Smilodon sabertooth bite marks, new perspectives on prey-predator models in South American Pleistocene

Traditionally, saber-toothed cats, particularly species of the genus Smilodon, were considered unable to bite through bones due to the apparent fragility of their teeth. We analyze here megafauna fossil remains of Salto de Piedra palaeontological site (Humid Pampas), one of the few extensively radiometrically dated Pleistocene sites in Argentina and systematically excavated. Salto de Piedra is located near contemporary human occupation sites that have yielded megafauna remains. This spatial and chronological proximity provides an exceptional scenario to compare the prey ranges hunted by humans and by Smilodon.

Here we provide Smilodon populator tooth mark patterning on their megaherbivore preys. Mark morphology demonstrates that Smilodon canines penetrated and processed cortical bone, refuting interpretations that considered them functionally fragile. These tooth marks have been compared with other carnivores and saber-toothed felid fossil sites. This spatial and chronological control provides direct evidence of predatory activity and constrains its temporal span, enabling evaluation of the ecological role of this hypercarnivore.

This extraordinary discovery provides direct evidence of Smilodon 's predation, and confirms that it was an apex predator, focused almost exclusively on megafauna. These tooth marks can be identified and extrapolated to other fossil sites. Furthermore, this diagnostic tooth-mark pattern enables reconstruction of prey preferences. Our results indicate that Smilodon exerted sustained predatory pressure across a broad range of megaherbivores. Despite the spatial proximity and temporal overlap between Smilodon predation (22.2-13.3 ka BP) at Salto de Piedra and human arrival to this area (~ 14 ka BP), no interactions between these two predators has been observed indicating certain avoidance between them. The complex predator-prey trophic interactions during the Late Pleistocene-Holocene transition in the Pampas region of South America, support multicausal models, integrating ecological and anthropogenic drivers.