Morphological evidence on blattarian phylogeny: “phylogenetic histories and stories” (Insecta, Dictyoptera)

Abstract

The hypotheses of Grandcolas and Klass on blattarian phylogeny are compared through a scrutiny of all stages of phylogenetic analysis: taxon sampling; morphological studies; establishment of topographic homologies; treatment and presentation of characters, states, and state distribution over taxa; cladistic analysis and support of clades. Grandcolas' hypothesis is difficult to test due to inadequate documentation in several stages. In the testable subsets, many mishomologizations between male genital (phallomere) components and many misscorings for phallomere and non-phallomere characters are demonstrated. Through the use of high-rank taxa as terminal taxa in the matrix, much support for alternative clades is neglected. With the exception of some that are uncontroversial, the clades established by Grandcolas lose their support when the necessary character revisions are performed; of the 52 clade autapomorphies indicated in the cladogram, 7 remain here untested, 7 are accepted, and 38 are refuted. The scenarios of dictyopteran evolution built by Grandcolas on his phylogenetic results cannot he upheld. Klass' hypothesis appears more strongly supported, being based on more detailed morphological observations and homology discussions and considering evidence from all character state distributions. It is thus another demonstration that extensive and detailed morphological investigations and structural comparisons between taxa are essential for morphology-based phylogenetic work, in particular if complex and diverse organs (such as the dictyopteran phallomeres) are used. Some further general issues of phylogeny reconstruction are also referred to, and the importance of the topographic homology level, which precedes the primary (s.s.) and secondary levels, is emphasized.