Bayesian Inference of the Metazoan PhylogenyA Combined Molecular and Morphological Approach

Abstract

The molecular data matrix used in our study, published in the September 21 issue of Current Biology on pages 1644–1649, was obtained from the small subunit (SSU) rRNA database and not directly from GenBank. The sequences in the (SSU) rRNA database are provided in aligned format in accordance to the secondary-structure information derived by comparative sequence analysis of numerous sequences. We used sequences from this database in order to ensure that the original model-based approach of the paper was followed through all steps of data treatment, from alignment to the Bayesian inference of the phylogeny. The database, however, contains only a subset of the sequences available from GenBank and only one ascertained species of nemertodermatids (Nemertinoides elongates). This sequence has been shown to be problematic and has, as Ronald Jenner of the University of California, Davis correctly pointed out, been retracted from GenBank but not from the SSU database. This was a fact we unfortunately were unaware of. To investigate the effects of the use of another, less ambiguous, nemertodermatid sequence on the Bayesian analyses of the data matrix, we downloaded the sequence of Nematoderma westbladi (AF327726) from GenBank, replaced it with the original Nemertinoides elongatus sequence, and performed a new combined analysis. Figure 1 shows the resulting tree. In contrast to the original analysis, the representative of Nemertodermatida came out as a sistergroup to all other bilaterian metazoans except Acoela. This is in agreement with Jondelius et al. [1Jondelius U. Ruiz-Trillo I. Baguna J. Riutor E. The Nemertodermatida are basal bilaterians and not members of the Platyhelminthes.Zool. Scr. 2002; 31: 201-215Crossref Scopus (136) Google Scholar], who were the first to report solid molecular support for a basal bilaterian position of the flatworm taxon: Nematodermatida. However, the phylogenetic tree inferred in the new analysis did not effect a positional change in any of the major taxonomic groups, and Deuterostomia, Protostomia, Ecdysozoa, Lophotrochozoa are all identical to the original analysis. Only groups that were originally positioned close to the replaced nemertodermatid sequence were affected in the new analysis. The remaining flatworms now form a sister group to Syndermata, whereas the two chaetognaths are united with Gnatostomolida to form an unresolved basal protostomian trichotomy with Ecdysozoa and a clade consisting of Lophotrochozoa, Syndermata, Gastotricha, the flatworms, Cycliophora, and the entoprocts. Importantly, the new analysis provides better support for the key taxonomic groupings already discussed in the original paper.