Technical Comments

Middle Miocene Hominoid Origins

Science  31 Mar 2000:
Vol. 287, Issue 5462, pp. 2375
DOI: 10.1126/science.287.5462.2375a

Ward et al. (1) ably show that samples of thickly enameled Middle Miocene hominoids that they attribute to a new genus, Equatorius, are distinct fromKenyapithecus. They fail to show, however, howEquatorius differs from Griphopithecus. In so doing, they may have missed the hominoid connection between Eurasia and Africa by 2 to 3 million years.

The authors note the presence of a well-developed buccal cingulum on the type specimen of G. darwini, an M3 from Devı́nska Nová Ves, Slovakia, that they say distinguishes this taxon from Equatorius. Samples from Paşalar andÇandır in Turkey that have been assigned to G. alpani, however, show that M3 cingulum expression is variable in Griphopithecus; indeed, specimens from these samples are quite similar to M3 specimens ofEquatorius. In addition, the upper molars of G. darwini lack well-defined lingual cingula. This is probably a shared derived character with other thickly enameled Middle Miocene hominoids, and distinguishes both Griphopithecus and the Maboko and Kipsaramon samples from Early Miocene forms. Nothing in the diagnosis of Equatorius excludes most of the specimens currently attributed to G. darwini or G. alpani; the portions of the diagnosis of Equatorius that can be compared to the current hypodigm of Griphopithecus could apply just as well to it. This is true even for the more apomorphic characters of Equatorius related to the inferior transverse torus, sublingual plane, and I2.

What Ward et al. really provide is evidence for a link between Eurasian and African thickly enameled hominoids of 15.5 million years ago (Ma) in addition to the link they describe between ∼14 MaKenyapithecus (sensu stricto) and the unnamed second taxon from Paşalar. This is most effectively communicated by recognizing two taxa—Griphopithecus for the samples from Devı́nska Nová Ves, Paşalar, Çandır, Kipsaramon, the Maboko Formation, Nachola, and the Tugen Hills, andKenyapithecus for the samples from Fort Ternan and the Paşalar unnamed second taxon.

Taxonomic issues aside, recognition of this more complex biogeographic pattern is significant: in both cases, the earlier-occurring taxon is Eurasian, not African. Both the unnamed taxon from Paşalar and a molar from Engelsweis, Germany, that cannot be distinguished fromGriphopithecus predate their African counterparts by at least 1 million years (2–4). This pattern may result from an incomplete fossil record, but it more likely means that African Middle Miocene hominoids represent successive radiations of forms dispersing from Eurasia—a pattern thought to hold for many other mammals, including muroid rodents, the bovid Eotragus, gomphotheriids, amebelodontines, carnivores, and at least one suid (5–7). The confirmation of a taxonomic and phyletic distinction between Kenyapithecus andGriphopithecus, and the proposed link betweenKenyapithecus and earlier hominoids from Paşalar, strengthen the Eurasian-origins hypothesis by providing yet another example of related hominoids that appear in Eurasia before they appear in Africa (8, 9).


In describing a new Middle Miocene ape species, “Equatorius africanus,” Ward et al. (1) use a hypodigm that includes samples from two different genera, Nacholapithecus kerioi (Afropithecini) (2–5) and Kenyapithecus africanus(Kenyapithecini) (6–8), with the latter cited as a junior synonym of “E. africanus.”Nacholapithecus and Kenyapithecus have different dietary and locomotor adaptations, however, and the description of “Equatorius” encompasses diagnostic features that are not present in both of these genera. Features aligning “Equatorius” with Afropithecus, including robusticity of the canine and breadth of the premolars, are present only in Nacholapithecus (2,4). The lower male canines of K. wickeriand K. africanus—which are identical in size and shape (Table 1)—are taller and more gracile than those ofNacholapithecus (9), and the upper premolars of both Kenyapithecus species are narrower, relative to M1 breadth, than are those of Afropithecus andNacholapithecus (4, 8, 10). The mandible of “Equatorius” is described asAfropithecus-like in having a long and proclined symphysis, but this feature is found only in K. africanus and K. wickeri among fossil apes (11). Mandibles ofNacholapithecus and Afropithecus are taller relative to corpus width (4) and have higher molar roots than those of K. africanus and K. wickeri(12).

Table 11

Dimensions (in mm) of complete lower male canines ofKenyapithecus wickeri (KNM-FT 28) from the type site of Fort Ternan, and of Kenyapithecus africanus (KNM-MB 32348) from the type site of Maboko. L, mesiodistal length; W, buccolingual width; CH, cusp height from tip to cervix on the buccal aspect. Values for ratios are multiplied by 100.

View this table:

Partial skeletons of Nacholapithecus, unlike specimens ofK. africanus and K. wickeri, lack a posteriorly retroflected medial epicondyle of the distal humerus (3,5), a derived feature distinguishing the genusKenyapithecus from all other Miocene ape genera (8,10, 13–15). Nacholapithecus was an arboreal quadruped and was more Proconsul-like in its postcrania (3, 5,16). It lacks both the terrestrial features of K. africanus—low, flat humeral head, short and straight phalanges, strong dorsal ridge of the distal metacarpal, and habitually adducted hallux (8, 10, 13–15)—and also lacks characters of the radius and cuboid that K. africanus shares only with the African apes and humans (15, 17).

The identification of “Equatorius africanus” rests on the notion that, because K. africanus is referable to“E. africanus,” K. wickeri and K. africanus are not congeneric. Such a claim, based on differences in zygomatic root and maxillary sinus heights (1, 18) and morphology of upper incisors and premolars, is not supported by evidence. All incisor variation observed in the combined K. wickeri and K. africanus sample is found within subfossil orangutan populations (19). Female P4s and P3s of K. africanus from Maboko match those of K. wickeri in shape, pattern of occlusal crests and fissures, cingulum, and cusp development (9). Relative to M1 area, zygomatic root heights of K. africanus(BMNH 16649) and K. wickeri (FT 46/47) fall below the regression lines observed for extant apes and do not exceed ranges of variation observed for modern hominoid species (8, 10). The difference between Kenyapithecus male (BMNH 16649) and female (FT 46/47) maxillary sinus height does not exceed that observed within fossil species [Sivapithecus indicus andAustralopithecus afarensis (20)] and living species (21, 22).

In light of these observations, we suggest that the new hominoid genus identified by Ward et al., “Equatorius,” is actually a combination ofKenyapithecus and Nacholapithecus.


Response: Notwithstanding Begun's claim that we failed to distinguish Equatorius from the Middle Miocene European genus Griphopithecus, we explicitly comparedEquatorius to the type series of Griphopithecusfrom Devı́nska Nová Ves, Slovakia—which, as we noted, is limited to four teeth (1). Begun, however, would substantially increase the number of possible morphological comparisons by expanding the hypodigm of Griphopithecus to include the large sample from Paşalar and a single mandible fromÇandır, both in Turkey. We chose not to use this expanded hypodigm because, without a more morphologically representative and diagnostic sample of Griphopithecus from the type site, assigning any additional samples to the hypodigm remains highly uncertain.

Even if one does include the Turkish samples, Begun's assertion that “the portions of the diagnosis of Equatorius that can be compared to the current hypodigm of Griphopithecus could apply just as well to it” is incorrect. Unlike Equatorius, lower molars from the Turkish sites mostly have prominent buccal cingulae (variation in M3 cingulum expression at Paşalar is less than Begun supposes). Paşalar upper lateral incisors lack the characteristic spiral morphology found inEquatorius, and the upper central incisors have a distinctive, prominent lingual pillar, a feature shared with later, more derived Eurasian hominoid genera (2). We acknowledge that Equatorius and the Paşalar specimens assigned toGriphopithecus share many features of dental and gnathic morphology. It strikes us as unsound practice, however, to use a taxon based on a type with almost no diagnostic morphology as a “catch basin” for Middle Miocene species spread throughout southern Europe, western Asia, and East Africa. Likewise, although our report (1) provides an implied justification for assigning the unnamed second species at Paşalar to Kenyapithecus, such a step seems ill advised before the species has been fully described or even named.

Our report (1) was not overly concerned with which taxa show the earliest connections between Africa and Eurasia. Our aim was to point out a very compelling case for these connections involving what appear to be two very closely related species. We do not understand how Begun concludes that one biogeographic scenario is more likely than another; many scenarios can be imagined that fit the available record equally well. In our view, he too easily dismisses the very real limitations imposed by the scanty and geographically limited fossil record of the Middle Miocene.

Begun's out-of-Eurasia hypothesis for Middle Miocene hominoids must be regarded as tenuous, in view of the imprecise dating of most European sites and the nearly complete lack of Early Miocene catarrhines anywhere in Eurasia. Although the latter could be another manifestation of an imperfect record, East Africa, at least, has numerous Early Miocene catarrhine lineages from which Middle Miocene hominoids can be derived. Some of these lineages also persisted into the Middle Miocene of Africa (3–5).

Benefit and McCrossin view Equatorius africanus as a chimera composed of two other taxa, the newly named Nacholapithecus kerioi (6) and “Kenyapithecus africanus, and hold that the diagnostic characters underpinning our identification ofEquatorius are not present in both of those taxa. We do not dispute the latter point and, in fact, agree with Benefit and McCrossin regarding the taxonomic status of N. kerioi, in light of reports at the recently held Kyoto conference that they cited. Our inclusion of the Nachola material in the Equatoriushypodigm was based on the best data available at the time of writing [see note 17 in (1)]. We stress, however, that the material now included in Nacholapithecus did not contribute to the formal diagnosis of Equatorius, which was based solely on the Tugen Hills specimens and on available material from Maboko Formation localities in western Kenya formerly assigned toK. africanus. Thus, differences between N. kerioiand “K.africanus enumerated by Benefit and McCrossin have no bearing on the validity of Equatorius, other than to better characterize the distinctiveness of the latter.

We did not assert, as Benefit and McCrossin suggest, that the mandible of Equatorius resembles that of Afropithecus“in having a long and proclined symphysis.” The only mandibular similarities that we note between Equatorius andAfropithecus are the presence of an inferior transverse torus and a robust mandibular corpus. [The claim by Benefit and McCrossin that a long, proclined symphysis is restricted to the two African fossil apes K. africanus and K. wickeriis wrong in any event; this morphology is also found in a mandible from Paşalar (2).] The comment also implies that we failed to note premolar differences between Equatorius and other genera, stating that “Upper premolars of bothKenyapithecus species are narrower relative to M1 breadth than are those of Afropithecus andNacholapithecus.” We cannot speak to the largely undescribed Nacholapithecus sample; we did note, however, that the upper premolars of Equatorius samples from Maboko and Kipsaramon “are longer relative to breadth”—in other words, narrower—than those of Afropithecus.

With respect to features that distinguish Equatoriusfrom Kenyapithecus, Benefit and McCrossin criticize our alleged lack of appreciation of intraspecific variability in incisor, premolar, and maxillary morphology. The morphology of upper central incisors does indeed vary within many extant and fossil taxa (7), but the temporal and geographic distribution of the variation must also be considered. The morphology of severalEquatorius upper central incisors from three different sites is remarkably uniform, and is distinctly different from the one incisor of K. wickeri.

We likewise do not contest that there is intraspecific variability in maxillary features. We have emphasized, however, the coincidence of differences between the two maxillae in question in several morphological features (including the anterior extent of the maxillary sinus, which was not mentioned by Benefit and McCrossin and has not been shown to vary according to sex), rather than intraspecific variability in any given feature. Moreover, in their analysis of zygomatic root height in the two specimens (8), Benefit and McCrossin provide no information on how they derived linear measurements from what is essentially a continuously curving surface. We can say little about the supposed similarities in premolar morphology between E. africanus and K. wickeri, as the specimens collected from Maboko by Benefit and McCrossin have not been made available for study; those from earlier collections, however, differ from the K. wickeri specimens.

The morphology of the male lower canine from Maboko may indeed be very similar to that of K. wickeri from Fort Ternan (their Table 1), but that merely implies that the Maboko canine is a specimen ofKenyapithecus rather than Equatorius. Indeed, we suggest that the Maboko sample represents two, possibly time-successive hominoid taxa: the earlier E. africanus and a later species perhaps belonging to Kenyapithecus. There are two relatively complete large hominoid humeri from Maboko, one expressing a primitive catarrhine morphology combined with specific adaptations for terrestriality (9), and the other having the derived, straight-shafted morphology characteristic of extant great apes (10). It seems inconceivable that these two fundamentally different morphological patterns can be embraced by a single genus.

Other recently recovered postcranial remains from Maboko, including the radius and cuboid mentioned in Benefit and McCrossin's critique, are also described as being more like those of extant great apes. The figured Maboko male lower canine is described as coming from the brown clay of Bed 5 at Maboko Main, a younger level than Bed 3, which has apparently produced most of the Equatorius material. While the stratigraphic level of the more recently recovered specimens has not been reported, we would not be surprised if the postcranial specimens with more derived morphology are also from Bed 5. We are confident that proper sorting of the Maboko sample will further validate Equatorius.


Navigate This Article