The anthozoan sub-class Octocorallia includes over 3500 nominal species of soft corals and gorgonian sea fans, many of which serve as critical foundation species in benthic marine ecosystems in shallow waters to the deep sea. Despite their familiarity and ecological impor-tance, the diversity and taxonomy of octocorals remain poorly known. All of the orders, subordinal groups, and a majority of families have been recognized to be poly- or paraphyletic, but poor resolution of the deeper nodes in mitochondrial or single-locus nuclear gene trees have hindered formal revision of the higher-level taxonomy of the group. We used sequence data from target-capture of 739 ultraconserved and exon loci to reconstruct a fully resolved phylogeny for 185 octocoral taxa represent-ing 55 of 63 currently recognized families. We use this phylogeny, sup-plemented with a gene tree for mitochondrial mtMutS for an additional 107 taxa, to guide a revision of the families and orders of Octocorallia. We (1) elevate the anthozoan sub-classes Octocorallia and Hexacorallia to the rank of Class; (2) replace the three currently recognized orders of Octocorallia (Alcyonacea, Pennatulacea, Helioporacea) with two new orders reflecting reciprocally monophyletic major clades; and (3) revise all families with the exception of the 15 recognized families of sea pens, which we accommodate within a new superfamily. The revised classifi-cation of Octocorallia thus comprises 79 families, including 18 that are newly described and three that have been reinstated or elevated in rank. In addition, two new genera are described and another three reinstated. We leave the family assignment of 46 of 413 genera as incertae sedis until further molecular or morphological data can be obtained to confirm their phylogenetic affinities.

The anthozoan sub-class Octocorallia, comprising approximately 3000 species of soft corals, gorgonians, and sea pens, remains one of the most poorly understood groups of the phylum Cnidaria. Efforts to classify the soft corals and gorgonians at the suprafamilial level have long thwarted taxonomists, and the subordinal groups in current use are widely recognized to represent grades of colony forms rather than clades. Molecular phylogenetic analyses of the sub-class do not support either the current morphologically based subordinal or familial-level taxonomy. To date, however, the resolution necessary to propose an alternative, phylogenetic classification of Octocorallia or to elucidate patterns of morphological evolution within the group is lacking. Attempts to understand boundaries between species and interspecific or intraspecific phylogenetic relationships have been hampered by the very slow rate of mitochondrial gene evolution in Octocorallia, and a consequent dearth of molecular markers with variation sufficient to distinguish species (or sometimes genera). A review of the available ITS2 sequence data for octocorals, however, reveals a yet-unexplored phylogenetic signal both at sequence and secondary-structure levels. In addition, incongruence between mitochondrial and nuclear gene trees suggests that hybrid speciation and reticulate evolution may be an important mechanism of diversification in some genera. Emerging next-generation genomic-sequencing technologies offer the best hope for a breakthrough in our understanding of phylogenetic relationships and of evolution of morphological traits in Octocorallia. Genome and transcriptome sequencing may provide enough characters to resolve relationships at the deepest levels of the octocoral tree, while simultaneously offering an efficient means to screen for new genetic markers variable enough to distinguish species and populations.