BackgroundDomestication generally implies a loss of diversity in crop species relative to their wild ancestors because of genetic drift through bottleneck effects. Compared to native Mediterranean fruit species like olive and grape, the loss of genetic diversity is expected to be more substantial for fruit species introduced into Mediterranean areas such as apricot (Prunus armeniaca L.), which was probably primarily domesticated in China. By comparing genetic diversity among regional apricot gene pools in several Mediterranean areas, we investigated the loss of genetic diversity associated with apricot selection and diffusion into the Mediterranean Basin.ResultsAccording to the geographic origin of apricots and using Bayesian clustering of genotypes, Mediterranean apricot (207 genotypes) was structured into three main gene pools: ‘Irano-Caucasian’, ‘North Mediterranean Basin’ and ‘South Mediterranean Basin’. Among the 25 microsatellite markers used, only one displayed deviations from the frequencies expected under neutrality. Similar genetic diversity parameters were obtained within each of the three main clusters using both all SSR loci and only 24 SSR loci based on the assumption of neutrality. A significant loss of genetic diversity, as assessed by the allelic richness and private allelic richness, was revealed from the ‘Irano-Caucasian’ gene pool, considered as a secondary centre of diversification, to the northern and southwestern Mediterranean Basin. A substantial proportion of shared alleles was specifically detected when comparing gene pools from the ‘North Mediterranean Basin’ and ‘South Mediterranean Basin’ to the secondary centre of diversification.ConclusionsA marked domestication bottleneck was detected with microsatellite markers in the Mediterranean apricot material, depicting a global image of two diffusion routes from the ‘Irano-Caucasian’ gene pool: North Mediterranean and Southwest Mediterranean. This study generated genetic insight that will be useful for management of Mediterranean apricot germplasm as well as genetic selection programs related to adaptive traits.

In order to give insights into the origin and historical selection process of Tunisian apricot propagated by grafting, 31 cultivars from three areas presenting contrasting ecological conditions -Kairouan, Testour and Ras Jbel were compared to cultivars from Europe, North America, North Africa, Turkey, Iran and China, using 234 AFLP markers. The phenetic analysis allowed to distinguish 5 clusters, the four previously defined groups: -'diversification', 'geographically adaptable', 'continental European' and 'Mediterranean' -groups and the Tunisian one. The partitioning of genetic diversity within and between cultivar groups assessed according to the Bayesian approach and assuming Hardy-Weinberg equilibrium, showed a loss of 21.81-38.49% of genetic diversity in Tunisian apricot compared to Mediterranean and diversification groups, respectively. Genetic variation occurred within Tunisian subgroups rather than among (F ST = 0.060) evidencing a narrow genetic pool. Mediterranean and Tunisian groups were the least differentiated. Comparing them, 24 AFLP fragments discriminated the Mediterranean group from the Tunisian group but most of them where also shared by the other groups. Strongly differentiated gene pool and low genetic diversity are probably the result of bottleneck effects linked to the occurrence of propagation by seedlings rather than by grafting during the introduction periods in the North and the Centre of Tunisia. This study points at the propagation by seedlings as an important factor which should be taken into account to understand the evolution of apricot in South Mediterranean areas.

Bourguiba et al. Worldwide Cultivated Apricot Genetic Structure used to define an apricot core collection based on molecular marker diversity which will be useful for further identification of genomic regions associated with commercially important horticultural traits through genome-wide association studies to sustain apricot breeding programs.