Anatomical, morphometric, and stratigraphic analyses of theropod biodiversity in the Upper Cretaceous (Campanian) Dinosaur Park Formation1

Abstract

The Dinosaur Park Formation (DPF) of Alberta, Canada, has produced one of the most diverse dinosaur faunas, with the record favouring large-bodied taxa, in terms of number and completeness of skeletons. Although small theropods are well documented in the assemblage, taxonomic assessments are frequently based on isolated, fragmentary skeletal elements. Here we reassess DPF theropod biodiversity using morphological comparisons, high-resolution biostratigraphy, and morphometric analyses, with a focus on specimens/taxa originally described from isolated material. In addition to clarifying taxic diversity, we test whether DPF theropods preserve faunal zonation/turnover patterns similar to those previously documented for megaherbivores. Frontal bones referred to a therizinosaur (cf. Erlikosaurus), representing among the only skeletal record of the group from the Campanian–Maastrichtian (83–66 Ma) fossil record of North America, plot most closely to troodontids in morphospace, distinct from non-DPF therizinosaurs, a placement supported by a suite of troodontid anatomical frontal characters. Postcranial material referred to cf. Erlikosaurus in North America is also reviewed and found most similar in morphology to caenagnathids, rather than therizinosaurs. Among troodontids, we document considerable morphospace and biostratigraphic overlap between Stenonychosaurus and the recently described Latenivenatrix, as well as a variable distribution of putatively autapomorphic characters, calling the validity of the latter taxon into question. Biostratigraphically, there are no broad-scale patterns of faunal zonation similar to those previously documented in ornithischians from the DPF, with many theropods ranging throughout much of the formation and overlapping extensively, possibly reflecting a lack of sensitivity to environmental changes, or other cryptic ecological or evolutionary factors.