ANDERSON, F.E.*; FOSTER, P.G.: Tests of ecdysozoan monophyly using molecular sequence data

Some analyses of 18S ribosomal RNA sequence data have supported monophyly of Ecdysozoa, a superphylum composed of taxa that possess periodically molted chitinous cuticles (e.g., Arthropoda, Onychophora, Priapula and Nematoda). The Ecdysozoa hypothesis already has been accepted by many researchers, and this hypothesis (particularly the proposed close relationship between nematodes and arthropods) has been used to understand enigmatic developmental and genetic patterns within Nematoda. Recently, however, evidence from multiple nuclear protein-coding genes has been presented suggesting that Ecdysozoa (or at least a Nematoda + Arthropoda clade) is not monophyletic. To evaluate support for ecdysozoan monophyly, we added nematode sequences to previously published alignments for the 18S gene and several nuclear protein-coding genes and reanalyzed the data using maximum parsimony and maximum likelihood (including heterogeneous ML models, in which model parameters such as base composition can be estimated separately for particular branches). We also analyzed alignments of two other genes (the nuclear elongation factor-1 alpha gene and the mitochondrial cytochrome c oxidase subunit I gene). Finally, we used a taxon jackknifing approach with the 18S, EF-1 alpha and COI alignments to assess the role of taxon sampling in estimating ecdysozoan relationships. Our findings suggest that support for Ecdysozoa across multiple data sets and taxon samples is not unambiguous, and we argue that the use of the Ecdysozoa hypothesis for inferring character evolution within Nematoda may be premature.