UC San Diego

Coral systematics have been plagued by a host of problems. Traditional identification methods based on gross anatomy often fail to recognise natural taxa because of morphological convergence, environment-induced phenotypic variability, and recent speciation. I first summarised levels of paraphyly for taxa within Scleractinia and showed that paraphyly was also a significant problem in Faviidae, the second-most speciose reef-building family. Phylogenetic reconstruction based on two mitochondrial markers amplified for 41 species representing 13 genera illustrated that at least five genera and Faviidae were paraphyletic. Morphological characters currently used to identify these corals similarly failed to recover many genera. The analysis was then expanded to the clade 'Bigmessidae' that comprised Faviidae, Merulinidae, Pectiniidae and Trachyphylliidae. To reconstruct a robust and resolved molecular phylogeny, three nuclear and two mitochondrial loci were sequenced from 76 of the 132 'Bigmessidae' species collected from five reef regions in the central Indo-Pacific and Atlantic. Results indicated numerous examples of cryptic taxa due to unexpected phylogenetic placements of several species, but the recovery of most 'Bigmessidae' genera with only minor degrees of paraphyly offered hope for impending taxonomic amendments. Congruence between molecular data and morphology was then determined by mapping 47 corallite and subcorallite characters onto the 'Bigmessidae' (= Merulinidae) molecular phylogeny. Subcorallite traits diagnostic of natural groups included characters associated with wall microstructure and septal tooth micromorphology. Based on these results and a larger collection of corals, I conducted a taxonomic revision of Merulina, type genus of Merulinidae, and its sister group, Goniastrea. Finally, I used the supertree approach to reconstruct a comprehensive tree of life for Scleractinia that included all 837 living reef species, one-third of which are threatened with extinction. A composite measure of phylogenetic distinctiveness and extinction risk was used to identify the most endangered lineages, some of which would not be given top conservation priority on the basis of risk alone. Tests for phylogeny-associated patterns showed that corals susceptible, resistant or resilient to impacts such as bleaching and disease tended to be close relatives. Intensification of these threats or extirpation of the endangered lineages could therefore result in disproportionate pruning of the coral tree of life