Phylogenetic relationships of family groups in Pentatomoidea based on morphology and DNA sequences (Insecta: Heteroptera)

Phylogenetic relationships within the Pentatomoidea are investigated through the coding and analysis of character data derived from morphology and DNA sequences. In total, 135 terminal taxa were investigated, representing most of the major family groups; 84 ingroup taxa are coded for 57 characters in a morphological matrix. As many as 3500 bp of DNA data are adduced for each of 52 terminal taxa, including 44 ingroup taxa, comprising the 18S rRNA, 16S rRNA, 28S rRNA, and COI gene regions. Character data are analysed separately and in the form of a total evidence analysis. Major conclusions of the phylogenetic analysis include: the concept of Urostylididae is restricted to that of earlier authors; the Saileriolinae is raised to family rank and treated as the sister group of all Pentatomoidea exclusive of Urostylididae sensu stricto; a broadly conceived Cydnidae, as recognized by Dolling, 1981, is not supported; the placement of Thaumastellidae within the Pentatomoidea is affirmed and the taxon is recognized at family rank rather than as a subfamily of Cydnidae, although its exact phylogenetic position within the Pentatomoidea remains equivocal; the Parastrachiinae is treated as also including Dismegistus Amyot & Serville and placed within a broadly conceived Corimelaenidae, the latter group being treated at family rank; the family-group taxa Dinidoridae and Tessaratomidae probably represent a monophyletic group, but the recognition of monophyletic subgroups will benefit from additional representation in the sequence data set; and the Lestoniidae is treated as the sister group of the Acanthosomatidae. The Acanthosomatidae and Scutelleridae are consistently recovered as monophyletic. The monophyly of the Pentatomidae appears unequivocal, inclusive of the Aphylinae and Cyrtocorinae, on the basis of morphology, the latter two taxa not being represented in the molecular data set. The Willi Hennig Society 2008. The recognition of a taxon with a composition similar to the present superfamily Pentatomoidea goes back at least to Leach, 1815 (Leston, 1953a). Although several authors have considered Pentatomoidea to be a natural group, there has been substantial disagreement on the relationships among family-level and lower categories (Kirkaldy, 1909; Leston, 1953a; Pendergrast, 1957; Scudder, 1959; Štys, 1961; McDonald, 1966; Gross, 1975, 1976; Štys and Kerzhner, 1975; Cobben, 1978; Schuh, 1986; Gapud, 1991; Henry, 1997; Cassis and Gross, 2002). The majority of previous studies of pentatomoid relationships used one—or at most two—suites of characters, and in many of these studies taxa were grouped on the basis of symplesiomorphic resemblance. Hypotheses of phylogenetic relationships for the Pentatomoidea are presented in explicit diagrammatic form in Fig. 1. These schemes, taken from Bonatto (1988), represent, respectively, the theories of: Fig. 1(a)—Singh-Pruthi, 1925 (diagram and discussion); Fig. 1(b)—Leston, 1958 (fig. 5); Fig. 1(c)—China and Miller, 1959 (fig. 1); Fig. 1(d)—Cobben, 1968 (figs 269– 270); Fig. 1(e)—Cobben, 1978 (several figures and text); and Fig. 1(f)—Gapud, 1991 (fig. 28). As can be seen, these classifications contain many conflicts concerning relationships among the pentatomoid families. The first contribution on pentatomoid relationships using cladisticmethodologywas that ofGapud (1991).He analysed 41 characters in 13 terminal taxa, following the prior general schemes that had divided the superfamily into 11 families. Thaumastella Horváth was included within the Cydnidae; Aphylidae and Megarididae were *Corresponding author: E-mail address: [email protected] The Willi Hennig Society 2008 Cladistics 10.1111/j.1096-0031.2008.00224.x Cladistics 24 (2008) 1–45