Introduction

Analyses of functional morphology typically apply phenomenological methodologies, where investigators seek to identify links between morphological and ecological data, or mechanistic approaches, where engineering principles are applied to explain how and why particular morphologies are associated with specific behaviours and ecologies. Often these two are combined within a comparative context.

Although there is certainly debate over basic assumptions and the degree to which structure might be expected to predict function (Gould, 2002), for the majority of biologists, determining relationships between form and function is fundamental to understanding the evolution of behaviours and ecologies, the nature of morphological convergence, and the prediction of habitus in living and extinct taxa.

With respect to the mammalian carnivore skull, numerous studies invoking a range of phenomenological and mechanistic approaches have identified some correspondence between these variables (Savage, 1977; Buckland-Wright, 1978; Radinsky, 1981a, 1981b; Werdelin, 1986; Van Valkenburgh, 1989; Therrien, 2005a; Wroe et al., 2005; Christiansen and Wroe, 2007; Wroe and Milne, 2007). However, the degree to which any skull might be optimised for feeding is not well understood. The vertebrate skull is not simply a food- processing mechanism, it also houses major sensory and neural apparatuses (Dumont et al., 2005), and, in mammalian carnivores, it may be subject to considerable external stresses generated in the subjugation and killing of prey (Preuschoft and Witzel, 2004). Consequently, skull morphology may represent compromise between various competing influences (Hylander et al., 1991; Preuschoft and Witzel, 2004). Identifying such compromise is difficult to achieve using traditional methods.