Endocranial Casts of Pre-Mammalian Therapsids Reveal an Unexpected Neurological Diversity at the Deep Evolutionary Root of Mammals

Abstract

The origin and evolution of the mammalian brain has long been the focus of scientific enquiry. Conversely, little research has focused on the palaeoneurology of the stem group of Mammaliaformes, the Permian and Triassic non-mammaliaform Therapsida (NMT). This is because the majority of the NMT have a non-ossified braincase, making the study of their endocranial cast (sometimes called the “fossil brain”) problematic. Thus, descriptions of the morphology and size of NMT endocranial casts have been based largely on approximations rather than reliable determination. Accordingly, here we use micro-CT scans of the skulls of 1 Dinocephalia and 3 Biarmosuchia, which are NMT with a fully ossified braincase and thus a complete endocast. For the first time, our work enables the accurate determination of endocranial shape and size in NMT. This study suggests that NMT brain size falls in the upper range of the reptilian and amphibian variation. Brain size in the dicynodont Kawingasaurus is equivalent to that of early Mammaliaformes, whereas the Dinocephalia show evidence of a secondary reduction of brain size. In addition, unlike other NMT in which the endocast has a tubular shape and its parts are arranged in a linear manner, the biarmosuchian endocast is strongly flexed at the level of the midbrain, creating a near right angle between the fore- and hindbrain. These data highlight an unexpected diversity of endocranial size and morphology in NMT, features that are usually considered conservative in this group.

Virtual reconstruction of cranial endocasts of traversodontid cynodonts (Eucynodontia: Gomphodontia) from the upper Triassic of Southern Brazil

“…We estimated the body mass and calculated the EQs following previous protocols (Benoit, Fernandez, et al, ). The highest values of body mass estimation were provided by BMb (Hu et al, ), and the lowest by BMc (Castanhinha et al, ), as previously observed by Benoit, Fernandez, et al ().…”

“…In the last few years, several studies have been carried out, improving the knowledge of the endocranial morphology of several non‐mammaliaform therapsids through digital reconstruction of the endocast as follows: the biarmosuchian Lemurosaurus (Benoit, Fernandez, Manger, & Rubidge, ); the dinocephalian Moschops (Benoit, Manger, Norton, Fernandez, & Rubidge, ); the dicynodonts Pristerodon (Laaß, ), Diictodon (Laaß, Schillinger, & Kaestner, ), Kawingasaurus (Laaß & Kaestner, ), and Niassodon (Castanhinha et al, ); the gorgonopsian Cynariops (Bendel, Kammerer, Kardjilov, Fernandez, & Fröbisch, ); the early non‐mammaliaform cynodont Galesaurus (Pusch, Kammerer, & Fröbisch, ), and the prozostrodontid non‐mammaliaform cynodonts Riograndia (Rodrigues et al, ) and Brasilodon (Rodrigues, Ruf, & Schultz, ).…”

Endocranial morphology of the Brazilian Permian dicynodont Rastodon procurvidens (Therapsida: Anomodontia)

“…However, the EQs of R. procurvidens are still lower than those of the cynodont Brasilitherium (0.14 AE 0.05 and 0.24 AE 0.09; Rodrigues et al 2013), and shows markedly less encephalization than Mammaliaformes, as seen in Hadrocodium (0.24 AE 0.08, 0.51 AE 0.19) and Morganucodon (0.20 AE 0.07, 0.35 AE 0.13) (Rowe et al 2011) (Table S1). This demonstrates the progressive increase of EQs approaching the early evolution of mammaliaforms in comparison with early Therapsida (Benoit et al 2017b(Benoit et al , 2018 (Fig. 9).…”

A new species of burnetiid (Therapsida, Burnetiamorpha) from the early Wuchiapingian of South Africa and implications for the evolutionary ecology of the family Burnetiidae

“…Posteriorly, the brain endocast descends posteroventrally so that there is a moderate degree of flexure between the forebrain and hindbrain of ~141° (Benoit et al . ). This is a lesser degree of flexion than is present in both basal burnetiamorphs ( Lemurosaurus ) and non‐burnetiamorph biarmosuchians ( Hipposaurus and Herpetoskylax ), but its prevelance in burnetiids is as yet unknown (Benoit et al .…”

“…This is a lesser degree of flexion than is present in both basal burnetiamorphs ( Lemurosaurus ) and non‐burnetiamorph biarmosuchians ( Hipposaurus and Herpetoskylax ), but its prevelance in burnetiids is as yet unknown (Benoit et al . ). The bony labyrinth of SAM‐PK‐K11112 is also largely ossified, as in other biarmosuchians and dinocephalians (Benoit et al .…”