The cercopithecoids are a diverse group of primates, which are believed to have split from the hominoids some 25–23 Ma (Gingerich and Schoeninger, 1979; Glazko and Nei, 2003), though the split may have occurred as long as 34–29 million years ago (Steiper et al., 2004). Compared to hominoids, cercopithecoids are characterized by having bilophodont molars, which is most commonly associated with a folivorous diet and seed eating (Happel, 1988; Cuozzo and Yamashita, 2006). This molar form is thought to emphasize the use of interlocking transversely aligned blades to efficiently both shear and crush foods (Gregory, 1922; Kay and Hylander, 1978; Benefit, 2000; Rasmussen et al., 2019). One cercopithecoid tribe, the Cercopithecini, includes the vervets (Chlorocebus), guenons (Cercopithecus and Allochrocebus), talapoins (Miopithecus), Allen's swamp monkey (Allenopithecus), and the Patas monkeys (Erythrocebus). Cercopithecini are distinguished from their sister clade, the Papionini, by lacking the distal-most cusp, the hypoconulid, on the M3 (Delson, 1975). Having a M3 hypoconulid seems to be the primitive state for cercopithecoids (Delson, 1975), particularly as a hypoconulid is observed in stem cercopithecoids such as Victoriapithecus and Prohylobates (Benefit, 2002), as well as in Nsungwepithecus gunnelli (Stevens et al., 2013). Loss of the hypoconulid in papionins is reportedly rare (Delson, 1975), though it has been observed in some taxa such as Papio, Cercocebus albigena, and Macaca mulatta (Phillips-Conroy, 1982). Phillips-Conroy (1982) describes the symmetrical loss of the hypoconulid in a specimen of Macaca mulatta, resulting in its molar row resembling that of Chlorocebus in overall form and proportions. Although all cercopithecins lack the hypoconulid, it is currently unclear as to why this is the case. Presumably, the loss of this cusp relates to dietary differences between cercopithecins and other cercopithecoids, though diet varies significantly among living cercopithecins (Table 1). It could be argued, however, that the loss of a cusp would result in a decrease in overall complexity and shearing potential in the molar row (Evans et al., 2007, Kay, 1975; Bunn et al., 2011), which perhaps relates to a shift in foods that are easier to process, such as fruit. Indeed, guenons are primarily frugivorous (Fleagle, 2013), though this is not necessarily the case for all cercopithecins.

The following study investigates the functional ramifications of losing the hypoconulid in extant cercopithecins by examining patterns of molar functional morphology using dental topographic analysis (DTA). Considering dental topographic metrics are known to covary with diet in primates and other related taxa (Evans et al., 2007; Boyer, 2008; Bunn and Ungar, 2009; Bunn et al., 2011; Ledogar et al., 2013; Winchester et al., 2014; Pampush et al., 2016; Winchester, 2016a; Selig et al., 2019; Avià et al., 2022), these methods should provide a useful means for linking diet to dental form in cercopithecins. Loss of the hypoconulid may have an effect on the overall shearing ability of the molar row, but may also effectively shorten the molar row in total length given that the distal-most cusp is removed. Evidence suggests that shortening the face can have the effect of increasing bite force (Radinsky, 1981; Greaves, 1988; Spencer, 1999; Dumont et al., 2009; Freeman and Lemen, 2010), which may have been selected for among cercopithecins. Therefore, mandibular length will also be addressed in light of the loss of the M3 hypoconulid to determine if cercopithecins have relatively shorter faces compared to macaques.

As with potential differences in function, the development of the lower molars may differ between Cercopithecini and Papionini. For example, previous research notes that papionins are characterized by a molar area pattern of M1<M2<M3, where the M1 is the smallest tooth and the M3 the largest, whereas guenons are characterized by a pattern of M1<M2>M3 (Swindler, 2002; Schroer and Wood, 2015; Winchester, 2016b; Roseman and Delezene, 2019). This contrasting pattern has been explained as representing different levels of molar activation/inhibition during dental developmental (Winchester, 2016b) under the framework known as inhibitory cascade model (ICM; Kavanagh et al., 2007). Mammalian molars develop sequentially, with the precursor cells that produce the first molar extending posteriorly, giving rise to the adjacent distal molar (Kavanagh et al., 2007). The ICM postulates that a balance in inhibitor and activator activity from embryonic signaling centers in earlier developing teeth (i.e., M1 and/or the deciduous molars) affects the development of subsequent teeth (i.e., M2 and ultimately M3), and this balance produces a molar row that is proportioned in size in a predictable, linear manner (e.g., M1<M2<M3). Taxa whose molars are not proportioned linearly (e.g., like guenons, M1<M2>M3) are said to violate the expectations of the ICM (Roseman and Delezene, 2019). Winchester (2016b) suggests that it might be earlier crown termination of the M3 or a reversal of the ICM pattern among guenons that explains why the M3 is smaller than the M2 based on the predictions of the ICM. Roseman and Delezene (2019) argue that the ICM is not a good predictor of molar size covariation given that only Macaca fasicularis follows the predictions of the model, whereas the other taxa they examined, such as Cercopithecus cephus and Cercopithecus pogonias (both guenons), as well as select hominoid taxa (Hylobates lar, Gorilla, Pongo pygmaeus, and Pan troglodytes), do not. However, several other studies have found that the ICM does predict and explain molar size covariation across a broad distribution of both living and extinct mammals, and particularly when the deciduous molars are included (e.g., Kavanagh et al., 2007; Asahara, 2013; Halliday and Goswami, 2013; Schroer and Wood, 2015; Couzens et al., 2016; Evans et al., 2016; Selig et al., 2021). The following seeks to reevaluate molar area among cercopithecins in the context of the ICM by including more cercopithecin taxa and will consider how the presence or absence of the hypoconulid might affect the patterns noted among this group.

The development of individual molar cusps is also said to follow predictions of a developmental model. Similar to the ICM, the patterning cascade model (PCM) postulates that cusp development is governed by the timing of other developmental events, as well as the spacing and the size of other developing cusps (Jernvall, 2000; Salazar-Ciudad and Jernvall, 2002, 2010; Skinner and Gunz, 2010; Ortiz et al., 2018). As a multicuspate tooth germ develops, the enamel knots direct the forming and folding of the inner enamel epithelium, which is a major component of the enamel-dentine junction, a structure that acts as a blueprint of the fully developed tooth (Butler, 1956; Skinner, 2008; Ortiz et al., 2018). Enamel then forms above the enamel-dentine junction and dentine forms below the enamel-dentine junction and what ultimately become the cusp tips arise from the position of the enamel knots (Kraus, 1952; Butler, 1956; Skinner et al., 2010). Like earlier developing teeth, enamel knots produce signaling molecules that prevent the development of adjacent enamel knots (Jernvall and Thesleff, 2000). For a new enamel knot to form, it must be sufficiently distant from the earlier developing enamel knot such that it falls beyond the zone of inhibition created by that earlier structure (Jernvall and Thesleff, 2000; Ortiz et al., 2018). The PCM suggests that molar cusp formation is governed by the timing, size, and spacing of the enamel knots. Molar cusps form moving in order distally (protoconid, metaconid, hypoconid, entoconid, and ultimately hypoconulid in cercopithecoids; Swindler, 1983), meaning for the distal-most cusp to form (and particularly any accessory distal cusps), the more mesial cusps must be sufficiently distant such that they do not inhibit the formation of the ultimate cusp. Therefore, the size of the primary mesial cusps, and specifically the distance between those mesial cusps, has been shown to be a predictor of the formation of more distal cusps (Jernvall, 2000; Winchester, 2016b; Ortiz et al., 2018; Bermúdez de Castro et al., 2022). For example, Winchester (2016b) suggests that cercopithecoid mesial cusps (protoconid and metaconid together) and distal cusps (entoconid and hypoconid together) seem to be characterized by a degree of buccolingual migration moving away from one another, potentially as the inner enamel epithelium forms and folds. Therefore, the PCM would suggest that the distal cusps must be sufficiently distant from one another in order for a hypoconulid to form, or at least for a sufficiently large hypoconulid to form. Winchester (2016b) found variable support for this hypothesis in that M. fasicularis had relatively less constricted distal cusps compared to Cercopithecus mitis. However, M. fasicularis had the most prominent hypoconulids among the analyzed taxa, but did not have the greatest distance between the distal cusps as predicted by the PCM (the greatest distance was observed in Colobus guereza). Moreover, Ortiz et al. (2018) found a positive relationship between relative metacone-hypocone spacing and the development of cusp 5 (metaconule) in taxa such as Gorilla, Australopithecus, and Paranthropus. The following seeks to evaluate the same hypothesis as Winchester (2016b), but by including a larger sample of cercopithecins. It is again predicted that the cercopithecins will be characterized by more constricted distal cusps given that they lack a hypoconulid.

When it comes to the ancestral cercopithecoid molar morphotype, Delson (1975) argues that it is Macaca that is characterized by the most plesiomorphic molars. Macaca has relatively large but shallow trigonids, bunodont cusps, and a well-developed M3 hypoconulid, for example (Delson, 1975). Therefore, Macaca, and specifically Macaca fasicularis, has been chosen in the present analysis to represent the ‘ancestral’ cercopithecoid condition for comparison with the cercopithecins. The cercopithecins differ from Macaca and other cercopithecoids in missing their hypoconulid. Virtual removal of the hypoconulid from the third molars of Macaca may provide additional insight into questions of the function ramifications associated with the loss of this cusp. In this study, Macaca third molars are therefore virtually dissected and cropped such that the hypoconulid is removed to test how this changes functional, and developmental interpretations of these molars. The cropped teeth will be subjected to DTA, along with the uncropped teeth and those of the cercopithecins, to test for functional differences among the sample. Does the presence/absence of this cusp have a significant effect on the overall efficiency of the M3? Does removal of the hypoconulid change the perceived functional properties of the macaque M3? How does removal of a portion of the tooth change interpretations with regard to the ICM? Does the loss of the hypoconulid in cercopithecins have an impact on their non-conformity to the inhibitory cascade model?