Morphology and phylogenetic relationships of the earliest known hippopotamids (Cetartiodactyla, Hippopotamidae, Kenyapotaminae) Author Boisserie, Jean-Renaud Author Lihoreau, Fabrice Author Orliac, Maeva Author Fisher, Rebecca E. Author Weston, Eleanor M. Author Ducrocq, Stéphane text Zoological Journal of the Linnean Society 2010 2010-10-28 158 2 325 366 http://dx.doi.org/10.1111/j.1096-3642.2009.00548.x journal article 10.1111/j.1096-3642.2009.00548.x 0024-4082 5438148 KENYAPOTAMUS CORYNDONAE PICKFORD, 1983 ‘creature very close to the basal stock from which true hippopotamus was derived’; Coryndon, 1978a: 291 . ‘a genus of the Suidae similar to Bunolistriodon ’; Coryndon, 1978: 291, fig. 18:7B. Kenyapotamus coryndoni ; Pickford, 1983: 195–202 , figs 3–12, pl. 1 (figs 1, 2, 4), pl. 2 (figs 1, 2). Kenyapotamus coryndoni ; Pickford, 1990 : tab. I. Kenyapotamus sp. ; Pickford, 1990 : pl. I. Nomenclatural remark: We followed Weston & Boisserie (in press) here and used a feminized species name, this species being named after a woman, Shirley Coryndon. Geographical distribution: Kenyan Rift, Kenya , and Afar depression, Ethiopia , eastern Africa; central Tunisia , northern Africa. Temporal distribution: Late Miocene, roughly between 10.5 and 8.5 Mya. Attributed material: Nakali: KNM-NA 187 (distal half of right M 3 ) † ; KNM-NA 188 (right M 1 ); KNM-NA 192 (distal half of right lower molar); KNM-NA 194 (right lower premolar, most probably P 4 ); KNM-NA 203 (right P 3 or, maybe, P 4 ); KNM-NA 246 (right mandibular corpus with P 4 , M 1 , distal half of M 2 , M 3 ); KNM-NA 247 (apical fragment of lower incisor). Ngeringerowa: KNM-BN 1289 (lower incisor fragment); KNM-BN 1320 (fragmentary right M 1 or M 2 ); KNM-BN 1321 (right M 3 – holotype with BN 2075); KNM-BN 1322 (fragmentary right M 1 ); KNM-BN 1323 (distal half of lower premolar); KNM-BN 1483/ 1487 (right P 2 ); KNM-BN 1489 (distal half of left M 1 or M 2 ); KNM-BN 1490 (right P 2 or P 3 ); KNM-BN 1492 (right P 2 ) † ; KNM-BN 1493 (left P 4 ); KNM-BN 1494 (fragmentary palate with left M 1 and right M 2 ); KNM-BN 1715 (left P 3 ); KNM-BN 1717 (right P 1 ); KNM-BN 1802 (left P 4 ); KNM-BN 2075 (fragmentary right M 2 – holotype with BN 1321). Samburu Hills : KNM-SH 14789 (fragmentary left mandibular corpus with M 1 or, less likely, M 2 ); KNM-SH 14792 (fragmentary left mandibular corpus with M 1 and emerging M 2 ); KNM-SH 15850 (right M 1 or 2? ) † ; KNM-SH 15851 (left M 1 ); KNM-SH 15857 (fragmentary mandible with: two canine fragments, left P 3 , P 4 , fragmentary M 1 , M 2 , M 3 ; fragmentary right P 3 , P 4 , M 1 , M 2 , M 3 ); KNM-SH 18001 (fragmentary upper molar); KNM-SH 40142 (fragmentary right mandibular corpus with P 3 and P 4 ) . Emended diagnosis: Exhibits features also found in Hippopotaminae : deep median indentation of P 3 labial cervix; upper molars with parastyle confounded or very close to mediomesial mesiostyle. Salient postectohypocristid on lower molars occurring infrequently. Molar crowns relatively shallower than in Hippopotaminae . Discussion: Some examined specimens from Ngeringerowa could be related to K. coryndonae , but their fragmentary condition did not warrant certain attribution. This was the case for two molar fragments, KNM-BN 1488 and KNM-BN 1601 ( Fig. 5G ), as well as for KNM-BN 1353, a fragment of canine enamel. This fragment displays the fine ridging typical of hippopotamids, similar enamel thickness, and could be a portion of the linguomedial enamel band lining the distal groove of upper canines. If this identification was proven correct, this fragment would correspond to a canine similar in size to Archaeopotamus harvardi , i.e. larger than expected for K. coryndonae . We therefore referred these fragmentary remains to K. cf. coryndonae . Similarly, all postcranial remains identified as K. coryndonae ( Fig. 8 ) should be better referred to K. cf. coryndonae until the discovery of clearly associated dental and postcranial material. KNM-NA 250 includes a canine fragment and right M 3 ( Fig. 7E ), with morphologies similar to other specimens of K. coryndonae . However, the M 3 appears closer in size to KNM-FT 3322 ( Fig. 7F ) attributed to K. ternani (dimensions of KNM-NA 250 M 3 : mesiodistal length = 35.5 mm ; mesial lobe maximal width = 20.0 mm). Given the incompleteness of the Fort Ternan specimen and the small available sample, we prefer to attribute KNM-NA 250 to K. cf. coryndonae . The dental remains of Kenyapotamus from the Beglia Formation were found to be of similar size to the dentition of K. coryndonae and most similar to specimens from Nakali ( Pickford, 1990 ). We did not examine this material, but in our opinion its publication clearly demonstrated the occurrence of Kenyapotamus in northern African. However, it was not made clear whether the author preferred to attribute this material to ‘ K. coryndoni ’ or ‘ Kenyapotamus sp. ’ ( Pickford, 1990 : tab. I, pl. I). We have therefore retained here an attribution to K. cf. coryndonae . Closer affinities mentioned between specimens from Beglia and those from Nakali were proposed on the basis of higher molar crowns and a more robust P 4 with better developed metaconid and distal cuspids compared to those found at Ngeringerowa ( Pickford, 1990 ). We could not identify a premolar displaying clear hippopotamid P 4 features in the material from Ngeringerowa ( contra Pickford, 1983 ), whereas a figured Beglia molar ( Pickford, 1990 : pl. I E, F) exhibits a similar crown height as molars from Ngeringerowa and Nakali, with the exception of one specimen from the latter locality. KNM-NA 251 ( Fig. 5B ) is a right M 3 moderately worn (all cusps display dentine isles) exhibiting significant difference in crown height to other specimens. Despite wear, the crown of KNM-NA 251 appears relatively as high as or higher than that of unworn molars from Nakali, Ngeringerowa, and Beglia. Its hypsodonty index was estimated to be similar to that of hippopotamines ( H probably close to 80). This difference appears significant enough in morphofunctional terms to cast doubt on the attribution of this specimen to K. coryndonae . We prefer for now to refer it to K . aff. coryndonae . These proposed specimen reassignments are of little consequence to the existing taxonomy of late Miocene kenyapotamines. However, they may indicate a greater diversity amongst these kenyapotamines than initially thought, with the likely occurrence of two forms separated by crown height differences, and maybe of a third form characterized by its smaller size.