Comparison between classical and Bayesian methods to investigate the history of olive cultivars using SSR-polymorphisms
Introduction
It is important to clarify a crop's connection to its wild ancestors for improved efficiency in the management of genetic resources and for revealing cultivar origins. For botanists the olive is Olea europaea subsp. europaea var. europaea, whereas oleaster is var. sylvestris [1]. The olive is cultivated whereas the oleaster is wild. Several hundred geographically diverse olive cultivars exist in the Mediterranean basin. They are distinguished by leaf morphology, drupe shape and colour, oil composition and phenology. The main descriptive points correspond to 42 characteristics on leaf, fruit and stone morphology, for their identification [2].
The current designations for olive cultivars come from a mixture of generic names. These are based on the colouring of drupes from green to black (‘Rougette’, ‘Blanquette’, ‘Noirette’, ‘Verdale’), place names, sometimes reference to drupe size (‘Grossane’), their shape (Amygdalolia), as well as their use for oil production (‘Frantoio’). Throughout history records for olive cultivar names have not reflected an accurate description of drupes. Consequently, it is uncertain whether a name such as ‘Picholine’, which has been in use since at least 1712, refers to a singular cultivar or a group of cultivars used as a table olive. It has already been shown that the same cultivar (a clone) may have different names in different countries and languages all around the Mediterranean basin, so the name of a cultivar cannot guarantee its geographic origin [3]).
However, the difference between the wild and cultivated olive is unclear regarding drupe morphology [4] and DNA molecular markers [5]. The olive may escape from cultivation and return to an apparently wild state [6], [7], [8]. In most countries some cultivars bear small-sized fruit, i.e., ‘Arbequina’ in Spain, ‘Cailletier’ in Continental France and ‘Sabina’ in Corsica (France). Without care, old olive trees from many cultivars may look like oleasters.
It is recognised that the olive is derived from the oleaster [6], [9], but where and how olive cultivars were created remains undocumented. Olive growers have different views on olive origins but none are documented although olive sales are boosted by informal but informed website histories. The general belief is that olive cultivars were identified in the east of the Mediterranean basin (Near East) and that the Phoenicians, Greeks and Romans brought them further west [10]. The olive was introduced into the western Mediterranean basin and documented [11], [12]. Numerous websites state that oleasters in the western Mediterranean basin were simply those which escaped from cultivation and became wild. This belief is not documented and has been rejected [5].
Different olive cultivar classifications have been considered in the literature based upon a variety of characters. Genetic distances have been estimated from archaeological remains [13]. Morphological characters such as drupe shape [14], stone morphology [15], and fruit size [16]. Geographic situation has also been used and lead to 13 clusters from olives obtained from different countries [17]. Distinct groups from the East have been described [5], [6], [18], [19], West, and from other countries. One found that locally cultivated oleasters differed from those introduced later [20]. Olive uses either for oil, table or mixed uses were grouped using molecular markers [6]. A polyphyletic origin of the olive based on three repeated Olea sequences which are distributed throughout the Mediterranean basin [21] has been suggested. Each of these methods has brought information on relationships but not on their eventual origins.
An approach was developed based on the Bayesian clustering and admixture analysis [22] used in population genetics, to establish olive cultivar origins in oleaster populations [23]. In contrast to all previous methods that are merely based on similarities, the software can infer direct genetic relationships based on molecular data. Individual oleasters have been previously studied and all the methods used to reconstruct seven Glacial Refugee Populations (GRPs) for oleaster trees have been detailed [23]. Depending on the Factorial Correspondence Analysis (FCA), Fst, and Bayesian methods used, six or seven GRPs have been found in the Mediterranean basin. Others have found domestication and pre-domestication, respectively, in the Near-East and in central Spain [10], [15], [24], [25]. There is no data showing olive cultivar domestication locations.
Here the sample includes cultivars from the East and the West and cultivars from the Mediterranean islands of Corsica and Sicily. The objectives are broadly to determine whether the Bayesian method is more informative than the dendrogram. We attempted to answer the following (1) of the seven oleasters Bayesian-based clusters how many may have led to cultivars; (2) whether the method used was appropriate in linking cultivars to Glacial Refugee Populations; (3) whether among the set of cultivars some may have origins in one or several GRPs. Results were discussed within a methodological and domestication context, as this novel diversity could be applied to genetic resource management as well as olive production programs.
Section snippets
Materials and methodology
A total of forty cultivars (Table 1) were chosen which form twenty-three clusters in a dendrogram [3]. They represent the olive RAPD-based diversity. They come from the East and West Mediterranean basin and, according to their names, are from several countries and islands. They also carry different chlorotypes CE1 from the East or COM and CCK from the West of the Mediterranean basin [26]. Consequently the forty-cultivar set covers most of the cultivated olive genetic diversity. A few of these
Cultivar diversity
As a reference, oleasters displayed 171 alleles. On average, each locus of the twelve loci, displayed 14.25 alleles. The twelve SSR loci revealed n = 1.18 as allele content (average number of alleles). In cultivars, the twelve loci displayed 117 alleles representing 68.4% of the oleaster diversity with 11 alleles per locus and 1.02 allele content. In all cases cultivars are highly heterozygous: mean Hobs about 0.75, He 0.72 [23 for details].
Multivariate and similarity analyses
Multivariate analysis (FCA) calculated on cultivar
Discussion
Routine cultivar identification has been set up in different countries [2], [33], [34], [35], [36] and was not addressed in this study. Leaf samples from all trees did not enable to us to characterise each gene pool by differential morphologic traits. Cultivar relationships have been extremely well documented, and several teams have indicated that in the west, cultivars originated (at least, partly) in local oleasters [6] based on chlorotypes or mitotypes, respectively. However, the
Conclusion
The Bayesian method used indubitably brought more information on cultivar origins than classical methods. Although the set of cultivars was limited, we revealed several domestication pools and with the oleaster sampling we can locate, broadly, the domestication site. New samples of oleaster will be needed for a more accurate definition of these sites. The same method may be suitable for grapes or other crops as it enables cultivars to be classified according to their diverse origins. Other
Conflict of interest
We declare no conflict of interest.
Acknowledgements
Dr. L. Baldoni, Dr. I. Trujillo, and Dr. V. Bronzini provided some samples from Italy, Spain, and Corsica France, respectively. The study was partly supported from within the framework of the Trial-Cost, Phylogene-INRA action and DADP2 programme (2001–2005) sponsored by the Conseil Régional du Languedoc Roussillon. Deep thanks are due to Dr. Jenny Guerin, University of Adelaide, for her suggestions and improvements.
References (47)
- et al.
Beginnings of olive cultivation in eastern spain in relation to holocene bioclimatic changes
Quaternary Res.
(1996) - et al.
Multiple origins of the Mediterranean olive (Olea europaea L. ssp. europaea) based upon mitochondrial DNA polymorphisms
CR Acad. Sci. Life Sci.
(2000) A revision of Olea L. (Oleaceae)
Kew Bull.
(2002)- et al.
Using molecular markers for olive germplasm management in one collection
Theor. Appl. Genet.
(2003) - et al.
Cultivar identification in the olive (Olea europaea L.) based on RAPDS
J. Am. Hort. Sci.
(2001) - H. Hannachi, C. Breton, M. Msallem, S. Ben El Hadj, M. El Gazzah, A. Bervillé, Differences between native and...
- et al.
Using multiple types of molecular markers to understand olive phylogeography
- et al.
Olive domestication from structure of oleasters and cultivars using RAPDs and mitochondrial RFLP
Genet., Select., Evolut.
(2001) - et al.
Ancient wild olives in Mediterranean forests
Nature
(2001) - et al.
Allozyme variation of Oleaster populations (wild olive tree) (Olea europaea L.) in the Mediterranean Basin
Heredity
(2004)