of association between the original similarity matrix and the tree matrix in both morphological and AFLP analysis. Comparison between both methods was performed for the accessions for which both morphological and AFLP data were available by calculating the correlation between the two data sets using the Mantel test in NTSYS-pc.

Results and Discussion

Morphological analysis

The UPGMA dendrogram obtained using morphological characters (Figure 1) separated the mandarin accessions into two main groups, which diverged at a similarity index of 0.41. The larger group, Group 1, contained 41 accessions, mainly from mandarin and Mediterranean groups, and consisted of two subgroups (A and B) and three accessions (Karaji P.I., Kansu and Australiana) which did not fall into a subgroup. Group 2 also formed two subgroups (C and D) and consisted of 18 accessions, mainly hybrids from mandarins with other species such as Tangelos and Tangors (subgroup D) and some mandarins like Satsuma, Clementine, Tamurana, Furuni, Kara, and Taiwanica (subgroup C). Group 2 contained one outlier (Vero Hybrid) which did not fall into either subgroup.

The average genetic similarity among the mandarin accessions was 0.65, with values ranging from 0.10 to 0.90. Accessions Swen Kat and Tinkat showed a very high degree of similarity (0.90), as did accessions Tizon and Bowers (0.89), indicating that these pairs are closely related varieties. On the other hand, the Vero Hybrid and Clark mandarin accessions had the lowest similarity values (0.10). Although there were accessions with similar names (for example, Nucellar Satsuma and Satsuma Owari), they do not appear to be related to each other.

The cophenetic analyses comparing the UPGMA cluster analysis and the simple matching similarity matrix demonstrated a correlation of r= 0.82, indicating that data in the matrix was well represented by the dendrogram. Accessions were clustered mainly based on size of flowers and fruits. Individuals from Group 1 tended to have smaller flowers and fruit than did individuals from Group 2. The classical morphological classification of mandarin divides this species into four principal groups: Satsuma mandarin, common mandarin, Mediterranean ‘Willowleaf’ mandarin, and natural and man-made hybrids (Davies and Albrigo, 1994). However, these four groups were not completely clear in the morphological dendrogram. In Group 1, there were accessions from at least 3 of the 4 main groups listed, while in Group 2 there were mainly accessions from Satsuma and natural and man made hybrids. Characterization of mandarins using morphological characters have been used by several authors (Domingues et al., 1999; Koehler-Santos et al., 2003). Cluster analysis results from these authors and the present one were different; although it is important to consider that some accessions and characters were different in those works.

Principal components analysis (PCA) was used to identify multidimensional relationships among characters for the definition of groups. The first 5 principal components accounted for 56.1% of the total variability and the first two for 37.3% of the variance; 28.8% in the first and 8.6% in the second (Table IV). The first PCA was most highly influenced by traits related with flower and fruit morphology, the second PCA was most influenced by characteristics of the axes and the third PCA was most influenced by leaf characters (Table III). The projection of the 59 mandarin accessions onto the graph defined by the 2 first principal components confirmed the same groups reflected in the dendrogram (Figure 2). Group 1 showed more diversity than Group 2. Tangor and Tangelo accessions were clearly grouped in the graph, along with some of few other mandarin accessions from mandarins. Only Vero Hybrid and Kansu accessions were separated from the rest of the accessions, which was agreed with the subgroup generated from the dendrogram.

AFLP analysis

Level of polymorphism. An initial analysis was carried out with 24 combinations of EcoRI-MseI and PstI-MseI primers with three nucleotide extensions, in order to determine the best combination of enzymes and selective amplification primers. The Mse CAG plus Eco ACA, and Mse CAA plus Eco AGG were the best combinations, based on the number of bands which could be unambiguously scored. A total of 54 and 55 bands were obtained, respectively, similar to what Vos et al. (1995) reported previously. Of the 109 total bands, 94 variable bands were recorded (86% polymorphism).

The mean genetic similarity between each pair of genotypes was 0.75 and ranged from 0.94 between the accessions Robinson and the Tankau, and Kinokuni and Swen-Kat, to 0.21 between Clementine and the Tangor Tarraco. UPGMA cluster analysis of the AFLP genetic similarity matrix resulted in the dendrogram in Figure 3. A comparison of the cophenetic values resulted in a correlation of r= 0.92, indicating that data in the matrix was very well represented by the dendrogram. Similar to the morphological analysis, two main groups were distinguished and they diverged at a genetic similarity coefficient of 0.53. Group 1 (containing two subgroups, A and B) was composed of 42 accessions, mainly from mandarin and Mediterranean groups. Group 2 contained 21 accessions; subgroup C mainly integrated by some mandarins from the Satsuma group and tangelos, and subgroup D composed of the tangores. The subgroups diverged at a genetic similarity of 0.65.

The AFLP primers used in this work were able to discriminate between all mandarin accessions analyzed, including those with the same common name. This high discrimination between (presumably) closely related accessions is the result of the large heterogeneity in the mandarin group. The level of variation in mandarins has been previously reported by several authors using different techniques including isozymes (Torres et al., 1978; Torres et al., 1982), ISSR (Fang and Roose, 1997; Fang et al., 1998), RFLP and RAPD (Federici et al., 1998) and microsatellites (Koehler-Santos et al., 2003). All of them found a large heterogeneity within this group, and all concluded that the mandarin group is the most variable of the three true Citrus species (C. grandis, C. medica and C. reticulata), both when measured with molecular markers and morphologically. Breto et al. (2001) analyzed 24 Clementina mandarin accessions by AFLP and found a low polymorphism level between accessions; however, this result was logical because all cultivars of Clementine mandarins were derived from a single plant and thus the genetic variability level was low.

Comparison between AFLP and morphology. Comparison of matrices of AFLP and morphological data shows a low correlation between dendrograms (r= 0.31, P=1.0) following 500 random permutations with the Mxcomp procedure from the NTSYS program. Despite this low correlation between morphological and molecular analysis, there were similar groups formed in the respective dendrograms. The formation of two main groups was consistently found in both analysis; however, some discrepancies between the two dendrograms can be found. For example, the Vero Hybrid was clearly separated in the morphological analysis, while in the AFLP analysis the same accession was grouped in the C subgroup. Another discrepancy concerns the Karaji P.I., Kansu and Australiana accessions, which did not fall into one of the subgroups in the morphological dendrogram, but in the molecular analysis they clustered closely to other accessions within a subgroup.

The low correlation between AFLP and morphological traits had been reported in other studies in European barley varieties (Schut et al., 1997), synthetic hexaploid wheats and their parents (Lage et al., 2003), and Squash germplasm (Ferriol et al., 2004). Similar results were found in mandarins by Koehler-Santos et al. (2003), who detected differences between dendrograms generated from morphological and SSR data, and they suggest that morphological and molecular differences are apparently independent, due to different selection and evolutionary factors. Although AFLP markers can cover a high proportion of the genome because of the high number of bands scored in each analysis, due its neutral origin, there is no guarantee that such bands fall in coding regions of the genome involved in morphological and agronomic traits.

Traditionally, germplasm has been classified on the basis of morphological and agronomical traits, but recently the use of molecular markers to study diversity and characterization in plants has become common. In this study the characterized accessions were mainly grouped according to flower and fruit morphology, which are complex and multigenic characters. Such characters are environmentally affected and therefore liable to subjective evaluation. In this sense, the molecular characterization is more efficient in the generation of an unbiased picture of diversity than an agronomic approach. However, the agronomic characterization is still important in germplasm management, and determination of molecular diversity should not be seen as replacing traditional characterization but rather as a complement to it.

Although the correlation between the morphological and AFLP data was low, both methods allowed grouping the mandarin accessions analyzed in this work. Despite the fact that morphological traits were relatively less efficient for precise discrimination of closely related genotypes, the cost and time invested were lower than for the molecular analysis. Therefore, based on cost, efficiency, and information gained, both techniques appeared complementary in mandarin characterization.

Although the characterization of mandarin germplasm has been documented in several works, this is the first time that the mandarin accessions from the Campo Citrícola Experimental Francisco Villa citrus collection were characterized using molecular markers since it was created in 1974. This is one of the largest collections in México, and includes not only accessions with commercial characteristics (such as Satsuma and Clementina) but also accessions used as rootstocks (Sunki mandarin and Citrus amblicarpa) and as sources of disease resistance (Citrus depressa). Despite the high diversity present in the mandarin group, only a small number of cultivars are used commercially, and so the exploitation of the genetic diversity in the in situ collections to support breeding programs and other research efforts is important. The characterization information of the broad genetic variation found in this work can be used in germplasm management, variety protection, and new efforts in citrus improvement.

Conclusions

Both morphological and molecular markers show a high degree of variation among the mandarin accessions analyzed. These accessions should represent an important source of genetic diversity in citrus and can be used in future breeding programs; there were no redundancies inside the collection. Both dendrograms separate the mandarin accessions analyzed in two main groups and four subgroups with some differences. Although the correlation between morphological and AFLP data is low, both techniques can be used complementarily in mandarin characterization.

REFERENCES

1. Andrade-Rodríguez M, Villegas-Monter A, Carrillo-Castañeda G, García-Velásquez A (2004) Polyembryony and identification of volkamerian lemon zygotic and nucellar seedlings using RAPD. Pesq. Agropec. Bras. 39: 551-559.        [ Links ]

2. Breto MP, Ruiz C, Pina JA, Asíns MJ (2001) The diversification of Citrus clementina Hort. Ex Tan., a vegetatively propagated crop species. Mol. Phylogen. Evol. 21: 285-293.        [ Links ]

3. Cervera MT, Cabezas JA, Sancha JC, Martínez F, Martínez JM (1998) Application of AFLPs to the characterization of grapevine Vitis vinifera L. genetic resources. A case study with accessions from Rioja (Spain). Theor. Appl. Genet. 97: 51-59        [ Links ]

4. Corazza-Nunes MJ, Machado MA, Nunes WMC, Cristofani M, Targon MLPN (2002) Assesment of genetic variability in grapefruits (Citrus paradisi Macf.) and pummelos (C. maxima (Burm.) Merr.) using RAPD and SSR markers. Euphytica 126: 169-176.        [ Links ]

5. Davies FS, Albrigo LG (1994) Crop production Science in Horticulture. 2: Citrus. CAB International. Wallingford. UK. 254 pp.        [ Links ]

6. Domingues ET, Souza VC; Sakuragui CM, Pompeu Junior J, Teófilo Sobrinho J, Souza JP (1999) Caracterizacao morfológica de tangerinas do banco ativode germoplasma de citros do Centro Citricola Sylvio Moreira/IAC. Sci. Agric, Piracicaba, 56: 197-206.        [ Links ]

7. Deng ZN, Gentile A, Nicolosi E, Vardi A, Tribulato E (1995) Identification of in vivo and in vitro lemon mutants by RAPD markers. J. Hort. Sci. 70: 117-125.        [ Links ]

8. Fang D, Roose ML (1997) Identification of closely related citrus cultivars with inter-simple sequence repeats markers. Theor Appl. Genet. 95: 408-417.        [ Links ]

9. Fang D, Roose ML, Krueger RR, Federici CT (1997) Fingerprinting trifoliate orange germplasm accessions with isozymes, RFLP’s, and inter simple sequence repeat markers. Theor Appl. Genet. 97: 211-219.        [ Links ]

10. Fang D, Krueger RR, Roose ML (1998) Phylogenetic relationships among selected citrus germplasm accessions revealed by Inter.-simple sequence repeat (ISSR) markers. J. Amer. Soc. Hort. Sci. 123: 612-617.        [ Links ]

11. Federici CT, Fang DQ, Scora RW (1998) Phylogenetic relationships within de genus Citrus (Rutaceae) and related genera as revealed by RFLP and RAPD analysis. Theor Appl Genet. 96: 812-822.        [ Links ]

12. Ferriol M, Picó B, Fernandez deCP, Nuez F (2004) Molecular diversity of germplasm collection of squash (Cucúrbita moschata) determined by SRAP and AFLP markers. Crop. Sci. 44: 653-664.        [ Links ]

13. Green RM, Vardi A, Galun E (1986) The plastome of citrus cultivars and species a comparison with related genera. Theor. Appl. Genet. 72: 170-177.        [ Links ]

14. Guo WW, Deng XX, Yi HL (2000) Somatic hybrids between navel orange (Citrus sinensis) and grapefruit (C. paradisi) for seedless triploid breeding. Euphytica 116: 281-285.        [ Links ]

15. IPGRI (2000) Descriptores para los cítricos Citrus spp. Instituto Nacional de Recursos Fitogenéticos. Roma, Italia. 75 pp.        [ Links ]

16. Kim MS, Moore PH, Zee F, Fitch MM, Steiger DL, Manshardt RM, Paull RE, Drew RA, Sekioka T, Ming R (2002) Genetic diversity of Carica papaya as revealed by AFLP markers. Genome 45: 503-512.        [ Links ]

17. Karp A, Kresovich S, Bhat KV, Ayad WG, Hodgkin T (1997) Molecular tools. In: plant genetic resources conservation: a guide to the technology. IPGRI Bulletin 2. 47 pp.        [ Links ]

18. Khan IA, Roose ML (1988) Frequency and characteristics of nucellar and zygotic seedling in three cultivars of trifoliate orange. J. Am. Soc. Hort. Sci. 113: 105-110.        [ Links ]

19. Kijas JMH, Fowler JCS, Thomas MR (1995) An evaluation of sequence-tagged microsatellite site markers for genetic analysis within Citrus and related species. Genome 38: 349-355.        [ Links ]

20. Koehler-Santos P, Dornelles ALC, de Freitas LB (2003) Characterization of mandarin citrus germplasm from southern Brazil by morphological and molecular analyze. Pesq. Agropec. Bras. 38: 797-806.        [ Links ]

21. Lage J, Warburton ML, Crossa J, Skovmand B, Andersen SB (2003) Assessment of genetic diversity in synthetic hexaploid wheat and their Triticum dicoccum and Aegilops tauschii parents using AFLPs and agronomic traits. Euphytica 134: 305-317.        [ Links ]

22. Liou PC, Gmitter FG, Moore GA (1996) Characterization of the citrus genome through analysis of restriction fragment length polymorphism. Theor. Appl. Genet. 92: 425-435.        [ Links ]

23. Martínez, L, Cavagnaro P, Masuelli R, Rodríguez J (2003) Evaluation of diversity among Argentine grapevine (Vitis vinifera L.) varieties using morphological data and AFLP markers. Electr. J. Biotechnol. 16(3): 11.         [ Links ]

24. Moore GA (2001) Oranges and lemons: clues to the taxonomy of citrus from molecular markers. TRENDS in Genetics 17: 536-540.        [ Links ]

25. Perera L, Rusell R, Provan J, McNicol J, Powell W (1998) Evaluating genetic relationships between indigenous coconut accessions from Sri Lanka by mean of AFLP profiling. Theor. Appl. Genet. 96: 545-550.        [ Links ]

26. Ramalho LD, Duarte VA, Dos Santos SFW, De Jesus BC (2000) Identificacao de seedlings zigóticos de citros mediante o uso de marcadores RAPD. Rev. Bras. Frutic. 22: 181-185.        [ Links ]

27. Rohlf FJ (2000) NTSYS-pc: Numerical Taxonomy and Multivariate Analysis System. Version 2.1. Exceter Sofware, New York, USA.        [ Links ]

28. Saghai-Maroof MA, Soliman K, Jorgensen RA, Allard RW (1984) Ribosomal DNA spacer length polymorphism in barley: Mendelian inheritance, chromosomal location, and population dynamics. PNAS 81: 8014-8018.        [ Links ]

29. Schut JW, Qi X, Stam P (1997) Association between relationships measures based on AFLP markers, pedigree data and morphological traits in barley. Theor. Appl. Genet. 95: 1161-1168        [ Links ]

30. Sugarawa K, Oowada A (1995) Identification of Citrus chimeras by RAPD analysis. Hort. Science 30: 1276-1278.        [ Links ]

31. Torres AM, Soost RK, Mau-Lastovicka T (1978) Leaf isozymes as genetic markers in citrus. Amer. J. Bot. 65: 869-881.        [ Links ]

32. Torres AM, Soost RK, Mau-Lastovicka T (1982) Citrus isosymes. J. Hered. 73: 335-339.        [ Links ]

33. Vos P, Hogers R, Bleeker M, Reijans M, de Lee TV, Hornes M, Frijters A, Pot J, Peleman J, Kuiper M, Zabeau M (1995) AFLP: a new technique for DNA fingerprinting. Nucl. Acids Res. 2: 4407-4414.        [ Links ]