CONCLUSIONS

The cranial joints of Sphenodon demonstrate a variety of forms, many of which are complex. These joints include abutments (mostly midline), overlaps (laterally) and interdigitation. However, as in other lepidosaurs (e.g., Daza et al. 2010) extensive interdigitations are rare. This observation is surprising, given that recent in vivo research suggests such joints are suited to reducing compressive stress (Rafferty and Herring 1999; Herring and Teng, 2000; Rafferty et al. 2003; Markey and Marshall 2007a), something that must occur within the lepidosaur skull during feeding. Overlaps are the most common joint type in the skull of Sphenodon. Koskinen (1975) suggested that large overlapping joints might reflect more rapid growth because a larger growth surface is provided relative to the external seam width. However, Sphenodon is not a rapidly growing animal (e.g., Castanet et al. 1988): females, for example, take at least 13 years to reach sexual maturity (Gaze 2001). Moreover, as in other lepidosaurs (Bell et al. 2002), the degree of overlap between bones is larger in mature animals than in juveniles.

The greater degree of joint overlap in adult animals may instead relate to differences in stress. Adult animals possess larger adductor muscles and are capable of applying larger bite forces (Jones 2008; Jones and Lappin 2009; Herrel et al. 2010). Studies of diet suggest that post-hatchling Sphenodon eat smaller prey than adults (Ussher 1999) and, compared to females and juveniles, the larger adult males have a greater tendency to feed on vertebrates such as sea birds (Cartland-Shaw et al. 1998; Cree et al. 1999; Cree personal commun., 2004). Consumption of the latter provides fatty acids that may be seasonally important for reproduction (Cartland-Shaw et al. 1998). Therefore, the large overlapping joints are better interpreted as a way of maximising the surface area available for soft tissues that can dissipate and/or redistribute stress while maintaining the rigidity of the skull. Such joints associated with the temporal bars maybe be particularly important during long-axis bending and torsion of the skull.

The potential for assessing the capacity for "skull kinesis" (active or passive) using skeletal material alone is admittedly limited (Schwenk 2000; Metzger 2002). Nevertheless, there is nothing in the morphology of the cranial joints of Sphenodon to suggest they could accommodate or promote any of the forms of "active" cranial kinesis frequently discussed in the literature (e.g., mesokinesis, amphikinesis). A small degree of metakinesis remains possible in hatchlings. As others have said (e.g., Ostrom 1962; Iordansky 1990; Rieppel 1992; Holliday and Witmer 2008), observations of live juveniles are necessary to understand kinesis in Sphenodon. However, because hatchling bones are less mineralised than those of adults (e.g., Erickson et al. 2003), imaging resolution needs to be sufficiently fine to distinguish between bending within bone and between individual bones. The most likely location of potential flexibility is that between the base of the premaxillae and rest of the skull. Slight expansion at these joints may be necessary to accommodate anterior loading of the premaxillae or the impact on the premaxilla by the lower jaw during prooral shearing.