Morphological and molecular systematics of scleractinian corals (Cnidaria, Anthozoa), with emphasis on deep-water species

Ubiquitous to all oceans, the phylum Cnidaria is the second most basal metazoan group in the tree of life, with plausible fossil record dating back from the pre-Cambrian. The vast majority of cnidarians are restricted to the marine realm, although few species are known to occur in fresh water systems. The approximately 9000 extant representatives of this phylum are divided into 2 sub-phyla, of which the Anthozoa embraces nearly 7000 species including approximately 1400 extant scleractinians, also commonly known as corals. Corals are the subject of intense scientific, public and media interest because of their ecological and economic importance, and because of the uncertain fate of coral reefs in the face of ever increasing anthropogenic challenges. Despite this, the scleractinian origins are not well understood. They suddenly appeared in the Middle Triassic already represented by a wide variety of solitary and colonial forms. From the level of colony integration to the microstructural organization within individual corallites, the range of morphological variation in the Triassic fossils is comparable to that observed in modern scleractinians. In fossil specimens in which aragonite is preserved, coralla show at least four basic types of microstructural organization suggesting an extensive Palaeozoic evolutionary history for the Order. Two long-lived hypotheses for scleractinian coral evolution are that they are either descendants of late Paleozoic rugose corals that survived the mass extinction at the Permian/Triassic boundary, or that they evolved from soft-bodied corallimorpharian-like ancestors by independently gaining the ability to deposit a calcified skeleton. In the last few years, a third hypothesis known as "naked corals" has received extensive media attention, advocating that some scleractinian lineages, as result of ocean acidification, have the ability to undergo skeleton loss. These contradicting hypotheses all suffer from the absence of data from deep-water corals. Indeed, prior to this thesis, the vast majority of molecular data available was retrieved from shallow-water zooxanthellate species, and despite of accounting for approximately half of the extant species of the order, azooxanthellate corals had rarely been included in molecular phylogenetic reconstructions. Nonetheless, starting from the premise that taxonomy remains essential to credible biological research, and acknowledging that reliable and accurate molecular based phylogenies, and its interpretation relies on taxonomic knowledge, a large segment of the present thesis is dedicated to classical taxonomy of extant azooxanthellate scleractinians from New Caledonia. The examination of more than 3000 specimens, collected between 80 and 1200 m deep, revealed the occurrence of 170 species (including 3 new families, 1 new genus, and at least 12 undescribed species) to this small southwestern Pacific Archipelago. Species rarefaction analysis suggests that the species diversity for this region still underestimated, implying a much more diverse azooxanthellate scleractinian coral fauna than all previously examined regions in the world. Additionally, to better understand broad patterns of coral evolution, I generated molecular data for a broad and representative range of deep-water scleractinians, and conducted the most comprehensive molecular phylogenetic analyses of the order to date. Primarily, I show that there is a striking discrepancy between the taxonomic validity of coral families consisting predominantly of deep- or shallow-water representatives. Most families composed predominantly of deep-water azooxanthellate species are monophyletic but, by contrast, most families composed predominantly of shallow-water zooxanthellate taxa are polyphyletic, although Acroporidae, Caryophylliidae, Poritidae, Pocilloporidae, and Fungiidae were exceptions to this general pattern. Furthermore, using a relaxed molecular clock calibrated against the oldest Mesozoic fossil records that can be indisputably assigned to extant genera/families, I show not only that the divergence of the two major coral clades, the "Robust" and "Complex" corals, took place more than 110 My earlier than previously thought, but also that two families composed exclusively of deep-water corals, the Gardineriidae and Micrabaciidae, diverged even earlier, pushing coral origins to the Middle Ordovician; ca. 450 Ma. These results genetically and morphologically connect the elusive "scleractiniamorph" fossils with extant scleractinians, confirm that the order Scleractinia is monophyletic, suggest that corals evolved from Paleozoic soft-bodied ancestors, and show that modern colonial reef-building corals, which are dependent on symbionts, had solitary, non-symbiotic precursors. In conclusion, I hypothesise that global environmental changes over the past 460 My may have been a major driving force behind the colonization of deep environments by scleractinian corals. Taken together, these results significantly improve our understanding of coral evolution and significantly enrich the debate about how corals are capable to withstand global climate changes