Abstract

Phylogenetic analysis of the extinct anthozoan clade Tabulata yields new hypotheses concerning their pattern of diversification in the Ordovician. Two separate phylogenetic analyses, one based on primitive rugose corals as the outgroup (RUGSGRPS), and the other based on Lichenaria as the ancestral tabulate coral (LICHGRPS) yielded different phylogenies. The phylogenies generated are broadly different from previously proposed phylogenies based on possibly subjective morphological interpretations, and on biostratigraphical and/or biogeographical hypotheses alone. Character analysis based on consistency index (a measure of homoplasy of characters) yielded four suites of morphological characters: (1) suites with a high consistency index (CI) that differentiate major groups; (2) suites with a high CI that differentiate subgroups; (3) suites with a low CI that differentiate major groups and (4) suites with a low CI that differentiate subgroups. Therefore, CI does not necessarily correspond with the potential for differentiating major groups. The most useful characters in differentiating major groups of Ordovician tabulates are colony architecture, wall thickness, mural pores, microstructure, corallite shape, and coenenchyme, whereas those not particularly useful in differentiating major groups are tabulae, septa, rows of septal spines, columella, and stereozone. The phylogenetic analyses corroborate the taxonomic integrity of the presently defined Auloporida, Favositida, Halysitida, Heliolitida and most Sarcinulida and falsify the taxonomic integrity of the Chaetetida and the Lichenariida. As presently defined the Halysitida should be separated from the Heliolitida.