Skeletal and dental morphology supports diphyletic origin of baboons and mandrills

Abstract

Numerous biomolecular studies from the past 20 years have indicated that the large African monkeys Papio, Theropithecus, and Mandrillus have a diphyletic relationship with different species groups of mangabeys. According to the results of these studies, mandrills and drills (Mandrillus) are most closely related to the torquatus–galeritus group of mangabeys placed in the genus Cercocebus, whereas baboons (Papio) and geladas (Theropithecus) are most closely related to the albigena–aterrimus mangabeys, now commonly placed in the genus Lophocebus. However, there has been very little morphological evidence linking mandrills on the one hand and baboons and geladas on the other with different groups of mangabeys. In a study of mangabey locomotion and skeletal anatomy, we have identified features of the postcranial skeleton and the dentition that support the molecular phylogeny and clearly link mandrills with Cercocebus and Papio with Lophocebus. Moreover, the features linking Cercocebus and Mandrillus accord with ecological studies of these species indicating that these two genera are a cryptic clade characterized by unique adaptations for gleaning insects, hard nuts, and seeds from the forest floor.

Mangabeys are a group of large African monkeys characterized by moderately projecting snouts, large incisors, hollow cheek bones, long limbs, and long tails (1–3). They are found throughout much of sub-Saharan Africa from Senegal on the west to Kenya and Tanzania on the east (4). Traditionally placed in a single genus Cercocebus (5–6), these monkeys are commonly divided into two species groups—the terrestrial species galeritus, torquatus, atys, and agilis in one group, and the arboreal species albigena and aterrimus in another group. The three largest African monkeys, mandrills, baboons, and geladas have long been considered a single radiation of closely related species, commonly placed in three separate genera Mandrillus, Papio, and Theropithecus. In 1976, immunological studies by Cronin and Sarich (7) and Hewett-Emmett et al. (8) demonstrated that the two species groups of mangabeys were polyphyletic with respect to other papionins. The albigena group was more closely related to baboons and geladas, whereas the torquatus–galeritus group was more closely related to mandrills and macaques. Groves (2) reported numerous cranial differences between the two species groups and urged separation of the albigena group into a separate genus, Lophocebus. Most recently Nakatsukasa (9–11) has documented skeletal differences between the two genera of mangabeys that distinguish the more terrestrial (12–15) Cercocebus from the more arboreal (16–19) Lophocebus. Subsequent molecular studies, especially by Disotell (20, 21) and colleagues (22, 23) have confirmed the polyphyly of the mangabeys and clarified that the central African mandrills and drills (Mandrillus) are the sister taxon of Cercocebus whereas Lophocebus, Papio, and Theropithecus form a separate, unresolved clade.

Despite considerable molecular evidence that mandrills are phylogenetically distinct from baboons and geladas and morphological studies documenting differences between Cercocebus and Lophocebus, there is limited morphological evidence supporting the Cercocebus–Mandrillus clade (3, 20). In the course of a study designed to identify skeletal features associated with the different locomotor habits of the two mangabey groups (24), we found many osteological features that distinguish the more arboreal Lophocebus on the one hand from the terrestrial Cercocebus as well as all the larger papionins, (Mandrillus, Papio,and Theropithecus). However, in addition to the features that commonly distinguish arboreal and terrestrial primates, we found features of the postcranial skeleton that link Cercocebus uniquely with Mandrillus and also link the arboreal Lophocebus with the terrestrial Papio and Theropithecus (Table 1; Fig. 1).

Table 1.

Morphological and dental features distinguishing Cercocebus and Mandrillus from Lophocebus and Papio

Index	Cercocebus	Mandrillus	Lophocebus	Papio	P value
(Supinator crest height/humerus length) × 100	35.2  (1.6), 6	33.2  (0.85), 6	31    (3.34), 6	30.2  (2.68), 5	0.002
(Deltoid plane width/humeral head width) × 100	88.3 (6.7), 6	84   (10.7), 6	73.1  (8.13), 6	70.1  (9.8), 5	<0.001
(Coronoid width/articular notch width) × 100	54.9 (6.14), 6	55.6  (8.36), 6	71.9  (9.77), 4	70.7  (7.13), 5	<0.001
(Tibia mid AP diameter/mid ML diameter) × 100	132.3 (1.29), 6	123.5  (.964), 6	152.5  (0.77), 4	144.2  (4.32), 5	<0.001
(Scapula maximum height/maximum length) × 100	90.9 (8.22), 6	91.2  (3.67), 6	80.9  (3.49), 7	72.9  (4.32), 5	0.002
(Ilium minimum width/maximum acetabular diameter) × 100	111.8 (8.85), 5	117  (9.8), 5	86.2  (6.56), 4	100.6  (1.17), 6	<0.001
(P4 m-d × b-l/M1 m-d × b-l) × 100	90.3 (10.9), 10	95.3  (13.66), 12	57    (7.13), 13	61.1  (9.2), 6	<0.001
(P4 m-d × b-l/M1 m-d × b-l) × 100	85.6 (6.4), 10	86   (14.4), 12	57.6  (5.3), 13	55.5  (4.2), 6	<0.001

The means of Cercocebus and Mandrillus and those of Lophocebus and Papio were compared using the Games and Howell method and were not significantly different. Subsequently the combined means of Cercocebus–Mandrillus and Lophocebus–Papio were compared using Student’s t test. All measurements are of males. Values given as mean (SD), n. m-d = mesiodistal length, b-l = buccolingual breadth. 

Figure 1.

Morphological and dental features distinguishing Cercocebus, Mandrillus, and Macaca nemistrina (M) from Lophocebus, Papio, and Theropithecus (T).

Compared with Lophocebus and Papio, Cercocebus and Mandrillus are characterized by a scapula that is relatively deep, with an expanded supraspinous fossa and prominent inferior angle. The shaft of the humerus of Cercocebus and Mandrillus is characterized by a very broad deltoid plane, a proximally extended supinator crest, and a broad brachialis flange (Fig. 2). The distal articulation of the humerus of Cercocebus and Mandrillus is characterized by a narrow olecranon fossa with a deep lateral ridge. The ulna articulation is characterized by a narrow coronoid process and a relatively large radial notch. Compared with Lophocebus and Papio, the ulna and radius of Cercocebus and Mandrillus are distinguished by prominent interosseus lines. Overall, the upper extremity of Cercocebus and Mandrillus is marked by skeletal features indicative of powerful limb adduction and flexion of the elbow and digits. The latter accords with earlier observations that mandrills have a much greater forelimb flexor mass than Papio (25).

Figure 2.

Anterior view of the left humerus of Mandrillus, Cercocebus, Lophocebus, and Papio showing the prominent deltoid plane and proximally extending supinator crest in Mandrillus and Cercocebus.

The lower extremity of Cercocebus and Mandrillus is distinguished by a relatively robust ilium, a femur with a reduced gluteal tuberosity, subequal and sharp borders on the patellar margins, and a tibia with a more rounded midshaft dimension. The most striking dental feature that unites Cercocebus and Mandrillus is their common possession of relatively large, rounded posterior premolars in both the lower and upper dentition (Table 1, Fig. 3). Both the large size and the extreme wear of these teeth indicate that they are used in powerful crushing.

Figure 3.

Right upper dentitions of Cercocebus, Lophocebus, Mandrillus, and Papio showing the enlarged posterior premolars of Cercocebus and Mandrillus.

Because of their rapid terrestrial locomotion and shyness resulting from their susceptibility to human hunting, Cercocebus and Mandrillus are among the most poorly known of all nonhuman primates. Nevertheless, there are several unusual shared features of their foraging behavior that suggest a similar overall ecological adaptation associated with their unique anatomical adaptations. Both Cercocebus and Mandrillus are terrestrial monkeys that forage predominantly on the forest floor, where they manually search through the decomposing wood of fallen trees and the leaf litter for arthropods and fallen fruit seeds and nuts (26–29) but also climb trees for arboreal foods. Mandrills have been reported to regularly rip apart rotten wood and termite mounds in search of food. This powerful manual foraging as well as their habit of climbing vertical tree trunks is probably responsible for their prominent development of forelimb flexor muscles. Both mandrills and Cercocebus regularly consume extremely hard nuts and seeds that they find on the forest floor (18, 26–29). Hard nuts and seeds that can lie on the forest floor for months without decomposing are the major food of mandrills in Cameroon during the dry season when fruits are scarce; in Ivory Coast, Cercocebus are most easily located by the loud sound of cracking nuts with their teeth. The suite of dental and postcranial features that enable Mandrillus and Cercocebus to uncover and break open hard nuts and seeds that they harvest from the leaf litter of the forest floor enables this group of African monkeys to utilize a unique adaptive niche in the African rainforests.

Identifying which of these features distinguishing the Cercocebus–Mandrillus clade from Lophocebus and Papio are derived and which represent the primitive condition can only be evaluated after a broader phylogenetic study of cercopithecine morphology. However, a simple comparison with a single individual of Macaca nemistrina, a member of the immediate outgroup to mangabeys and baboons, indicates that the skeletal features characterizing Cercocebus and Mandrillus are probably primitive for papionins (Fig. 1). Indeed, the foraging strategies and locomotor adaptations of Macaca nemistrina have been noted to be similar to those of mandrills and terrestrial mangabeys (ref. 30 and N. Shah, personal communication). In contrast, premolar enlargement in Cercocebus and Mandrillus seems to be a derived feature of that clade.

Any attempt to identify morphological homologies and synapomorphies within a closely related group of animals ultimately depends on the phylogenetic relationships among the taxa (31). Although such a phylogenetic criterion is potentially circular, it is nevertheless inevitable. For this group, the molecular evidence supporting a Cercocebus–Mandrillus clade is extremely strong in contrast with the largely impressionistic traditional morphological evidence that has been advanced so far in support of alternative phylogenies (22). A more comprehensive phylogenetic analysis of this group, including the many fossil papionins, is long overdue. Clearly there are some morphological features that distinguish the large-bodied African papionins from mangabeys (5) and many that distinguish the arboreal Lophocebus from the terrestrial taxa (24) as well as those that seem to accord with the molecular phylogeny. How these morphological features are best characterized and analyzed in conjunction with the molecular data are difficult issues with no easy answers.

In recent decades, molecular systematics has identified many cryptic species and clarified a variety of debated issues in primate phylogeny. The discovery of a diphyletic origin of mandrills and baboons among the mangabeys is one of the most striking departures from traditional views of primate phylogeny and has not been widely accepted. However, careful examination of the postcranial skeleton and dentition of these monkeys reveals a suite of morphological features that support the molecular phylogeny and identify Cercocebus and Mandrillus as a cryptic clade of African monkeys with unique adaptations to terrestrial foraging. In the absence of a good molecular phylogeny for this group of primates, it is quite likely that the skeletal and dental adaptations of this clade of poorly known and difficult-to-study primates would have gone unrecognized.