Brachypotherium minor, Geraads & Miller, 2013

Geraads, Denis & Miller, Ellen, 2013, Brachypotherium minor n. sp., and other Rhinocerotidae from the Early Miocene of Buluk, Northern Kenya, Geodiversitas 35 (2), pp. 359-375 : 361-366

Brachypotherium minor n. sp.

Brachypotherium sp. Geraads, 2010: 670 .

HOLOTYPE. — KNM WS-46072, relatively complete skull, lacking the premaxillae and right cheek teeth, but preserving the left P2−M3. Housed in the National Museums of Kenya, Nairobi.

ETYMOLOGY. — In reference to its small size.

EXAMINED MATERIAL. — Except for the specimens mentioned below under “ Rhinocerotidae gen. et sp. indet.”, we assign all the Rhinocerotidae material from Buluk to this new species.

TYPE HORIZON. — Buluk Member, Bakate Formation.

TYPE LOCALITY. — Buluk, Northern Kenya, Early Miocene (c. 17 Ma), 4°15’N, 36°36’E.

DIAGNOSIS. — A Brachypotherium of relatively small size. Skull short and broad, dorsal profile almost flat, temporal lines far apart, nasals not much shortened, perhaps bearing (a) small terminal horn(s), orbit does not reach very far anteriorly. No sub-aural contact between postglenoid and post-tympanic processes. Upper teeth with poor development of accessory crests, protoloph and metaloph of premolars directed transversely and of even thickness, protocone of molars not much constricted, antecrochet moderate. Upper I1 and lower i2 sexually dimorphic.

DESCRIPTION

Skull

The holotype skull KNM WS-46072 ( Fig. 2 ) is the most instructive specimen. It is not much crushed, but is slightly affected by what White (2003) called “expanding matrix distortion” (EMD), i.e. the skull consists of a mosaic of fragments that are not contiguous, but are separated by intervening matrix, resulting in some inflation of the specimen and increasing its apparent size.

Measurements (in mm): length from condyle to front of P2 = 460; bizygomatic width 2 × 175- (width increased by EMD); maximum occipital width = c. 240; length from condyle to rear of M3 = 250; bicondylar width = 112; length from tip of nasals to top of occipital = 520- (length increased by EMD).

In lateral view, the top of the skull is slightly concave from the top of the occipital to the anteorbital area, where preservation is not very good, and it is likely that the anterior part of this profile, to the tip of the nasals, was approximately straight. The occipital plane is distinctly inclined forwards, so that the nuchal crest is more anterior than the occipital condyles. The zygomatic arch is robust, but not extremely so, and has a gentle sigmoid curvature; its ventral border has a large notch behind the orbit. The latter is not well preserved; its anterior border is above the rear part of M1. The nasals have been somewhat crushed down into the nasal notch; therefore, the outline of the latter is not clear, but its bottom was above P2/P3. The nasals show a hint of dorsal convexity and of longitudinal swellings that suggest that they might have born some kind of small horn, but this is far from certain, as the preserved part of their surface is smooth.

In dorsal view, the skull is broadest at orbital level; behind this, the transverse diameter of the fronto-parietal roof, between the temporal lines, decreases slowly caudally, but it remains broad even in its narrowest part. The occipital has a flat top border between the temporal lines. Rostrally, the decrease in width is not abrupt either, as the nasals are well anterior to the orbit.

In ventral view, the well preserved sub-aural region shows that the postglenoid process remains separated by several millimeters from the transversely elongated paroccipital process, but the rest of the cranial base is poorly preserved. We do not know whether the vomer and basioccipital were rounded or keeled. The location of the condylar foramen is not quite clear, but it is probably rather close to the condyles. The choanae reach the level of the metaloph of M2.

Upper teeth

The cheek-teeth are well-worn, but their main features remain visible. There is an alveolus for a P1 or DP1. All teeth have a flat or very slightly concave buccal wall behind the moderate paracone fold. The metaloph of P2 is directed slightly mesi - ally and comes into contact with the protoloph. On P3, the protoloph and metaloph are long and transverse, without accessory crests (save perhaps for a vestigial crochet) or constriction, so that their thickness is regular for most of their length; the lingual cingulum is faint, except at the opening of the central valley, where it is stronger. P4 displays the same features, except that it has a small crochet and a hint of protocone constriction. On the much worn M1, constriction of the protocone is distinct but remains shallow, as on M2 and M3, which have a small crochet. M3 is triangular, with a robust paracone fold.

Its premaxillae being absent, the upper incisors are unknown in the holotype, but there is no doubt that they were present, as they are represented by several isolated specimens. Two of them are much larger than the other three ( Table 1; Fig. 3 A-E), and we assume that they represent two male and three female individuals, because such sexual dimorphism (or rather “dimetrism”, as there is no major morphological difference between the male and female upper incisors) is well known in brachypotheres (e.g., Roger 1900; Heissig 1972b; Cerdeño 1993), and because the non-brachypothere rhino of Buluk (see below) is a rare form that is unlikely to be represented by three upper incisors. On the male specimens, the root is shorter than the mesio-distal length of the crown. KNM WS-11 is unworn; in occlusal view, the crown is long and of regular width; in WS-147, instead, the central part of the crown is distinctly broader, but none of these specimens match any of the types illustrated by Antoine (2002: fig. 82). In both male and females, wear first affects the buccal part of the crown in the mesial half, and then proceeds distally and buccally.

Upper cheek teeth are represented by a number of other specimens, so that the morphological and metric variations can be estimated ( Table 2; Fig. 3 F-I). None bears cement. KNM WS-12633 ( Fig. 3H ) is a left maxilla including the rounded with a *) from Buluk.

ventral orbital border; its mesial border reached farther rostrally than in the type, above the mesial part of M1. The P3-M3 series shows almost exactly the same features as the holotype, except that the lingual cingulum is better indicated, especially on the molars, and that the metaloph is slightly longer than the protoloph on P4, instead of slightly shorter. WS-12858 ( Fig. 3G ) is a series P3-M 1 in medium wear. The well-preserved premolars have a stronger lingual cingulum than in the previous specimens, but are otherwise similar in their long, simple, parallel lophs showing no indication of a constriction or lingual connection, and with a small crochet as the only accessory spur. It is also present on M1, of which, unfortunately, the protocone is broken off; the distal part of the ectoloph is directed somewhat disto-buccally, so that the buccal wall is more concave than in the previous M1s. WS-12849 is a series P2-P4 (buccal part of P4 missing), again with the same characters, but completely lacking the lingual cingulum. WS-133 ( Fig. 3I ) is a series M1-M3, moderately worn; the buccal walls of M1 and M2 also have a paracone fold and a concave distal part; all molars have a small crochet and a robust but short antecrochet, less distinct on M3. WS-49467 ( Fig. 3F ) is a toothrow P4-M2; on P4, the protoloph and metaloph are slightly convergent lingually, but remain far from each other; the antecrochet is poorly indicated on the molars.

KNM WS-143 is a very small incompletely cleaned DP2, with a poor central buccal rib and no other buccal fold, and short crochet and crista that remain far from each other, no prefossette, a transverse metaloph and a long, oblique protoloph. WS-12851 is a P4 similar to those of WS-46072 and WS- 12633 in its parallel lophs and some indication of an incipient antecrochet. WS-30722 is another P4 on which the antecrochet is better indicated, but remains moderate; the lingual cingulum is weak. WS-30726 is still another P4, unworn, also with a weak lingual cingulum and poorly expressed antecrochet; its height (42 mm) is intermediate between its length and width and the tooth can therefore be qualified as mesodont.

KNM WS-12844 is a slightly worn M1 or M2 with distinct pinching of the protocone, moderate antecrochet, and small crista; as on the molar series mentioned above, the parastyle is short and much less buccal than the robust paracone rib. The tooth is about as high as it is broad. WS-99 is another upper molar, in medium wear, with the same characters; the distal part of the buccal wall is distinctly concave. There are three M3s (KNM WS-14, WS-15 and WS-12848) that are clearly triangular but with some basal distobuccal expansion, a weak crochet, and a poorly expressed antecrochet.

On the whole, these upper cheek-teeth are notable for their simplicity, mesodonty, poor development of accessory structures, variable cingulum, transverse orientation of the protoloph and metaloph, and lack of lingual contact between them.

Mandible

The best mandibular specimen is KNM WS- 45979 ( Fig. 3J ); it lacks the rostral part and most of the ascending ramus, but the large size and great depth of the corpus anterior to the cheek-teeth leave no doubt as to the presence of large incisors. The ventral border of the corpus is virtually straight for its entire length. The mental foramen is below the mesial root of p3. From what remains of the posterior part, it is likely that the gonial area was at least somewhat expanded. The mandible preserves p4-m 2 in medium wear ( Fig. 3J 2 ). Rhinoceros lower teeth are less distinctive than upper ones, but these ones display some typical features: on the molars, the paralophid is short lingually, the protolophid is oblique rather than transverse, the hypolophid is more crescent-shaped than L-shaped, and the ectoflexid (buccal groove) is shallow. The premolars were certainly short compared to the molars; from its alveoli, p2 was quite small, and there was certainly no p1.

Lower teeth

KNM WS-12850 ( Fig. 3K ) and WS-12861 are p2s; they are small, simple, with a reduced paralophid and shallow valleys. WS-146 consists of several mandibular pieces with unworn p3, p4, m1, and partial m2, showing the same characters; like the upper ones, they are mesodont, being about as high as they are long. The oblique lophids and shallow ectoflexid are also quite clear on the m1-m2 WS-23 and on the large m2, WS-136. A few isolated lower cheek-teeth display similar features. Measurements of lower cheek-teeth are given in Table 3.

KNM WS-12866 is a large fragment of the crown of a large lower i2 (diameters: 42 × 24). In its unworn part, the cross-section is amygdaloidal, i.e., the dorsal face is as convex as the enamel-covered ventral one. It is likely from a male individual, while the much worn i2 WS- 12864 ( Fig. 3L ), which preserves its very long root, is much smaller (26.5 × 17) and therefore probably represents a female.

Post-cranials

Rhinocerotid post-cranial remains are rare at Buluk. KNM WS-13021A and WS-13021B are two distal ulnae. The latter is more robust and the radial facet is more vertical; there is a strong tuberosity on the anterior face but no lunar facet, as in European Brachypotherium ( Antoine 2002) and no accessory facet for the radius is visible; we tentatively assign it to Brachypotherium .

KNM WS-12857 is a distal central metapodial, with a marked plantar concavity above the articulation

KNM WS-12853 and WS-12854 are two right astragali ( Fig. 3M, N ) that share the same low and broad proportions ( Table 4), a poorly concave ectal calcanear facet with a marked distal extension, and a much reduced or absent distal calcanear facet. WS-12854 differs in that, unlike in most rhinos, the proximal profile of the lateral lip of the trochlea is flat or even slightly concave in front view, and the cuboid facet is much narrower compared to the navicular one. It is likely that these differences reflect a different dispatching of the weight on the digits, but the similar size and proportions of these astragali prompt us to assign both of them to Brachypotherium .

Although a second rhino species is certainly present at Buluk (see below), we assign all specimens described above to a single species. Features of the upper teeth, as well as those of the lower ones, are homogeneous, and there is no reason to doubt their association.

Figures

FIG. 2. — Brachypotherium minor n. sp., holotype skull KNM WS-46072 from Buluk in: A, dorsal; B, occipital; C, left lateral views; D, occlusal view of left P2-M3. Scale bar: A-C, 20 cm; D, 10 cm.

FIG. 3. — Brachypotherium minor n. sp. from Buluk: A-E, upper I1s in labial view; A, WS-11 (male?); B, WS-147 (male?); C, WS-12862 (female?); D, WS-139 (female?); E, WS-12860 (female?); F, P4-M2, WS-49467; G, P3-M1, WS-12858; H, P3-M3, WS-12633; I, M1- M3, WS-133; J1, mandible WS-45979, medial view; J2, same, occlusal view of p4-m2; K, p2 WS-12850; L, left female i2 WS-12864, lateral view; M, right astragalus WS-12853; N1, right astragalus WS-12854, cranial view; N2, same, plantar view; N3, same, distal view. Scale bar: A-I, J2-N, 10 cm; J1, 1.25 cm.