Abstract
Most authors have identified two rapid increases in relative brain size (encephalization quotient, EQ) in cetacean evolution: first at the origin of the modern suborders (odontocetes and mysticetes) around the Eocene-Oligocene transition, and a second at the origin of the delphinoid odontocetes during the middle Miocene. We explore how methods used to estimate brain and body mass alter this perceived timing and rate of cetacean EQ evolution. We provide new data on modern mammals (mysticetes, odontocetes, and terrestrial artiodactyls) and show that brain mass and endocranial volume scale allometrically, and that endocranial volume is not a direct proxy for brain mass. We demonstrate that inconsistencies in the methods used to estimate body size across the Eocene-Oligocene boundary have caused a spurious pattern in earlier relative brain size studies. Instead, we employ a single method, using occipital condyle width as a skeletal proxy for body mass using a new dataset of extant cetaceans, to clarify this pattern. We suggest that cetacean relative brain size is most accurately portrayed using EQs based on the scaling coefficients as observed in the closely related terrestrial artiodactyls. Finally, we include additional data for an Eocene whale, raising the sample size of Eocene archaeocetes to seven. Our analysis of fossil cetacean EQ is different from previous works which had shown that a sudden increase in EQ coincided with the origin of odontocetes at the Eocene-Oligocene boundary. Instead, our data show that brain size increased at the origin of basilosaurids, 5 million years before the Eocene-Oligocene transition, and we do not observe a significant increase in relative brain size at the origin of odontocetes.