Neogene and Pleistocene African suids displayed convergent evolutionary trends in the third molar (M3) morphology, with increasingly elongated and higher crowns through time. While these features can prevent premature loss of masticatory functionality and potentially increase long-term reproductive success, changes in dental occlusal traits such as enamel complexity and thickness can also improve chewing efficiency and increase short-term energetic return. While both long-term and short-term benefits can contribute to the thriving of a lineage, the selective pressures associated with each category can be different. To examine how crown elongation correlates with these functional occlusal traits, we selected M3s of Kolpochoerus, Notochoerus, and Metridiochoerus from Kenya and South Africa, dated between 3.0 Ma and 0.4 Ma. To account for dental wear, we used micro-CT imaging of unworn/slightly worn M3s to simulate wear progression within each tooth. We compared morphometric representatives of occlusal enamel complexity and thickness among the specimens following their respective wear trajectories. We found that M3 elongation correlates with higher occlusal complexity and thinner enamel in Notochoerus and Metridiochoerus lineages through time. In Kolpochoerus, enamel complexity and thickness were generally maintained through time, despite M3 elongation. The differences in M3 morphometric trends suggest that Kolpochoerus likely experienced a different set of selective pressures on functional occlusal traits compared to Notochoerus and Metridiochoerus. The shared evolutionary trends of M3 specialization among Notochoerus and Metridiochoerus suggest similar selective pressures on their chewing efficiency and the possibility of a dietary niche overlap in more xeric habitats.

Micro-CT images of Kenyan and South African materials were collected using the Nikon XTH 225 high resolution X-ray CT scanner (2000 × 2000 pixels) at the Evolutionary Studies Institute, University of the Witwatersrand, South Africa, at 120-140 kV and 140-150 μA settings. The Sus scrofa specimens were scanned at PACEA / University of Bordeaux, using a General Electrics (GE) Vtome x|s X-ray microtomograph, at 100-120 kV and 200 μA settings. The Phacochoerus africanus specimens were scanned at the National Institute of Applied Sciences of Lyon, using a GE Phoenix Nanotom 180 at 150 kV and 90 μA. For fossil specimens, a 1 mm copper filter was used to limit the beam hardening effect. For extant specimens, a 0.1 mm copper filter was used.

Within each specimen, between 8 and 16 virtually simulated “occlusal surfaces” (slices) were created in Avizo 7.1 (Visualization Science Group) perpendicular to the growth axis of the tooth, using the mesial second pair of pillars as the reference. The number of simulated “occlusal surfaces” was determined based on the crown height of the specimen, and the spacing between “occlusal surfaces” is consistent within each specimen. After the slices were generated, they were manually processed in Adobe Photoshop CS6 to create segments of enamel, dentine, and coronal cementum if present (step by step Adobe Photoshop protocol is presented in Appendix S2). The segmented images were then exported to the open-source image analyzing software Fiji for measurements of occlusal features (step by step Fiji protocol is presented in the Supplementary document).

Please refer to ReadMe file.