Three-dimensional dental topography of fossil suids and paleoenvironmental reconstruction of earliest Vallesian (Late Miocene) sites from the Vallès-Penedès Basin (NE Iberian Peninsula)

Abstract

The earliest Vallesian (∼11.2 Ma) sites of Castell de Barberà (CB) and Creu de Conill 20 (CCN20), in the Vallès-Penedès Basin (NE Iberian Peninsula), are similarly well sampled but display differences in faunal composition that hint at paleoenvironmental differences. Most importantly, CB records one of the few co-occurrences of hominoids and pliopithecoids in the Miocene of Europe, while primates are seemingly absent from CCN20. Differences are also reflected in the more diverse suid assemblage from CB. To ascertain the paleoecological basis for such differences, here we investigate the diets of the suids from these sites—the suine Propotamochoerus palaeochoerus, the tetraconodontines Parachleustochoerus valentini and Versoporcus steinheimensis, and the listriodontine Listriodon splendens. We use 3D dental topography to infer paleodiet and provide further insight into the local paleoenvironmental differences between these sites. We measure four variables—orientation patch count, relief index, angularity, and mean surface slope—and distinguish herbivorous from omnivorous taxa, with the former categorized as browsers, mixed feeders, and grazers. The analyzed suids are similar in terms of angularity but differ in mean surface slope, orientation patch count and relief index. A canonical variate analysis classifies the CB and CCN20 suids as herbivorous (mixed feeders), except for P. palaeochoerus from CCN20, which displays greater similarities to omnivores (higher surface complexity), suggesting more varied diets in the latter locality. We conclude that, despite being roughly coeval and geographically close (∼10 km), these two sites record different local paleoenvironments—with CB representing a very humid and densely forested habitat, and CCN20 recording a more open, arid, and seasonal environment. This confirms previous inferences based on faunal composition and would explain the absense of primates from CCN20. Our study further highlights the relevance of local paleoenvironments for understanding differences in faunal composition among past communities from the same area.