**7. Relationships between the oleaster and the olive**

[70] Baldoni et al. concluded that most cultivars have been introduced into Central Italy regions from the outside and that Umbrian cultivars have originated by selection from local oleaster trees.[71] Belaj et al. concluded that the genetic structure (=density of alleles across the geographic distribution of individuals) is not strong enough to positively establish relationships between true oleaster trees and cultivated varieties. The impact of these studies has probably been limited due to the limited sampling of the wild forms. [30] Hannachi et al. (2009) has revealed that the cultivar sets can be split into those of local

origins and those introduced from the Near-East and western regions, making Tunisia in central Mediterranean a key-place for olive and oleaster diversity. [57] Breton concluded that the oleaster populations were structured in at least eleven ancestral populations, which colonized the Mediterranean basin after the last ice age, following mostly the sea-coasts. Based on coincidence of the sampling area and the clusters, some geographic zones for the refugee populations have been suggested: 1) Four in the East (Turkey, Cyprus (2) and Israel+ Lebanon), 2) four in the central Mediterranean (including North of Africa, and main islands Sicily, Sardinia, and Corsica) and 3) three in the western Mediterranean (Continental Spain (2) and South France). The zones were well defined in the East and wide in the West, probably due to limited sampling in Spain and continental Italy.

Origin and History of the Olive 15

All these data converge to sustain that in each region the present olive cultivar diversity is either or both the result of ancient introductions from the Near-East and/or from other area (North of Africa, Cyprus, Turkey), local selection from oleaster trees and from crosses between oleaster and ancient cultivar trees. However, the self-incompatibility system in the wild olive and the olive is still not yet known, leaving the selection pressures that occurred along the domestication processes unknown, which are required to gather enough S-alleles in a region to enable fruit set. [74] Breton & Bervillé have recently deciphered S-allele pair wise combinations for a few varieties, and it appeared which varieties may combine efficiently, at least *in silico*, but it remains to experimentally check coincidence in blossoming and other compatibility levels, which may affect development of pulp and embryo. The model developed infers which genotypes may coexist to ensure correct fruit set, even

though self-compatibility appears inherent to most varieties.

**Figure 4.** Abandonned olive trees along the Mediterraean coast (North of Catalogna, Spain)

© Catherine Breton
