Abstract

Coleopterida (Coleoptera + Strepsiptera) has been established as the sister group of Neuropterida (Megaloptera + Neuroptera + Raphidioptera) based on recent phylogenetic analyses of DNA sequence data obtained from genomes and transcriptomes. However, within the resulting clade (Neuropteroidea) the proposed sister-group relationship between the highly specialized endoparasitic Strepsiptera and the megadiverse Coleoptera still lacks convincing morphological support. Furthermore, relationships among the four suborders of Coleoptera remain controversial, with morphological characters strongly conflicting with results suggested by molecular evidence. A large morphological dataset comprising external and internal features of adults and immature stages is presented here and analysed phylogenetically. Our study is focused on deep splits in Coleopterida and on reconstructing character evolution on the phenotypic level. Parsimony analyses clearly support a sister-group relationship between Strepsiptera and monophyletic Coleoptera. Presumptive synapomorphies are characters linked with posteromotorism, but also features of the head and prothorax. We recover Archostemata as sister group of the remaining extant Coleoptera, and Polyphaga as sister group of the species-poor suborder Myxophaga. The most important character complex of Coleoptera is heavy sclerotization without exposed membranes and a simplification of the thoracic muscle apparatus. Non-archostematan beetles are characterized by further simplifications of the thoracic locomotor apparatus. This trend reaches its peak in Myxophaga and Polyphaga, and these suborders also share apomorphies of the larval legs. A pattern with Polyphaga as sister to all other suborders and a clade Myxophaga + Archostemata (as in recent molecular phylogenetic studies) requires ten additional steps with our dataset. This scenario implies that various simplifications of the thoracic exoskeleton and musculature have taken place several times independently, and also that a complex feeding apparatus suitable for saprophagy and sporophagy was ancestral in Coleoptera, with secondary reduction (or modification) in Archostemata and Adephaga. The coleopteran subordinal relationships remain a challenge, with morphological and molecular data suggesting distinctly different patterns. The earliest evolution of Coleopterida is not documented in the fossil record. The exploration of potential stem-group fossils is a high priority, as is the study of species from the Permian–Triassic transition zone, which are apparently important in the context of evaluating the relationships among beetle suborders.