Field observations of the sooty mangabey Cercocebus atys reveal that these monkeys consume hard foods year-round. (McGraw et al., 2011, 2012, 2014). Indeed, 25–80% of the Cercocebus atys monthly diet consists of Sacoglottis gabonensis seeds, which are protected by seed casings twice as hard as cherry pits (Daegling et al., 2011). By contrast, the grey-cheeked mangabey Lophocebus albigena relies on hard foods as dietary fallbacks, and does not eat them year-round (Lambert et al., 2004). A direct comparison of the hardness of these two species diets has yet to be made, but it is clear that the species differ in the frequency with which they consume hard foods. While fallback consumption of hard foods would be expected to select for fracture-resistance in teeth, the more frequent consumption of hard foods—as occurs in Cercocebus atys—would expose teeth to greater opportunity for fracture as well as increase their risk of fatigue failure (Guatelli-Steinberg et al., 2022). For these reasons, we previously hypothesized that the molars of Cercocebus atys would show evidence of greater resistance to fracture than those of Lophocebus albigena (Guatelli-Steinberg et al., 2022).

Consistent with this hypothesis, Cercocebus atys molars exhibit: 1) greater absolute crown strength (ACS; Schwartz et al., 2020), 2) thicker enamel in occlusal basins relative to overall enamel thickness, and 3) greater flare of cusps most directly involved in phase II of the chewing cycle (Guatelli-Steinberg et al., 2022). Schwartz et al. (2020) found that ACS—the square root of the product of average enamel thickness (AET; Martin, 1985) and half the bicervical diameter (BCD)—more closely approximates a tooth's resistance to fracture than does relative enamel thickness (RET; Martin, 1985). Proportionally thicker enamel in molar occlusal basins, where food is crushed and ground, tends to be found in hard-object feeders (Kono, 2004; O’Hara, 2021; Schwartz, 2000). Finally, flare of the lateral walls of cusps most directly involved in phase II of the chewing cycle may buttress molars against laterally directed chewing forces (Macho and Shimizu, 2009). That Cercocebus atys molars exhibit these features to a greater degree than those of Lophocebus albigena suggests that they are more fracture-resistant.

Differences in molar form between Cercocebus and Lophocebus have only recently been described (Guatelli-Steinberg et al., 2022), while differences in the size of their premolars—specifically their P4s—were noted years ago (Fleagle and McGraw, 1999, 2002). Compared to members of the Lophocebus–Papio clade, species in the Cercocebus-Mandrillus clade are characterized by P4s that are larger relative to their M1s. Molarization of P4s is linked to the unique feeding niche of the Cercocebus-Mandrillus clade in which hard nuts and seeds are harvested “from the leaf litter of the forest floor” (Fleagle and McGraw, 1999: 1159). Although neither enlarged P4s (Daegling et al., 2011) nor P4s (Scott et al., 2018) are unique to hard-object feeders (Daegling et al., 2011; Scott et al., 2018), the P4s of primate hard-object feeders are large relative to their M1s (Scott et al., 2018), likely reflecting functional integration of their P4s with their molars (Scott et al., 2018).

Field observations reveal that Cercocebus atys uses its P4s in concert with its molars to crush hard foods (McGraw et al., 2011). The hard seed casings are placed on the postcanine tooth row (Supplementary Online Material [SOM] Fig. S1), where they are shattered with a powerful bite (Daegling et al., 2011). Lack of significant microwear differences between the P4s and molars of Cercocebus atys also suggests that these tooth types have similar functions, at least in terms of comminution (Daegling et al., 2011).

Here, using a limited maxillary dental sample, we asked whether the P4s of Cercocebus (Cercocebus atys and Cercocebus torquatus) differ from those of Lophocebus (Lophocebus albigena and Lophocebus aterrimus) in ways that parallel the differences in their molars. Specifically, we predicted that compared to the P4s of Lophocebus, those of Cercocebus would have greater ACS, thicker occlusal basin enamel relative to AET, and greater flare of their lingual cusp walls, as lingual cusps of upper post-canine teeth are more directly involved in phase II crushing and grinding than buccal or ‘guiding’ cusps (Hillson, 1996). Finally, given known differences in molarization of the P4s between Cercocebus and Lophocebus (Fleagle and McGraw, 1999, 2002), we predicted that the pattern of ACS change across the P4–M3 series would differ between the two genera, with the P4 being most divergent.