Phylogeny of the Neuropterida:: a first molecular approach

In a first molecular approach specially dedicated to examining the phylogeny of the Neuropterida, two nuclear and two mitochondrial genes were tested: 18S rRNA, translation elongation factor-1alpha, cytochrome c oxidase subunit 3 and 16S rRNA. Molecular results are discussed in the light of a previous holomorphological cladistic analysis. The hypothesis of a sister-group relationship Raphidioptera + (Neuroptera + Megaloptera) put forward in recent morphological analyses is supported by our data, which is in contrast to the traditional view (Raphidioptera + Megaloptera) + Neuroptera. Furthermore, the Nevrorthidae (constituting the suborder Nevrorthiformia) as a sister group of all other Neuroptera is confirmed. The disruption of the suborder Hemerobiiformia is the most conflicting result of the molecular analysis. Sisyridae and Osmylidae do not cluster within Hemerobiiformia, but represent two distinct and widely separated branches. The remaining Hemerobiiformia emerge as the sister group of the suborder Myrmeleontiformia, which is once more confirmed as monophyletic. Among the genes tested, cytochrome c oxidase subunit 3 proved to be most potent for resolving the phylogenetic relationships among Neuropterida. The nuclear gene for the ribosomal 18S rRNA is too conserved within the alignable regions, whereas the variable sections are too divergent to be applicable within this evolutionary time frame. The elongation factor-1alpha gene proved to exist in more than one copy in Neuropterida, and thus is not applicable in the present state of knowledge. With respect to the mitochondrial sequences (cytochrome c oxidase subunit 3, 16S rRNA), saturation impedes the unambiguous resolution of deeper nodes. Apparently, due to early diversification of the heterogeneous Neuroptera, phylogenetic analysis of this group remains a challenge with respect to selection of the proper genes and mutatis mutandis the morphological approach.