Agronomic and molecular analyses for the characterisation of accessions in Tunisian olive germplasm collections
Financial support: The present work has been done in the framework of a research programme (2002-2005) funded by the Tunisian Ministry of Science Research, Technology and Competency Development.
Keywords: agronomic traits, AFLP, Olea europaea, SSR.
In this study, a total of 14 agronomic traits, five AFLP primer combinations and ten SSR loci were used to describe and to classify a group of Tunisian olive genotypes into groups based on molecular profiles and agronomic traits. The analysis of variance of the agronomical data revealed significant differences among accessions for all measured traits. The mean phenotypic dissimilarity (0.34 with a range of 0.08-0.6) was low in comparison to dissimilarity calculated using AFLP (0.50 with a range of 0.16-0.70) and SSR markers (0.76 with a range 0.35-0.94). The correlation between the agronomical dissimilarity matrix and the matrices of genetic dissimilarity based on SSR and AFLP markers was very weak: 0.156 (p = 0.05) and 0.185 (p = 0.05), respectively. The SSR-AFLP dendrogram based on unweighted pair-group cluster analysis using Jaccard’s index revealed that the genetic diversity was predominantly structured according to fruit size. A trend of clustering together of accessions originating from the same or adjacent regions was also observed. The data obtained can be used for the varietal survey and construction of a database of all olive varieties grown in Tunisia and providing also additional information that could form the basis for the rational design of breeding programs.
The olive tree (Olea europaea L.) is a subtropical species typical of the Mediterranean basin where it represents the most important oil-producing crop. It is a diploid, out crossing species with a very wide genetic patrimony. Since the beginning of its domestication, olive has been propagated vegetatively to exploit the best combination of genes which arose by random crosses or mutations (Carriero et al. 2002). As a result, a great number of varieties are present in all the countries where this species is cultivated, raising several problems for germplasm management and preservation (Carriero et al. 2002). Evaluation and characterisation of olive genetic resources is therefore crucial, since identification of olive cultivars is complicated by the large number of varietal synonyms and homonyms, the intensive exchange of plant material, the presence of varietal clones, and problems of varietal certification in nurseries (Bandelj et al. 2002).
Morphological and biological characters have been widely used for descriptive purposes and are commonly used to distinguish olive cultivars (Cantini et al. 1999). Agronomic characterization also allowed the classification of different olive cultivars (Barranco and Rallo, 2000). In order to supplement and refine the initial phenotypical based descriptions, different genetic markers such as randomly amplified polymorphic DNA (RAPD) markers (Fabbri et al. 1995; Wiesman et al. 1998; Belaj et al. 2001; Besnard et al. 2001b; Sanz-Cortés et al. 2001; Nikoloudakis et al. 2003), amplified fragment length polymorphisms (AFLP) (Angiolillo et al. 1999; Rotondi et al. 2003; Sanz-Cortés et al. 2003; Sensi et al. 2003), microsatellites (SSR) (Carriero et al. 2002; Cipriani et al. 2002) and intermicrosatellites (ISSR) (Hess et al. 2000) have been applied to identify and characterize different cultivars or genotypes and in order to provide information on olive origin and dispersal.
and molecular analyses were performed on 26 olive (Olea europaea
L.) accessions: 25 accessions belonging to 25 cultivars and one accession
without denomination. These were obtained from four collections maintained
in experimental orchards at the ‘Institut de l’Olivier’ (IO), ‘Institut
National des Sciences Appliquées et de
agronomic and biochemical characteristics were measured or analyzed
as mean of 2003 and 2004 (Table 2). At the mature stage, 200
olives (3 replicates per genotype) were picked randomly and then split
in two samples, one put in oven to dry, the other used for fresh fruit
and pit analysis and fatty acids determination. Average fruit weight
was determined and, after removing and cleaning the stones, flesh
and stone weights were also recorded. Oil content was determined by
extracting the dry material with 40–60°C petroleum ether using a Soxhlet
apparatus. Olive oil used for the qualitative analysis was extracted
from fresh material using methanol-chloroform (1:1, v/v) according
to Allen and Good (1971). Fatty acids were determined,
as methyl ester, by gas chromatography. Methylesters were released
by transesterefication with BF3/methanol after saponification with
methanolic KOH. The fatty acid pattern was assessed with a HP 4890
D (Agilent Technologies) equipped with a split-splitless injector,
a FID detector and a
hundred milligrams of young leaves were collected, rinsed with tap
water, frozen in liquid nitrogen, ground to a fine powder in a reciprocal
grinding apparatus (Mixer Mill MM300, Retsch,
Haan, Germany) and kept at
developed primer pairs for olive microsatellite loci (Sefc
et al. 2000) were selected for their polymorphism and the clarity
of their electrophoretic profiles. The loci amplified by these primer
pairs were designated as: ssrOeUA-DCA1, ssrOeUA-DCA3, ssrOeUA-DCA4,
ssrOeUA-DCA7, ssrOeUA-DCA9, ssrOeUA-DCA11 ssrOeUA-DCA13, ssrOeUA-DCA15,
ssrOeUA-DCA16 and ssrOeUADCA18. PCR reactions were performed in a
total volume of 20 µL containing
AFLP protocol was performed according to the procedure described by
Vos et al. (1995). 500 ng of genomic DNA was double
digested with EcoRI and MseI enzymes (2.5 U each) and linked to adapters
(50 and 5 pmols of MseI and EcoRI adapters, respectively). Restricted
and ligated DNA (25 ng) was then pre-amplified using EcoRI and MseI
primers (25 ng) with one selective nucleotide. PCR pre-amplifications
were prepared in a total volume of 25 μl and amplification was
carried out in a PTC-100 thermal cycler (MJ Research Inc., Watertown,
MA, USA), using 20 cycles of
Analysis of variance was performed for all measured traits in order to test the significance of variation among accessions. The standardized traits mean values (mean of each trait was subtracted from the data values and the result divided by the standard deviation) were used to perform principal component (PCA) and cluster analyses using XLSTAT software. To group the accessions based on agronomical dissimilarity, cluster analysis was conducted on the Euclidean distance matrix with the unweighted pair group method based on arithmetic averages (UPGMA).
For molecular diversity analysis, AFLP and microsatellite results were scored for presence (1) or absence (0) of amplified markers. Genetic distance was calculated on the basis of Jaccard’s coefficient method (Jaccard, 1908). The combined SSR-AFLP similarity matrix was subjected to cluster analysis by UPGMA. The individual SSR-AFLP genotypes scores were plotted in a bidimensional space using principal coordinates analysis (PCoA) by computing the genetic distance matrix using the GENSTAT software, 8th edition. The relationships between the Euclidean distance matrix based on agronomic traits and genetic distance matrices obtained with SSR and AFLP markers were analyzed according to Mantel (1967).
The analysis of variance revealed significant differences among accessions for all of the studied traits (Table 3). The average fruit fresh and dry weights, flesh and pit fresh weights and palmitoleic acid percentage showed wide variation, while fruit oil content on dry weight basis and oleic acid percentage showed a narrower range of phenotypic variation. The 14 studied variables were analysed by PCA (data not shown). The eigenvalues obtained by PCA on the agronomic data indicate that two to three components provide a good summary of the data. In particular, the first two components (PC1 and PC2) accounted for 56.7% of total variance, and three components explained 68.8%. The other components contribute less than 10% each. Variables such as fruit fresh and dry weights, flesh and pit fresh weights (on PC1), palmitic and oleic acids percentages and fruit oil content on fresh and dry weight basis (on PC2) explained the largest portion of the variance. In the third PC, which explained 12.1% of total variation, the predominating traits were fatty acid percentages (excluding oleic acid).
The two molecular approaches used in this study could uniquely fingerprint each of the 26 olive accessions. The 5 primer combinations used to perform the AFLP analysis yielded a total of 418 bands with a percentage of polymorphism of 56.46% (Table 4). Only 129 well-defined bands were analysed in the whole set of data. A typical AFLP band pattern for the 26 accessions is shown in Figure 1a. Microsatellites were successfully amplified in all the analysed accessions with the ten primer pairs used. Patterns generated by primer pair ssrOeUA-DCA18 in accessions are shown in Figure 1b. All ten microsatellite markers were polymorphic across the screened genotypes, revealing a total of 86 alleles. The number of alleles for each locus varied from four at locus ssrOeUA-DCA15 to fourteen at locus ssrOeUA-DCA4, with an average of 8.6 (Table 5).
A histogram of pair wise dissimilarity for the 26 Tunisian olive accessions generated from SSR, AFLP and agronomic data is presented in Figure 2. The dissimilarity coefficients based on agronomic traits ranged from 0.08 to 0.60 with an average of 0.34. Based on SSR, these values ranged from 0.35 to 0.94 with an overall mean of 0.76. For AFLP, it ranged from 0.16 to 0.70 with an overall mean of 0.50.
To compare the extent of agreement between dendrograms derived from agronomic, AFLP and SSR data, a distance matrix was constructed for each assay and compared using the Mantel matrix correspondence test. The estimated correlation for the two molecular systems was significant but relatively low (r = 0.3, p = 0.001). The AFLP and SSR data were poorly correlated with the agronomic data (r =0.185, p = 0.05 and r = 0.156, p = 0.05, respectively).
dendrogram generated from the standardized phenotypic data is presented
in Figure 3. The UPGMA cluster analysis revealed
three main groups. Group 1 consisted of ‘Meski’, ‘Injassi’, ‘Besbassi’
and ‘Marsaline’, four accessions featuring large-sized fruits (4.95-
dendrogram generated based on a combined SSR and AFLP data set (Figure
4) showed three main groups: Group A, including 6 accessions,
Group B, including 5 accessions and Group C, with 13 accessions, while
‘Sayali’ and ‘Jemri Dhokkar’ showed the lowest similarity to all accessions
and were set apart from the three clusters. Group A consisted of six
Southern accessions originating from the arid regions of Sfax (‘Semni’,
‘Kbiret Louzir’, ‘Unknown’, ‘Chemlali’ and ‘Kchinet Sig’) and Zarzis
(‘Dhokkar’). All accessions in this cluster have small-sized fruits
and are typically used for oil production. Group B comprised five
small-fruited accessions used for oil production. Two of them originate
from the North of Tunisia (‘Chétoui’ and ‘Rakhami’); two (‘El Hor’
and ‘Oueslati’) originate from the semiarid zones of
The principal coordinates analysis (Figure 5) where the first two principal components accounted for 33.5% of the variance, seems to support the results obtained by cluster analysis. The pattern shown in Figure 4 was comparable to the clustering observed in the UPGMA dendrogram (Figure 5). With the exception of ‘Sayali’, genotypes that have either medium or large-sized fruits were scattered separated from the small-fruited genotypes. Again with one exception (‘Jdallou’), the PCoA separated the Southern oil producing accessions from those originating from the North and the Centre of Tunisia.
In this study, we used AFLP and SSR markers and agronomic traits to characterize a set of 26 Tunisian olive accessions. Polymorphism was evident for all three marker systems. This result is consistent with results from previous studies carried out on olive cultivars (Fabbri et al. 1995; Wiesman et al. 1998; Angiolillo et al. 1999; Baldoni et al. 2000; Rallo et al. 2000; Belaj et al. 2001; Besnard et al. 2001b; Sanz-Cortés et al. 2001), thereby confirming the great diversity within the cultivated olive germplasm. The distribution of values for agronomic dissimilarity and genetic dissimilarity (calculated with SSRs and AFLPs) differed markedly. The mean agronomic dissimilarity (0.34 with a range of 0.08-0.6) was low in comparison to dissimilarity calculated using AFLP (0.50 with a range of 0.16-0.70) and SSR markers (0.76 with a range 0.35-0.94). This data suggest that SSR and AFLP markers can better differentiate pairs of accessions than agronomic traits that show a low level of genetic variation. Comparing the two marker types, a higher level of polymorphism was obtained for SSR than for AFLP (Table 6) which highlights the discriminating power of the former. This result is in accordance with previous studies where SSRs were compared to other marker systems (Powell et al. 1996; Pejic et al. 1998; Belaj et al. 2003). The high variability observed at SSR loci was expected because of the unique mechanism by which this variation is generated: replication slippage is thought to occur more frequently than single nucleotide mutations and insertion/deletion events, which generate the polymorphisms detectable by AFLP analysis (Powell et al. 1996; Milbourne et al. 1997).
To provide an objective comparison, we examined correlations between distance matrices calculated on the basis of AFLP, SSR and agronomical data using a Mantel matrix correspondence test. The estimated correlation similarities for the two molecular systems was significant but relatively low. The type of genetic polymorphism detected by the two markers and the number of primers used may affect the correlations among them. The correlation between the two molecular markers was higher than the agronomy. When compared with DNA fingerprinting techniques, agronomic traits are relatively less reliable and efficient for precise discrimination of closely related accessions and analysis of their genetic relationships. Despite this limitation, numerical analysis of olive agronomic traits can be used as a general approach to establish from a practical viewpoint a first order of accessions classification within germplasm collections; it enables accession comparison and diversity conservation. Although both marker methods did not provide exactly the same description of relationships between the analysed accessions, there was some consistency. The best agreement between the two methods was present for accession pairs that were very distant (e.g. ‘Besbassi’-‘Chemlali’, ‘Besbassi’-‘Jemri Dhokkar’, ‘Zarrazi’-‘Kchinet Sig’) or very close (e.g. ‘El Hor’- ‘Jdallou’ and, to a lesser extent, ‘Chemlali’-‘Kchinet Sig’, ‘Meski’-‘Besbassi’).
UPGMA clustering and the PCoA Plot obtained from the AFLP-SSR distance
matrix showed a rather high variability among the accessions examined
and that most of Tunisian olive accessions clustered according to
their fruit size. For instance small-fruited accessions clustered
in Group A and B. Accessions that have medium to large sized-fruits
clustered in Group C. Genetic differentiation based on fruit size
and use has been observed in previous studies. Grati-Kamoun
et al. (2006), in their AFLP study, obtained a clustering of olive
cultivars into two main groups according to fruit size. Interestingly,
of the 29 cultivars included in their study, ten Tunisian cultivars
are also included here. Grati-Kamoun et al. (2006)
found a comparable grouping pattern among ‘Jemri Dhokkar’, ‘Chemlali’,
and ‘Oueslati’ that were grouped with the small-sized cultivars and
‘Marsaline’, ‘Meski’, and ‘Besbassi’ that were in the cluster of the
large-fruited cultivars. The same above-mentioned work showed that
‘Zarrazi’, ‘Chemchali’ and ‘Toffahi’, three medium-fruited cultivars
from Southern Tunisia, clustered with the small-sized olives originating
from the same part of the country while in our assay, these cultivars
were not well separated from those used for canning. The presence
of table olives from other Mediterranean countries in Grati-Kamoun
et al. (2006) study could change the clusters of these cultivars
in the Jaccard’s dendrogram. In a study based on RAPD markers, (Fabbri
et al. 1995) seventeen olive cultivars clustered into two main
groups according to fruit size and oil content. Using the same technique,
Nikoloudakis et al. (2003) reported that most of
Greek cultivars clustered according to their fruit size or commercial
use. Another group (Loukas and Krimbas, 1983) studied
22 Greek cultivars based on allozyme markers and found clustering
according to fruit size rather than to geographic origin. Wild and
feral olives are characterized by very small fruits. The relationship
between electrophoretic profiles and the common utilisation of fruits
(oil, canning, or both uses) may be due either to a single origin
of varieties with big fruits or to their less-close proximity with
wild populations due to a stronger or longer selection towards fruit
size (Besnard et al. 2001a). A trend of clustering
of cultivars originating from the same or adjacent regions was also
detected. With the exclusion of ‘Jdallou’, Southern oil producing
cultivars clustered separated from those originating from the North
and the Centre of Tunisia. This was also the case for the pairs of
northern cultivars, ‘Gerboui’-‘Neb Djemel’ and ‘Meski’-‘Besbassi’
which feature medium- and large-sized fruits, respectively. ‘Toffahi’,
‘Chemchali’ and ‘Injassi’, three accessions originating from the South,
formed a small separate cluster, which does not include the accession
‘Zarrazi’. The three first accessions come from interior regions (Gafsa
and Tataouine) where weather conditions are different from coastal
regions such as Zarzis from where originated the accession ‘Zarrazi’.
Clustering of cultivars according to their geographic origin was also
observed in a larger geographic scale study performed by Belaj
et al. (2001) with cultivars from several countries of the
ALLEN, C. Freeman and GOOD, Pearl. Acyl lipids in photosynthetic systems. Methods in Enzymology: Photosynthesis and Nitrogen Part A, 1971, vol. 23, p. 523-547. [CrossRef]
ANGIOLILLO, A.; MENCUCCINI, M. and BALDONI, L. Olive (Olea europaea L.) genetic diversity assessed by amplified fragment length polymorphisms. Theoretical and Applied Genetics, March 1999, vol. 98, no. 3-4, p. 411-421. [CrossRef]
BALDONI, L.; PELLEGRINI, M.; MENCUCCINI, M.; MULAS, M. and ANGIOLILLO, A. Genetic relationships among cultivated and wild olives revealed by AFLP markers. Acta Horticulturae, January 2000, vol. 521, no. 11, p. 275-283.
A.; TRUJILLO, I.; DE
BELAJ, A.; SATOVIC, Z.; CIPRIANI, G.; BALDONI, L.; TESTOLIN, R.; RALLO, L. and TRUJILLO, I. Comparative study of the discriminating capacity of RAPD, AFLP and SSR markers and of their effectiveness in establishing genetic relationships in olive. Theoretical and Applied Genetics, August 2003, vol. 107, no. 4, p. 736-744. [CrossRef]
BESNARD, Guillaume; BARADAT, Philippe and BERVILLE, André. Genetic relationships in the olive (Olea europaea L.) reflect multilocal selection of cultivars. Theoretical and Applied Genetics, February 2001a, vol. 102, no. 2-3, p. 251-258. [CrossRef]
BESNARD, Guillaume; BRETON, Catherine; BARADAT, Philippe; KHADARI, Bouchaib and BERVILLE, André. Cultivar identification in olive based on RAPD markers. Journal of the American Society for Horticultural Science, November 2001b, vol. 126, no. 6, p. 668-675.
Claudio; CIMATO, Antonio and SANI, Graziano. Morphological evaluation
of olive germplasm present in
CARRIERO, F.; FONTANAZZA, G.; CELLINI, F. and GIORIO, G. Identification of simple sequence repeats (SSRs) in olive (Olea europaea L.) Theoretical and Applied Genetics, February 2002, vol. 104, no. 2-3, p. 301-307. [CrossRef]
CIPRIANI, G.; MARRAZZO, M.T.; MARCONI, R.; CIMATO, A. and TESTOLIN, R. Microsatellite markers isolated in olive are suitable for individual fingerprinting and reveal polymorphism within ancient cultivars (Olea europaea L.) Theoretical and Applied Genetics, February 2002, vol. 104, no. 2-3, p. 223-228. [CrossRef]
CLAROS, M. Gonzalo; CRESPILLO, Remedios; AGUILAR, Maria L. and CANOVAS, Francisco M. DNA fingerprinting and classification of geographically related genotypes of olive-tree (Olea europaea L.) Euphytica, November 2000, vol. 116, no. 2, p. 131-142. [CrossRef]
FABBRI, A.; HORMAZA, J.I. and POLITO, V.S. Random amplified polymorphic DNA analysis of olive (Olea europaea L.) cultivars. Journal of the American Society for Horticultural Science, May 1995, vol. 120, no. 3, p. 538-542.
GEUNA, F.; TOSCHI, M. and BASSI, D. The use of AFLP markers for cultivar identification in apricot. Plant Breeding, December 2003, vol. 122, no. 6, p. 526-531. [CrossRef]
N.; MAHMOUD, F. Lamy; REBAI, A.; GARGOURI, A.; PANAUD, O. and
HESS, J.; KADEREIT, J.W. and VARGAS, P. The colonization history of Olea europaea L. in Macaronesia based on internal transcribed spacer 1 (ITS-1) sequences, randomly amplified polymorphic DNAs (RAPD), and intersimple sequence repeats (ISSR) Molecular Ecology, July 2000, vol. 9, no. 7, p. 857-868. [CrossRef]
MILBOURNE, Dan; MEYER, Rhonda; BRADSHAW, John E.; BAIRD, Eileen; BONAR, Nicky; PROVAN, Jim; POWELL, Wayne and WAUGH, Robbie. Comparison of PCR-based marker systems for the analysis of genetic relationships in cultivated potato. Molecular Breeding, April 1997, vol.3, no. 2, p. 127-136. [CrossRef]
N.; BANILAS, G.; GAZIS, F.; HATZOPOULOS, P. and METZIDAKIS, J. Discrimination
and genetic diversity among cultivated olives of
RALLO, P.; DORADO, G. and MARTIN, A. Development of simple sequence repeats (SSRs) in olive tree (Olea europaea L.) Theoretical and Applied Genetics, October 2000, vol. 101, no. 5-6, p. 984-989. [CrossRef]
ROTONDI, A.; MAGLI, M.; RICCIOLINI, C. and BALDONI, L. Morphological and molecular analyses for the characterization of a group of Italian olive cultivars. Euphytica, July 2003, vol. 132, no. 2, p. 129-137. [CrossRef]
SANZ-CORTÉS, F.; BADENES, M.L.; PAZ, S.; IÑIGUEZ, A. and LLÁCER, G. Molecular characterization of olive cultivars using RAPD markers. Journal of the American Society for Horticultural Science, January 2001, vol. 126, no.1, p. 7-12.
SANZ-CORTÉS, F.; PARFITT, D.E.; ROMERO, C.; STRUSS, D.; LLÁCER, G. and BADENES, M.L. Intraspecific olive diversity assessed with AFLP. Plant Breeding, April 2003, vol. 122, no. 2, p. 173-177. [CrossRef]
SEFC, K.M.; LOPES, M.S.; MENDONÇA, D.; RODRIGUES DOS SANTOS, M.; DA CAMARA MACHADO, M. Laimer and DA CAMARA MACHADO, A. Identification of microsatellite loci in olive (Olea europaea) and their characterisation in Italian and Iberian olive trees. Molecular Ecology, August 2000, vol. 9, no. 8, p. 1171-1193. [CrossRef]
SENSI, E.; VIGNANI, R.; SCALI, M.; MASI, E. and CRESTI, M. DNA fingerprinting and genetic relatedness among cultivated varieties of Olea europaea L. estimated by AFLP analysis. Scientia Horticulturae, February 2003, vol. 97, no. 3-4, p. 379-388. [CrossRef]
VOS, Pieter; HOGERS, Rene; BLEEKER, Marjo; REIJANS, Martin; VAN DE LEE, Theo; HORNES, Miranda; FRITERS, Adrie; POT, Jerina; PALEMAN, Johan; KUIPER, Martin and ZABEAU, Marc. AFLP: A new technique for DNA fingerprinting. Nucleic Acids Research, January 1995, vol. 23, no. 21, p. 4407-4414.
WIESMAN, Z.; AVIDAN, N.; LAVEE, S. and QUEBEDEAUX, B. Molecular characterization of common olive varieties in Israel and the West Bank using randomly amplified polymorphic DNA (RAPD) markers. Journal of the American Society for Horticultural Science, September 1998, vol. 123, no. 5, p. 837-841.
Note: Electronic Journal of Biotechnology is not responsible if on-line references cited on manuscripts are not available any more after the date of publication.
Home | Mail to Editor | Search | Archive