Sampling at appropriate spatial scales in the Southern Ocean is logistically challenging and may influence estimates of diversity by missing intermediate representatives. With the assistance of sampling efforts especially influenced by the International Polar Year 2007-2008, we gathered nearly 1500 specimens of the crinoid species Promachocrinus kerguelensis from around Antarctica. We used phylogeographic and phylogenetic tools to assess its genetic diversity, demographic history and evolutionary relationships. Six phylogroups (A-F) identified in an earlier study are corroborated here, with the addition of one new phylogroup (E2). All phylogroups are circumpolar, sympatric and eurybathic. The phylogeny of Promachocrinus phylogroups reveals two principal clades that may represent two different cryptic species with contrasting demographic histories. Genetic diversity indices vary dramatically within phylogroups, and within populations, suggesting multiple glacial refugia in the Southern Ocean: on the Kerguelen Plateau, in the East Weddell Sea and the South Shetland Islands (Atlantic sector), and on the East Antarctic continental shelf in the Dumont d'Urville Sea and Ross Sea. The inferences of gene flow vary among the phylogroups, showing discordant spatial patterns. Phylogroup A is the only one found in the Sub-Antarctic region, although without evident connectivity between Bouvet and Kerguelen populations. The Scotia Arc region shows high levels of connectivity between populations in most of the phylogroups, and barriers to gene flow are evident in East Antarctica.

There has been a significant body of literature on species flock definition but not so much about practical means to appraise them. We here apply the five criteria of Eastman and McCune for detecting species flocks in four taxonomic components of the benthic fauna of the Antarctic shelf: teleost fishes, crinoids (feather stars), echinoids (sea urchins) and crustacean arthropods. Practical limitations led us to prioritize the three historical criteria (endemicity, monophyly, species richness) over the two ecological ones (ecological diversity and habitat dominance). We propose a new protocol which includes an iterative fine-tuning of the monophyly and endemicity criteria in order to discover unsuspected flocks. As a result nine « full » species flocks (fulfilling the five criteria) are briefly described. Eight other flocks fit the three historical criteria but need to be further investigated from the ecological point of view (here called « core flocks »). The approach also shows that some candidate taxonomic components are no species flocks at all. The present study contradicts the paradigm that marine species flocks are rare. The hypothesis according to which the Antarctic shelf acts as a species flocks generator is supported, and the approach indicates paths for further ecological studies and may serve as a starting point to investigate the processes leading to flock-like patterning of biodiversity.