Smooth-shelled mussels, Mytilus spp., have an antitropical distribution. In the Northern Hemisphere, the M. edulis complex of species is composed of three genetically well delineated taxa: M. edulis, M. galloprovincialis and M. trossulus. In the Southern Hemisphere, morphological characters, allozymes and intron length polymorphisms suggest that Mytilus spp. populations from South America and Kerguelen Islands are related to M. edulis and those from Australasia to M. galloprovincialis. On the other hand, a phylogeny of the 16S rDNA mitochondrial locus demonstrates a clear distinctiveness of southern mussels and suggests that they are related to Mediterranean M. galloprovincialis. Here, we analysed the faster-evolving cytochrome oxidase subunit I locus. The divergence between haplotypes of populations from the two hemispheres was confirmed and was found to predate the divergence between haplotypes of northern M. edulis and M. galloprovincialis. In addition, strong genetic structure was detected among the southern samples, revealing three genetic entities that correspond to (1) South America and Kerguelen Island, (2) Tasmania, (3) New Zealand. Using the trans-Arctic interchange as a molecular clock calibration, we estimated the time since divergence of populations from the two hemispheres to be between 0.5 million years (MY) and 1.3 MY (average 0.84 MY). The contrasting patterns observed for the nuclear and the organelle genomes suggested two alternative, complex scenarios: two trans-equatorial migrations and the existence of differential barriers to mitochondrial and nuclear gene flow, or a single trans-equatorial migration and a view of the composition of the nuclear genome biased by taxonomic preconception.

Defining the scale of connectivity among marine populations and identifying the barriers to gene flow are tasks of fundamental importance for understanding the genetic structure of populations and for the design of marine reserves. Here, we investigated the population genetic structure at three spatial scales of the red gorgonian Paramuricea clavata (Cnidaria, Octocorallia), a key species dwelling in the coralligenous assemblages of the Mediterranean Sea. Colonies of P. clavata were collected from 39 locations across the Mediterranean Sea from Morocco to Turkey and analysed using microsatellite loci. Within three regions (Medes, Marseille and North Corsica), sampling was obtained from multiple locations and at different depths. Three different approaches (measures of genetic differentiation, Bayesian clustering and spatially explicit maximum-difference algorithm) were used to determine the pattern of genetic structure. We identified genetic breaks in the spatial distribution of genetic diversity, which were concordant with oceanographic conditions in the Mediterranean Sea. We revealed a high level of genetic differentiation among populations and a pattern of isolation by distance across the studied area and within the three regions, underlining short effective larval dispersal in this species. We observed genetic differentiation among populations in the same locality dwelling at different depths, which may be explained by local oceanographic conditions and which may allow a process of local adaptation of the populations to their environment. We discuss the implications of our results for the conservation of the species, which is exposed to various threats.

Combined action from over-harvesting and recent mass mortality events potentially linked to ongoing climate changes has led to new concerns for the conservation of shallow populations (5-60 m) of Corallium rubrum, an octocorallian that is mainly found in the Mediterranean Sea. The present study was designed to analyse population structure and relationships at different spatial scales (from 10s of meters to 100s of kilometres) with a focus on dispersal pattern. We also performed the first analysis of the distribution of genetic diversity using a comparative approach between regional-clusters and samples. Forty populations dwelling in four distinct regions between 14 and 60 m in depth were genotyped using 10 microsatellites. Our main results indicate (i) a generalized pair-sample differentiation combined with a weak structure between regional-clusters; (ii) the occurrence of isolation by distance at the global scale, but also within two of the three analysed regional-clusters; (iii) a high level of genetic diversity over the surveyed area with a heterogeneous distribution from regional-cluster to sample levels. The evolutionary consequences of these results are discussed and their management implications are provided.