Morphology, fossils and molecules elucidate the mystery of dung beetle (Coleoptera: Scarabaeidae: Scarabaeinae) evolution using novel Bayesian and parsimony approaches

A plethora of studies published over the last two decades produced numerous contradictory phylogenies leading to the lack of consensus on phylogenetic history in dung beetles. Thus, due to those contradictions, the community of dung beetle systematists could not use advantages of current phylogenetic techniques to improve the largely artificial classification established a half-century ago. Moreover, those conflicting phylogenies, along with unstudied fossils, have significantly hampered our understanding of dung beetle evolutionary history. In this thesis, I revisit the phylogenetics, attempt to improve the classification and explore fossil evidence of dung beetles through the acquisition of global morphological and molecular data in the framework of traditional and novel methods for their phylogenetic analysis. The global morphological (134 species, 232 characters) and molecular (8 genes 547 terminals) datasets were analyzed using parsimony, maximum-likelihood and Bayesian inference. To improve phylogenetic inference, I employed three innovative approaches: (1) use of the positional congruence index to eliminate characters supporting taxon instability in parsimony-based morphological phylogenies, (2) partitioning of morphological matrices using anatomy ontologies and (3) use of Bayesian posterior prediction for selecting data partitions in molecular analysis. The second is the Bayesian approach that assigns characters from different anatomical regions to different partitions, thus allowing to model heterogeneous evolutionary rates, improving model fit and reducing biases in phylogenetic reconstruction. The third approach fits model to data, then uses the estimated parameters to simulate replicated data, which are then compared to the observed data, thus allowing adequacy assessment of a model. The partitions in molecular analyses were selected based on the adequacy of the fitted model, which in turn resulted in improved phylogenetic interference. Both the molecular and morphological analyses yielded consistent results concerning the shallow and deeper nodes in the resulting phylogenetic trees, which supported the definition of new taxonomic concepts for three highly polyphyletic dung beetle tribes and stabilized dung beetle classification. Additionally, a new tribe, Parachoriini trib.n., was recognized to accommodate the new concept of the Oriental genus Parachorius sensu novo (= Cassolus syn.n.), which was revised using the cybertaxonomic tool 3i. The examination of 33 described dung beetle fossils revealed that only 21 of these (two herein described as new) can be reliably referred to dung beetles. Three best-preserved fossils were integrated into the morphological matrix to elucidate their phylogenetic placement. Based on reliable fossils, the dung beetle minimal age shifts to the Eocene (53 Ma), while present-day dung beetle genera had already evolved by the Oligocene–mid Miocene. However, the biogeographic pattern of the molecular and morphological phylogenies suggests a Late Gondwanian origin of dung beetles, which corroborates the Upper Cretaceous age of origin inferred by a recent global phylogeny of Coleoptera.