INTRODUCTION

The overarching function of most mammalian teeth is to mechanically process food, and, for a given food, tooth shape is a principal determinant of how effectively they carry out that function. One crucial aspect of the tooth-food interaction is how the teeth interact with each other when they meet, or occlude. However, a severe limitation on studying the fine-scale interactions between teeth during occlusion is that the act of coming into occlusion tends to obscure the very surfaces we are interested in. Occlusion also hides the food being fractured, and so impedes our ability to investigate this aspect.

This problem has been addressed by using proxies for tooth-tooth contact, mainly attrition facets on occluding teeth (Gregory 1920; Butler 1952; Mills 1955, 1967). The relative direction of tooth movement can also be determined by microfracture patterns (Gordon 1984) or dentine basin shape (Greaves 1973). Other techniques that have been used to estimate the position of non-contacting surfaces on opposing teeth include embedding and sectioning casts (Crompton and Hiiemae 1970) and simplified tooth occlusal reconstructions (Evans et al. 2001).

In a parallel set of studies, the jaw movement during mastication has been investigated using cinematographic, cineradiographic and electromyographic techniques (Crompton and Hiiemae 1970; Luschei and Goodwin 1974; Herring 1976; Gorniak and Gans 1980). These approaches have established the general pattern of jaw movement, but they do not allow visualisation of tooth-tooth contact between the postcanine teeth. This is because X-ray images cannot show proximity of teeth normal to the image, i.e., the third dimension is not resolved.

Here, we seek to relate dental occlusion studies to cinematographic studies and look more closely at the relationships between occluding tooth surfaces and the determinants of tooth movement. The current study will test whether observed jaw movements and the opposing tooth shapes themselves are capable of producing observed wear facets on teeth. Studies using dental paper have gone some way to establishing that this is true, but the application of this technique is greatly complicated when using small or highly complex teeth.

When considering jaw movement during mastication, Herring (1993) gave three possible determinants of power stroke direction:

1. the mandible being physically constrained, for example, by pre- and post-glenoid processes surrounding the mandible;
2. the mandible sliding along inclined planes formed either by the teeth or by the craniomandibular joint (CMJ); or
3. precise control of the power stroke by the muscles.

Our aim is to investigate fine-scale movements and relative position of tooth surfaces using three-dimensional reconstructions of jaw movement and tooth occlusion. This will allow us to explore the extent to which the first two factors constrain jaw movement, and therefore to what degree the movement is controlled by musculature only. Incorporating the shape of the jaw joint surfaces will identify whether pre- and post-glenoid processes significantly constrain mandible movement. This technique should achieve more accurate and revealing determinations of the trajectories of tooth movement than have previously been available using cinematography.

Evans and Sanson (2006) constructed models of several tooth forms found in carnivorans, including the carnassial and tribosphenic forms. These models assumed that the occlusal vector, the direction of movement of the lower tooth row with respect to the upper, did not change during an occlusal stroke, and therefore that the occlusal path, the path followed by the teeth during the stroke, was linear. Such tooth movement results in crests with attrition facets that are either planar (if the crest is linear) or like an extruded surface (see figure 4 in Evans and Sanson 2006). An extruded surface is one where a two-dimensional curve has been extended or 'extruded' at right angles to the plane of the curve to form a surface. Such surfaces can be likened to a plane that has been warped in only one direction, like corrugated iron or a curtain. The predominant view among dental workers, as demonstrated by Herring's (1993) description of inclined planes, implies that attrition facets are on the whole planar, and consequently that both the crests (viewed along the occlusal vector) and the occlusal path are linear. Our intention is to investigate the possible shapes of occluding attrition facets, and more specifically how facet shape affects jaw movement.

In this study we reconstruct the jaw and tooth movement in a sample of recent carnivoran species, incorporating tooth and jaw joint shape. We attempted to reconstruct chewing using only jaw rotation around the lateral axis of the condyles and lateral translation of the jaw. This system has been developed in living mammals, with carnivorans being the principal example, but it is also applicable to fossil organisms.