A molecular phylogeny of Hexapoda

We present a supermatrix approach to the phylogeny of Insecta that stemmed from a talk given at the 2nd Dresden Meeting on Insect Phylogeny (2005). The data included a fragment of the 28S (D1–D8) and complete sequences for the 18S, histone (H3), EF-1α, COI, COII, the 12S and 16S plus the intervening tRNA, and 170 morphological characters. Ribosomal RNA sequences were manually aligned to secondary structure. Two separate Bayesian likelihood analyses were performed, as well as a weighted parsimony analysis, on combined data. Partitioned datasets were also explored. Expected clades like Hexapoda, Insecta, Dicondylia, Pterygota, Neoptera, Dictyoptera, Paraneoptera, and Endopterygota were consistently recovered. However, conflicting hypotheses from independent datasets, as well as a lack of quantitative support from the combined supermatrix, suggest that the elucidation of relationships between non-holometabolous neopteran orders is far from resolved. Substitution rate heterogeneity among lineages, missing intermediate taxa, near simultaneous divergences, flawed phylogenetic models and nucleotide compositional bias are discussed as possible causes for unresolved interordinal relationships. The capacity of this dataset to convey information, its inherent limitations, and the role and responsibility of the systematist in interpreting data are explored