Vicariance patterns in the Mediterranean Sea: east–west cleavage and low dispersal in the endemic seagrass <i>Posidonia oceanica</i>

Abstract Aim The seagrass, Posidonia oceanica is a clonal angiosperm endemic to the Mediterranean Sea. Previous studies have suggested that clonal growth is far greater than sexual recruitment and thus leads to low clonal diversity within meadows. However, recently developed microsatellite markers indicate that there are many different genotypes, and therefore many distinct clones present. The low resolution of markers used in the past limited our ability to estimate clonality and assess the individual level. New high‐resolution dinucleotide microsatellites now allow genetically distinct individuals to be identified, enabling more reliable estimation of population genetic parameters across the Mediterranean Basin. We investigated the biogeography and dispersal of P. oceanica at various spatial scales in order to assess the influence of different evolutionary factors shaping the distribution of genetic diversity in this species. Location The Mediterranean. Methods We used seven hypervariable microsatellite markers, in addition to the five previously existing markers, to describe the spatial distribution of genetic variability in 34 meadows spread throughout the Mediterranean, on the basis of an average of 35.6 (± 6.3) ramets sampled. Results At the scale of the Mediterranean Sea as a whole, a strong east–west cleavage was detected ( amova) . These results are in line with those obtained using previous markers. The new results showed the presence of a putative secondary contact zone at the Siculo‐Tunisian Strait, which exhibited high allelic richness and shared alleles absent from the eastern and western basins. F statistics (pairwise θ ranges between 0.09 and 0.71) revealed high genetic structure between meadows, both at a small scale (about 2 to 200 km) and at a medium scale within the eastern and western basins, independent of geographical distance. At the intrameadow scale, significant spatial autocorrelation in six out of 15 locations revealed that dispersal can be restricted to the scale of a few metres. Main conclusions A stochastic pattern of effective migration due to low population size, turnover and seed survival is the most likely explanation for this pattern of highly restricted gene flow, despite the importance of an a priori seed dispersal potential. The east–west cleavage probably represents the outline of vicariance caused by the last Pleistocene ice age and maintained to this day by low gene flow. These results emphasize the diversity of evolutionary processes shaping the genetic structure at different spatial scales.