Introduction
Among the various theories presented for the killing bite technique of the iconic hypercarnivore sabertooth cat, Smilodon, one is so comprehensive as to constitute a testable hypothesis, the canine shear-bite (Akersten, 1985). This chapter takes a different approach to this question, as an engineering experiment based on the morphology of the skull, upper canine, and mandible. I use a fabricated mechanical attachment based on casts of the upper and lower dentition of Smilodon from Rancho La Brea to measure the force required to bite a suitable prey proxy, using any desired sequence conforming to a chosen bite model. Discretion is required because, limited only by the original geometry, the device can easily exceed the force of the original and do the impossible. In this experiment Bison bison bison is used as a proxy for B. bison antiquus. Bison bison antiquus is the most abundant large herbivore at Rancho La Brea, and Coltrain et al. (2004) demonstrated that bison was one of the primary prey species for both and Panthera atrox at Rancho La Brea. Since both cats shared the same prey base the smaller size of Smilodon might suggest that it was a more efficient predator.
The Rancho La Brea collection facilitated the experiment by providing an abundant, accessible sample of Smilodon from which to select representative individuals. Smilodon (referred to variously as S. fatalis, S. floridanus, and S. californicus) has been the primary taxon used to study canine function in sabertooth cats, and the leading work on the subject has been described with sufficient detail to present a testable hypothesis. The model tested has been labeled the canine shear-bite (Akersten, 1985). Using that work for the given conditions, a machine was constructed to the dimensions and parameters of the skull, mandible, and canines of Smilodon, capable of replicating any proposed biting action, using sabers of the same size and shape as those in Smilodon. Using the bite model theories proposed to reproduce saber movement through cadavers permitted the forces required to pierce the skin and determine the resultant injury by necropsy. This permitted the required force to be determined experimentally. Bites targeting both throat and abdomen were evaluated. Freeman and Lemen, (2007) also determined canine strength experimentally. The approach used by others, such as McHenry et al. (2007), has used sophisticated FEA analysis to determine the strength of the skull, but relies on the assumption that the bite theory is correct, and the available force is adequate. We found this was not the case. Our objective to verify this theory experimentally under real world conditions, would require that the necessary force required be a comfortable margin of safety less than that of the tooth strength.
This chapter summarizes my observations based on the experimental mechanical device constructed. In addition, a summary of my observations on wear found on the upper canine is presented. My conclusion is that overall the canine shear-bite model proposed by Akersten (1985) is unworkable. It still remains the best study of the subject, however, and many insights presented in that paper are remarkably accurate. It was immediately obvious that we know a lot more about the fossil (and extant) cats than we do about the properties, elasticity, and strength of the prey and their soft tissues. The basic unanticipated experimental observation entails the extreme distortion of prey tissues as bite force is applied and the independent movement of hide and subcutaneous structures. In order to be successful, the sabertooth cat killing bite has to fatally injure an animal that is (figuratively) within an armor-like leather bag.