GIRAFFOIDEA Simpson, 1931

Pickford, Martin, Attia, Yousri Saad & Ghany, Medhat Said Abd El, 2001, Discovery of Prolibytherium magnieri Arambourg, 1961 (Artiodactyla, Climacoceratidae) in Egypt, Geodiversitas 23 (4), pp. 647-652 : 648-650

Superfamily GIRAFFOIDEA Simpson, 1931 Family CLIMACOCERATIDAE Hamilton, 1978

(= CLIMACOCERIDAE Hamilton, 1978 ) Genus Prolibytherium Arambourg, 1961 Prolibytherium magnieri Arambourg, 1961 MATERIAL EXAMINED. — Frontal appendages with part of braincase (CGM 82975) ( Fig. 1 ); edentulous left mandible with roots of m1 to m3 (CGM 82976); distal left humerus (CGM 30794); distal radius (CGM 16016); vertebra (CGM 30792).

DESCRIPTION

The most diagnostic specimen is the pair of frontal appendages attached to the brain case ( Fig. 1 ). The morphology of the frontal appendages, their thickness, their growth lines, their continuity over the brain case and their extension distal to the nuchal bone are all features unique to Prolibytherium magnieri . In ventral view, the dorsal portion of the neurocranium is preserved and shows the impressions of the brain on its inner surface. The brain case is small and has thin lateral walls as in material from Gebel Zelten ( Arambourg 1961; Hamilton 1973, 1978a, b). The brain case measures 57.5 mm from side to side, and the overhang of the frontal appendages beyond the nuchals is 31.5 mm.

The edentulous mandible, and postcranial bones are of a size range compatible with Prolibytherium , although in the absence of associated elements there must remain some doubt about the identification of the specimens. The distal humerus is typically ruminant in morphology. The mediolateral diameter of the distal articular facet is 41 mm. A specimen assigned to this species by Hamilton (1973) measured 35 mm across the epicondyles. The distal radius and vertebra are also typically ruminant in morphology and their size indicates that they may represent Prolibytherium .

SYSTEMATIC POSITION OF PROLIBYTHERIUM Hamilton (1973) placed Prolibytherium in the Sivatheriidae Hamilton, 1973 ( Giraffoidea ) but the same author ( Hamilton 1978b) later exclud- ed the genus from the Giraffoidea because the lower canine was unknown. The cheek dentition of Prolibytherium is poorly known, because most of the described specimens are heavily worn or broken ( Hamilton 1973), but the little that can be discerned from the remains indicates that they are closer in overall morphology to those of Climacoceratidae ( Climacoceras [ MacInnes, 1936] and Orangemeryx [ Morales, Soria & Pickford, 1999]) than to Giraffidae . As far as is known, Orangemeryx does not possess a bilobed lower canine, and it is thus not a giraffoid in the sense of Hamilton (1978b). Climacoceras gentryi Hamilton, 1978 is reported to possess a bilobed canine ( Hamilton 1978b), but the association of the tooth with the mandible may not be a primary one ( Morales et al. 1999) and further evi- dence is required to settle the matter. Skull M 15543 in the NHM, London, the holotype of Nyanzameryx pickfordi Thomas, 1984 , was placed in the Climacoceratidae by Thomas (1984). However, as it has been pointed out by McCrossin et al. (1998), it could well belong instead to a primitive bovid, because the left frontal apophysis, in spite of the reconstruction, reveals a transition between the pedicle and the part that would have been covered by the horn, a defining feature of Bovidae .

Prolibytherium possesses several features of the skull ( Hamilton 1978b) that are reminiscent of Prodremotherium Filhol, 1877 , as do the genera Sperrgebietomeryx ( Morales, Soria & Pickford, 1999) and Orangemeryx ( Morales et al. 1999) from southern Africa ( Fig. 2 ). The skull of Climacoceras is still poorly known but, apart from the frontal appendages, it appears to be close to that of Orangemeryx . All in all, the available evidence suggests a closer relationship between Prolibytherium and Climacoceras than between the former genus and any of the Giraffidae . For this reason, we consider it likely that Prolibytherium is a Climacoceratidae and that the genera Sperrgebietomeryx and Orangemeryx belong to the same family. Counter to this suggestion is the observation that the neck vertebrae of Prolibytherium appear not to be elongated ( Hamilton 1973) as much as they are in Orangemeryx and Sperrgebietomeryx ( Morales et al. 1999) . Only more complete material will permit a clear determination of the systematic position of Prolibytherium , but out of the various families of African early Miocene ruminants, its affinities seem to be closest to Climacoceratidae .

BIOCHRONOLOGY OF WADI MOGHARA

Fourtau (1918 [1920]) correlated Wadi Moghara with the European Burdigalian, in particular the site of Eggenburg (MN3) and the Sables de l’Orléanais (MN3 to MN5). Pickford (1991a) placed it in Faunal Set PII, slightly younger than Rusinga and older than Gebel Zelten and considered that it correlated with MN3 of Europe (Fig. 3). Miller & Simons (1996) reached a similar conclusion that “Moghara and Zelten overlap in time at about 17 My., but that Moghara is slightly older than Zelten”.

Fourtau (1918 [1920]) mentioned that the fossiliferous zone extends from Wadi Faregh in the East to 30 km West of Wadi Moghara a distance of some 130 km. Miller & Simons (1996) report- ed that the fossils they studied came from approximately 40 fossiliferous localities spanning a maximum distance of 55 km. Miller & Simons (1996) did not provide details of the stratigraphic interval covered by the fossil samples but Fourtau (1918 [1920]) measured a stratigraphic section of the Moghara area over 200 metres thick, but did not indicate from which beds the fossil vertebrates were derived. It is likely therefore that at Wadi Moghara, as at Gebel Zelten, a significant span of geological time is represented in the fossil samples ( Pickford 1991b).

Prolibytherium magnieri described here is so similar to the type and other material from Gebel Zelten ( Arambourg 1961; Hamilton 1973), that there can be little doubt that the same time interval occurs in both places. Furthermore, study of the proboscideans from Gebel Zelten housed in the Natural History Museum, London, reveals that the Gebel Zelten succession spans a considerable period of time, and that it probably even has late Oligocene strata at its base, and late middle Miocene strata near the top. The main fossil levels at Gebel Zelten contain anthracotheres, proboscideans, sanitheres and ruminants that are so close to those of Wadi Moghara that there can remain little doubt that much of the two successions was laid down at the same time. Miller & Simons (1996) discounted the presence of Brachyodus at Gebel Zelten because only one metapodial had been recorded from the site ( Pickford 1991b). In the NHM collections, however, there are several more undescribed specimens (two distal tibia, a proximal tibia, three tali, a calcaneum, two complete metapodials), and recent work at the site has yielded much material (Dolores Soria pers. comm.). The Brachyodus from Gebel Zelten is similar in size to Brachyodus depereti (Fourteau, 1918) , and is larger than Brachyodus mogharensis Pickford, 1991 and Afromeryx africanus (Andrews, 1899) , both of which occur at Wadi Moghara. A suid talus from Wadi Moghara is the same size (pers. obs.) as those of Libycochoerus massai Arambourg, 1961 from Gebel Zelten ( Van der Made 1996), suggesting that this species occurs at Wadi Moghara, even though no dental remains have been found to confirm its presence there. The overall picture that emerges is that parts of the Gebel Zelten and Wadi Moghara sequences span the same period of time (Fig. 3). In both areas, it is clear that the old collections consist of mixed assemblages from various levels, but the bulk of the samples come from a relatively restricted period which best equates with MN 4 in Europe.

Figures

FIG. 1. — Partial brain case and frontal appendages of Prolibytherium magnieri Arambourg, 1961 (CGM 82975) probably from Wadi Moghara (early middle Miocene = MN4), Egypt; A, ventral view; B, dorsal view. Scale bar: 5 cm.

FIG. 2. — Geographic distribution of Climacoceratidae.