Endothermy, neuron counts, and other issues: Further remarks on neurocognitive evolution in fossil vertebrates

Last year, we challenged the view that large-bodied theropod dinosaurs such as Tyrannosaurus rex resembled primates in cognition and behavior, a proposition made by Herculano-Houzel in 2023. More recently, Jensen et al. have criticized our work on this topic, raising methodological and conceptual issues. Central to their argument is the assumption that tachymetabolic endotherms should be expected to converge in neurocognitive traits, which follows the recently proposed endothermic brain hypothesis. We here respond to their critique, address critical misconceptions, and argue that none of the points raised by Jensen et al. challenge the conclusions we have drawn. We show that the endothermic brain hypothesis lacks robust support from the fossil record. As of now, no compelling evidence suggests that endothermy coevolved with enlarged brains or elevated neuron densities in either the avian or mammalian lineage. Various fossil groups containing endothermic taxa retain plesiomorphic endocast traits and do not converge with birds and mammals in the relative size and proportions of their brains. Furthermore, we elaborate on our discussion on (forebrain) neuron counts as correlates of cognitive performance and highlight that neuron numbers evolve in tandem with body mass in birds and mammals, suggesting that comparatively high neuron number estimates for some Mesozoic dinosaurs do not require explanations that orbit around exceptional cognitive abilities. Despite these disagreements, we identify significant overlap in opinion between Jensen et al. and ourselves, including in the position that neuron count estimates for Mesozoic dinosaurs will remain unreliable and are unsuitable for inferring cognitive complexity.