© 1998 by The Society for Integrative and Comparative Biology
Considerations for Reconstructing Metazoan History: Signal, Resolution, and Hypothesis Testing1
Center for Theoretical and Applied Genetics, Rutgers—The State University of New Jersey New Brunswick, New Jersey 08901-8521
SYNOPSIS. Certain issues of metazoan phytogeny remain difficult to resolve even with advances in phylogenetic theory and molecular tools. This lack of resolution might be due to limited phylogenetic signal in regions of the metazoan tree. For example, 18S rDNA data are consistent with a rapid radiation early in the history of bilaterian animals. Despite this lack of resolution, our understanding of metazoan evolutionary history can be furthered by extracting phylogenetic signal from data with more appropriate evolutionary models and by explicitly testing alternative hypotheses. I examined 18S rDNA data across a diverse number of phyla to identify potential problems with using genetic data to reconstruct metazoan history. Although the 18S performs well at the most general level of metazoan phytogeny, it has shortcomings particularly among the non-arthropod protostomes. Examples of parametric bootstrap simulation are given to illustrate how we may begin to address methodological issues when there is limited phylogenetic signal. These simulations suggest that we are more likely to understand the bilaterian radiation event by increasing the length of nucleotide sequences collected than by employing faster genes.