The development of the Ethiopian biogeographic realm since the late Miocene is here explored with the presentation and review of fossil evidence from eastern Africa. Prostrepsiceros cf. vinayaki and an unknown species of possible caprin affinity are described from the hominid-bearing Asa Koma and Kuseralee Members (~5.7 and ~5.2 Ma) of the Middle Awash, Ethiopia. The Middle Awash Prostrepsiceros cf. vinayaki constitutes the first record of this taxon from Africa, previously known from the Siwaliks and Arabia. The possible caprin joins a number of isolated records of caprin or caprin-like taxa recorded, but poorly understood, from the late Neogene of Africa. The identification of these two taxa from the Middle Awash prompts an overdue review of fossil bovids from the sub-Saharan African record that demonstrate Eurasian affinities, including the reduncin Kobus porrecticornis, and species of Tragoportax. The fossil bovid record provides evidence for greater biological continuity between Africa and Eurasia in the late Miocene and earliest Pliocene than is found later in time. In contrast, the early Pliocene (after 5 Ma) saw the loss of any significant proportions of Eurasian-related taxa, and the continental dominance of African-endemic taxa and lineages, a pattern that continues today.
Citation: Bibi F (2011) Mio-Pliocene Faunal Exchanges and African Biogeography: The Record of Fossil Bovids. PLoS ONE 6(2): e16688. doi:10.1371/journal.pone.0016688
Editor: Peter Ungar, University of Arkansas, United States of America
Received: September 22, 2010; Accepted: December 21, 2010; Published: February 16, 2011
Copyright: © 2011 Faysal Bibi. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by a National Science Foundation Graduate Research Fellowship, a National Science Foundation International Research Fellowship Award (#0852975), the Institute International de Paléoprimatologie et Paléontologie Humaine (I.P.H.E.P.), the Agence Nationale de la Recherche (ANR-09-BLAN-0238), the Revealing Hominid Origins Initiative (R.H.O.I.), a Geological Society of America Ross Research Award, a Yale University Enders Grant, a Yale Institute for Biospheric Studies Center for Field Ecology grant, a Yale ECOSAVE Center grant, the Abu Dhabi Authority for Culture and Heritage, and the Rimo Ramses Nakhle Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The author has declared that no competing interests exist.
Wallace , following Sclater , classified the majority of Africa and Arabia into a single ‘Ethiopian’ biogeographic realm, extending from the Tropic of Cancer southwards to the Cape and Madagascar. Wallace was struck by both the high number of animal groups endemic to this area as well as the absence from it of many widespread Eurasian taxa. He wrote (p.253):
"The great speciality indicated by [the Ethiopian realm's] numerous peculiar families and genera, is still farther increased by the absence of certain groups dominant in the Old-World continent, an absence which we can only account for by the persistence, through long epochs, of barriers isolating the greater part of Africa from the rest of the world.”
More than 130 years on, the biogeographic scheme of Sclater and Wallace continues to form a basis for continental-scale geographic comparison of mammalian communities (Fig. 1). Any observer of modern Africa can quickly recognize the stark ecological boundary delimited by the Sahara Desert, with the vast diversity of African-endemic taxa restricted to regions to its south. With almost no African fossil record to consult, scientists of the 19th and early 20th centuries could only speculate on the age or historical development of this continent's biogeography. In contrast, the last 100 years of paleontological exploration have provided a wealth of information that allows for an investigation into the developmental history of African endemism as a whole, and the Ethiopian biogeographic realm in particular. Wallace's proposal of “long epochs” of isolating barriers can now be more precisely formulated and addressed.
Figure 1. Map showing the boundaries of modern day biogeographic realms and the locations of sites mentioned in the text.
1, Middle Awash; 2, Lothagam, Lukeino, Mpesida, Namurungule; 3, Langebaanweg; 4,Toros-Menalla; 5, Sahabi; 6, Baynunah; 7, Siwaliks.doi:10.1371/journal.pone.0016688.g001
I here investigate the development of African and Ethiopian realm endemism, relying primarily on the fossil record of Bovidae (antelopes, oxen, and kin). I begin by describing three fossil bovid specimens from the faunas of the Asa Koma Member of the Adu-Asa Formation (ASKM) and Kuseralee Member of the Sagantole Formation (KUSM) of the Middle Awash, Ethiopia, dated to 5.77–5.54 Ma and ~5.2 Ma, respectively . These specimens represent two new additions to the faunal lists from those assemblages, which include the hominid Ardipithecus kadabba . I follow with a review of the development of African biogeographic endemism over the last 8 Ma as informed by the presence of bovids of Eurasian affinities in the sub-Saharan African fossil record, outlining the history of the Ethiopian biogeographic realm since the late Miocene.
Systematic Paleontology Antilopini Gray 1821, Prostrepsiceros Major 1891, Prostrepsiceros cf. vinayaki (Pilgrim 1939)
A bovid of medium to small size characterized by horn cores that arise from above the orbit with moderate inclination and basal divergence, torsion that is anticlockwise in the right horn core and relatively helical, very strong mediolateral compression with an oval basal cross-section and flat antero-medial surface, a prominent anterior keel that originates anteromedially, and a posterior keel that is variable in expression. Shallow postcornual fossa present, supraorbital foramina small, pear-shaped, single or multiple.
GSI B799, left horn core from the locality of Nila, Dhok Pathan Formation, Siwaliks (: pl. I fig. 10).
ASK-VP-3/4, proximal portions of right & left horn cores. ALA-VP-2/31, base of right horn core (Fig. 2).
Figure 2. GSI B799, holotype left horn core of Prostrepsiceros vinayaki from the Dhok Pathan Formation, Siwaliks (reproduced from ).
AUH 441, right horn core of P. aff. vinayaki from the Baynunah Formation (U.A.E.) in anterior, lateral, and medial views. ALA-VP-2/31, right horn core of P. cf. vinayaki in anterior, lateral, and medial views. ASK-VP-3/4, right and left horn core fragments of P. cf. vinayaki, in anterior views (top row), lateral views (middle row), right horn core fragment in posterior view, and view of distal break of left horn core fragment with anterior keel position marked AK. Scale bar equals 10 cm total (approximative for GSI B799).doi:10.1371/journal.pone.0016688.g002
Sites ASK-VP-3 and ALA-VP-2 are situated within the Asa Koma Member of the Adu-Asa Formation, and are constrained to between 5.77 and 5.54 Ma, the respective ages of the LABT tuff and the MA95-7 basaltic lava .
ASK-VP-3/4: right and left horn core fragments: left is a basal portion preserving a bit of the pedicel-horn core boundary, right is from just above the horn core base. Both the anterior and posterior keels are strong, the medial surface is quite flattened and the lateral surface rounded. Transverse and antero-posterior diameters (DTxDAP) at just above the horn core base in the left is 23.7×31.8, and at the proximal break in the right 20.6×29.5 mm. ALA-VP-2/31: proximal right horn core with part of the frontal, somewhat weathered, larger in size than ASK-VP-3/4. Anterior keel is prominent and the posterior keel weak. Medial surface is quite flat and the lateral face rounded such that the basal cross-section is an asymmetric oval. A small foramen is present at the posterolateral side of the horn core-pedicel junction. In medial view, the break through the frontal shows no indication of frontal sinuses. Basal DTxDAP is 26.6×38.5 mm.
The ASKM specimens can be differentiated from Sivoreas eremita by horn cores that are much more mediolaterally-compressed, with cross-section more asymmetrical, and weaker torsion that is wider to the torsion axis (more helical), with significant lateral divergence of the horn cores above the base (torsion is tight and the horn core relatively straight in S. eremita). They differ from Prostrepsiceros libycus in the presence of a strong anterior keel, stronger mediolateral torsion, a medial surface that is flatter than the lateral one, and the absence of grooves running along the anterior and posterolateral faces. In all these characters, the ASKM specimens are a good match for Prostrepsiceros vinayaki.
Prostrepsiceros vinayaki is otherwise represented by only a few horn core specimens. These are the holotype , two other horn cores from the Siwaliks , and a horn core (and three postcranial specimens) from the Baynunah Formation referred to Prostrepsiceros aff. vinayaki . Additional material attributable to Prostrepsiceros vinayaki or P. cf. vinayaki and awaiting description is known from the Siwaliks (A. Gentry and J. Barry pers. comm.), Molayan in Afghanistan , and Marageh, Iran .
The paucity and poor representation of material of P. vinayaki limits comparisons. The two Awash specimens, particularly ALA-VP-2/31, are larger than all three previously known horn cores for which measurements are available (Fig. 3). Gentry  reports the Baynunah horn core to have a flatter latter than medial surface, though I have found the surface just medial to the anterior keel to be the flattest part of the horn core. A difference amongst the known specimens of this taxon concerns the prominence of the posterior keel, which is reported as absent in the holotype, weak in the Baynunah specimen, weak to prominent in the Middle Awash specimens, or even more prominent than the anterior keel in other Siwaliks specimens. Otherwise, the Middle Awash, Siwaliks, and Baynunah horn cores all bear a combination of horn core characters that is unique and characteristic, though more and better material will have to be found and described to confidently determine whether this sample represents one or more species.
Figure 3. Plot of basal horn core anteroposterior and transverse diameters in Prostrepsiceros vinayaki (except in ASK-VP-3/4 where measurements are from just above the base).doi:10.1371/journal.pone.0016688.g003
The assignment of ASK-VP-3/4 and ALA-VP-2/31 to Prostrepsiceros cf. vinayaki constitutes the first record of this taxon in Africa (Fig. 4). Prostrepsiceros vinayaki is recorded in the Siwaliks to span 9.3–7.9 Ma , both the Molayan and the Baynunah faunas are between 8 and 6 Ma , , , , and the new Middle Awash occurrences are between 5.77 and 5.54 Ma . These specimens then span around 4myr, which would be atypical for the duration of a single bovid species. Regardless, the morphological congruence of the Awash, Baynunah, Siwaliks, and Molayan records is here taken to be of phylogenetic (and resulting biogeographic) significance.
Figure 4. Maps plotting records of Prostrepsiceros vinayaki (including P. cf. vinayaki) and Kobus porrecticornis (including K. cf. porrecticornis), highlighting links between southern Asia, Arabia, and East Africa in the late Miocene.doi:10.1371/journal.pone.0016688.g004
Kostopoulos  presented additional information on the cranial morphology of P. vinayaki, presumably from the unpublished Molayan material. In a systematic review, Kostopoulos found the closest affinities to P. vinayaki in the Greco-Anatolian P. vallesiensis, stating that “these two species seem to comprise a morphological continuum” (p889). Recent Siwaliks data  indicate that P. vallesiensis and P. vinayaki have a similar first appearance datum, and the phylogenetic relationships proposed by Kostopoulos  are of interest in that they suggest the most recent common ancestor of these two species existed in the early late Miocene and probably inhabited the Greco-Iranian region.
Other reported occurrences of Prostrepsiceros from Africa come from Oued el Atteuch (Algeria) and Sahabi (Libya). The Oued el Atteuch record  consists of a tooth and a horn core fragment which, to my knowledge, have not been described or figured. The Sahabi Prostrepsiceros libycus  differs markedly from P. vinayaki and has recently been reassigned to Dytikodorcas . ALA-VP-2/31 was previously referred to Aepyceros cf. premelampus by Haile Selassie et al. . The only other ASKM specimen that these authors assigned to Aepyceros is ALA-VP-1/5, which in fact bears a triangular section and pronounced torsion and is better assigned to Tragelaphus moroitu. These reassignments now restrict the occurrence of Aepyceros in the Mio-Pliocene Middle Awash assemblages to the younger Kuseralee deposits, with implications discussed below.
Incertae Sedis cf. Caprini Gray 1821
AMW-VP-1/51, right and left horn cores (Fig. 5).
Figure 5. cf. Caprini.
AMW-VP-1/51, right and left horn cores. Shown in side views (top and bottom), and the better preserved of the two horn cores in anterior view (right) with sketch of basal cross-sectional shape labelled anterior (A) and posterior (P). Scale bar equals 10 cm total.doi:10.1371/journal.pone.0016688.g005
From the Kuseralee Member of the Sagantole Formation. Fossils come from just below the MA92-15 5.18 Ma basalt, so are estimated to be ≥5.2 Ma .
AMW-VP-1/51 consists of two horn cores from a single individual, one of which is almost complete and the second more fragmentary (Fig. 5). Given each horn core's great symmetry, it is not evident which is the left and which the right. Horn cores are extremely compressed mediolaterally, with sides flat, and curve posteriorly scimitar-like, lacking torsion. The anterior and posterior surfaces of the horn core are somewhat rounded and not keeled. The cross-section at the base is oval, lacking signs of basal swellings, with the widest transverse diameter located posteriorly. The more complete of the two horn cores preserves what appears to be the horn core base and an attached sliver of pedicel, the internal surface of which appears smooth-walled, indicating the presence of a simple, unstrutted, frontal sinus that reached the pedicel but did not enter the horn core proper. Basal measurements of the more complete horn core are (DTxDAP) 39.3×73.7 mm; approximate complete length along the anterior surface = 425 mm.
In its extreme horn core compression and simple posterior curvature, AMW-VP-1/51 is differentiated from known fossils of Hippotragini of Alcelaphini from the Mio-Pliocene of Africa , . The living Hippotragus niger often has very strongly compressed horn cores but always retains a rounder horn core base than in AMW-VP-1/51. The taxa most comparable to AMW-VP-1/51 are Skoufotragus ( = Pachytragus) laticeps, known best from late Miocene Samos (Greece) , , Bouria anngettyae from the 1 Ma Bouri Daka Member (Ethiopia) , and Pachytragus sp. from the Namurungule Formation (Kenya) , , . Skoufotragus laticeps and Bouria anngettyae both resemble the Awash specimen in posterior-curving horn cores that are very mediolaterally flattened, with hollowed pedicels. The degree of posterior curvature in AMW-VP-1/51 is more so than in S. laticeps, and more like the condition in B. anngettyae. AMW-VP-1/51 differs from Skoufotragus laticeps in stronger posterior curvature, larger size, greater medio-lateral compression, and in the lack of asymmetry or swelling at the basal horn core. It differs from Pachytragus sp. described from the Namurungule Formation in much the same characters. AMW-VP-1/51 differs from Bouria anngettyae in the constancy of the horn core curvature and cross-section, as in Bouria the horn core widens disproportionately and asymmetrically at the base.
The subfamily Caprinae traditionally comprises Pantholops and three taxonomic groups, Caprini+Rupicaprini+Ovibovini, none of which are likely to be monophyletic , . Hassanin and Douzery  suggested the use of the name Caprini in place of the traditionally-defined Caprinae, a view more in accord with emerging molecular phylogenetic evidence. I follow these authors' nomenclature, pointing out that Caprini here is synonymous to the Caprinae of many other authors.
The assignment of AMW-VP-1/51 to Caprini is not made with any certainty in large part because the identification of caprins in the Miocene fossil record is not straightforward. For example, the assignment to Caprini of both Skoufotragus ( = Pachytragus) laticeps and Bouria anngettyae, the two taxa most comparable to AMW-VP-1/51, is open to consideration. While Gentry ,  argued for the caprin status of Pachytragus, Pilgrim “doubtfully”  (p.73) and Bosscha-Erdbrink  took it to be a hippotragin. Likewise, Bouria anngettyae was identified by Vrba  as a caprin, but Gilbert  noted morphological similarities of this species to the alcelaphin Parmularius angusticornis. Skoufotragus laticeps, Bouria anngettyae, and AMW-VP-1/51 all share strong medio-lateral compression of the horn core and the presence of a smooth-walled frontal sinus extending into the pedicel. Bovids with strong medio-lateral compression of the horn core have often been taken to be caprins, but this character can be present in both hippotragins (e.g. Hippotragus) and alcelaphins (e.g. Parmularius, Damaliscus). Likewise, the smooth-walled pedicel sinus appears more similar to the condition in Hippotragini and Alcelaphini than to the normally strutted pedicel sinuses of Caprini (Capricornis and Naemorhedus may be exceptions ). Confusion in the assignment of Miocene fossils among Caprini, Hippotragini, and Alcelaphini is not surprising given that these three clades are in fact sister taxa . Certain fossil forms would then be expected to show a mosaic of primitive and derived characters that indicates affinity to the greater clade Caprini+Hippotragini+Alcelaphini, but makes confident assignment to any one of these tribes difficult.
AMW-VP-1/53 is recorded from the KUSM, while two postcranial specimens assigned to cf. Caprini are recorded from the ASKM (STD-VP-2/74, ALA-VP-2/281, Middle Awash database, http://middleawash.berkeley.edu). It is not possible to speculate as to whether the ASKM and KUSM specimens represent the same or different species.
I. Revision of bovid faunas from 5.6 and 5.2 Ma in the Middle Awash
At 5.77–5.54 Ma and ~5.2 Ma, the Asa Koma Member (ASKM) and the Kuseralee Member (KUSM) faunas of the Middle Awash, Ethiopia, sample terminal Miocene faunas in eastern Africa . The ASKM and KUSM mammalian faunas show the greatest taxonomic similarity to contemporaneous assemblages from Kenya, Chad, Libya, and the United Arab Emirates, followed by those from Iran, Spain, Greece, and Italy . Assemblages such as the ASKM and KUSM provide a window onto late Miocene biogeographic configurations in Africa, which already had an African-endemic (Ethiopian) character to them but retain a degree of taxonomic continuity with Eurasia, shortly before early Pliocene advent of increased continental endemism (discussed below).
The bovid taxa recorded from the Asa Koma and Kuseralee Members of the Middle Awash are listed in Table 1, updated to reflect the identification of new taxa and revision of specimens presented in this paper. Comparison of the bovid taxa from these two assemblages shows some taxonomic differences between these two assemblages, dated to between 5.77–5.54 Ma and just older than 5.2 Ma, respectively , . About half of the bovid taxa in each member are not found in the other. At minimum, four out of ten earlier ASKM bovid species are not represented in the later KUSM (Tragoportax sp. ‘large’, Ugandax sp., Prostrepsiceros cf. vinayaki, Zephyreduncinus oundagaisus). Seven out of thirteen KUSM bovids are not recorded from the earlier ASKM (Ugandax demissum, Aepyceros cf. premelampus, Gazella sp., three reduncins, and a hippotragin). None of these taxonomic differences are the result of clear examples of phyletic evolution, as found in other parts of the total assemblage comparison by Haile-Selassie et al. , but some may be explained by sampling biases probably related to the local absence of appropriate habitats. Ugandax sp. from the AKSM is not a likely ancestor for U. demissum (reassigned from Simatherium following Gentry ) from the KUSM, and the similarity of the former to related species from the Pliocene ,  suggests its absence from the KUSM is likely an artifact of sampling. Similarly, Hippotragini, Gazella, and Aepyceros are recorded from sites older than the KUSM (e.g. Lower Nawata) , meaning their absence from the ASKM is also likely to be the result of sampling bias. The taxonomic differences that remain significant are summarized by the presence of Zephyreduncinus oundagaisus and Prostrepsiceros cf. vinayaki in the ASKM, and three different reduncins and Ugandax demissum in the KUSM. Small sample sizes (Table 1), however, limit any confident interpretation for the differences. Overall, the bovid fauna does not provide strong evidence for any major evolutionary turnover taking place in the time between the two assemblages . The same record, however, does leave open the possibility of some degree of paleoenvironmental differences between the two members.
Table 1. Revised faunal lists for fossil Bovidae of the Asa Koma (5.77–5.54 Ma) and Kuseralee (~5.2 Ma) members of the Middle Awash.doi:10.1371/journal.pone.0016688.t001
The identification of Prostrepsiceros cf. vinayaki from the ASKM indicates biogeographic linkages with the Baynunah and the Siwaliks, but does not significantly alter Bernor et al. 's  biogeographic analysis of the Middle Awash Asa Koma fauna, which found the greatest overall resemblance to that of the Lothagam Nawata Formation. Rather, Prostrepsiceros cf. vinayaki highlights the presence of Eurasian elements in late Miocene eastern Africa while also demonstrating the relatively restricted nature of Eurasian-African faunal exchanges compared to faunal dispersion within Africa itself.
II. Review of African Fossil Bovids of Eurasian Affinities since 8 Ma by Tribe
Bovids are a widespread and diverse group ideally suited for biogeographic studies. Modern bovid ranges conform almost perfectly to Wallace's biogeographic zonations: though there are some 132 extant bovid species, practically none of these possesses a geographic range that significantly traverses any of the traditionally-defined biogeographic boundaries. Perhaps the only exception might be produced if the three ibex species (Capra ibex, C. sibirica, C. nubiana) were considered collectively , providing a range that covers parts of Africa, southern Europe, and Central Asia. In an attempt to better understand the history of development of the Ethiopian realm, I here review the record of sub-Saharan African fossil bovids with ranges significantly traversing the boundaries of modern biogeographic realms, focusing mainly on the record of the last 8myr.
Besides Prostrepsiceros cf. vinayaki described above, perhaps the only other occurrence of a sub-Saharan fossil antilopin of Eurasian affinity is that of Antilope aff. subtorta from the late Pliocene Member C of the Shungura Formation . Antilope is otherwise represented in the Pleistocene Pinjor Formation of the Siwaliks by Antilope subtorta ,  and today in the Indian subcontinent by the living Antilope cervicapra.
From the ca.9.5 Ma  Namurungule Formation in Kenya, Nakaya  reported ‘Ouzocerus? sp.’ (previously Palaeoreas sp. , ), an antilopin otherwise recorded from Greece  and Tunisia (Thomas in ). An updated listing of the Namurungule fauna, however, omits Ouzocerus .
The new Middle Awash cf. Caprini specimen joins a list of caprin or caprin-like taxa known from isolated occurrences in the sub-Saharan African late Neogene record. These include Pachytragus sp. from the Namurungule Formation , , , Budorcas churcheri from Hadar , Ovibovini indet. from Langebaanweg , Bouria anngettyae and Nitidarcus asfawi from the Middle Awash Pleistocene , Makapania broomi and related species from late Pliocene to Holocene sites of South Africa , , , and numerous records of Caprini gen. et sp. indet. from Turkana Basin sites aged 3–1 Ma , , , . To this list I would also add Brabovus nanincisus from Laetoli, originally assigned to Bovini . Brabovus nanincisus bears characteristics that preclude inclusion in Bovini, or even Bovinae, including horn cores lacking keels and with a prominent raised lip at horn core–pedicel border; a deep postcornual fossa; a rounded braincase that is wider anteriorly, with laterally-facing occipital surfaces, mastoids located fairly far anterior on the skull, and a flexed and ventrally extended basicranium; a lower p2 that is relatively reduced in size and morphology; and a dual infraorbital foramen . Vrba and Gatesy  ruled out Brabovus nanincisus from being a hippotragin. I propose that the presence of the above-mentioned traits rule out the possibility of this species being a bovin, and, along with the strutted frontal sinuses and small central incisor, favor the placement of Brabovus nanincisus in Caprini.
The biogeographic implications of the African caprin fossil record are not immediately evident. Vrba  interpreted the patchy fossil record of Caprini in Africa to represent repeated episodes of faunal immigration into the continent from Eurasia, coincident with episodes of global cooling and the opening of land bridge connections (her “traffic-light” model). This hypothesis might find support in the records of Pachytragus, Budorcas, and Makapania, taxa with demonstrated affinities to Eurasian clades, but less so by the large number of taxonomically indeterminate caprin fossils, or even Bouria, Nitidarcus, and Brabovus, that have not been associated phylogenetically with any Eurasian caprin clades. Without further information, the majority of the sub-Saharan African caprin fossil record might just as well be sampling endemic African caprin lineages. Given an ecological preferences for mountainous terrain, their rarity in rift-axial fluvial fossil deposits would not be a surprise.
Reduncin antelopes are today restricted to sub-Saharan Africa, though fossil reduncins are also recorded from North Africa, the Levant, and the Indian subcontinent (and doubtfully from Iran and Spain ). The fossil record of Reduncini from the Siwaliks is extensive, spanning the late Miocene to the Pleistocene (Dhok Pathan to Pinjor formations) and comprising diverse species , . The reduncin Kobus ( = Dorcadoxa) porrecticornis was first described by Pilgrim  from the Dhok Pathan deposits of the Siwaliks (recorded from 9.3 to 8.0 Ma) . Kobus porrecticornis (or K. cf. porrecticornis) has since been recorded from both the Middle Awash ASKM and the KUSM, and also from Mpesida, Lukeino, Baard's Quarry at Langebaanweg, and the Upper Nawata , , , . These sites all date to between ~6.5 and 5 Ma, including presumably the specimens from Baard's Quarry, which is a mixed assemblage . Other reducins are known from the Pliocene and Pleistocene Tatrot and Pinjor Formations . Though their relationships to contemporaneous African reduncins are not clear, there is some indication that Siwaliks Plio-Pleistocene reduncins evolved from the Dhok Pathan taxa without significant connection with African species , .
Though providing the necessary route between Africa and the Indian subcontinent, no fossil reduncins are recorded from the Arabian Peninsula (Fig. 4), neither from the Baynunah Formation , nor the Pleistocene of Nafud . Despite sample sizes being small, this absence remains intriguing, particularly since reduncins are recorded from relatively similar faunas and paleoenvironments at Toros-Menalla and Sahabi, being particularly abundant at the former site , . The absence of Reduncini among Arabian fossil faunas may reflect the absence of appropriate habitats to sustain these antelopes in the Peninsula for significant periods of geological time. Evidence for arid conditions contemporaneous with the Baynunah river system  suggests that the Arabian Peninsula has been characterized by aridity since at least the late Miocene, with climatic variations continuously acting to shift the availability and distribution of limited freshwater habitats. Assuming they had the ecological preferences of their modern counterparts, reduncins such as Kobus porrecticornis would have required permanent wetlands and watered habitats, and only managed intermittent passage through, but not long-term persistence in, the Arabian Peninsula. Perhaps a modern analog may be sought in the Nile River, which supports several reduncin species along its upper reaches in Uganda and Sudan, but none as it traverses the Egyptian Sahara. In contrast, antilopins such as Prostrepsiceros vinayaki may have been sufficiently adapted to semi-arid conditions to persist and mark their presence in the late Miocene Arabian fossil record.
Tragoportacini , including primarily species of Tragoportax and Miotragocerus, is well represented throughout the late Miocene of Europe and Asia, but is poorly known from Africa. Recent years have seen the documentation of several tragoportacin taxa from late Miocene African sites, and I here note similarities between these and certain Eurasian tragoportacins. Tragoportax abyssinicus was named by Haile-Selassie et al.  from the Kuseralee Member (and might be represented in the ASKM by T. cf. abyssinicus). Similar bovids to T. abyssinicus from the African fossil record are Tragoportax sp. ‘A’ from the Lower and Upper Nawata members at Lothagam , and to a lesser extent Tragoportax sp. ‘B’ from the same deposits. The above-named taxa all share small size and straight, upright horns often marked by a prominent anterior demarcation and anterior-ward recurvature towards the tips. However, further comparisons suggest to me that T. abyssinicus finds its closest morphological match with Miotragocerus monacensis, represented by the syntype calvarium from Oberföhring, near Munich . I have examined a cast of this at the Natural History Museum (London), and the differences between the Awash specimens and T. monacensis come down to a few basic characters. T. abyssinicus differs from M. monacensis in larger size, horn cores that are relatively longer and more quadrangular in cross-section, and in the presence of a raised ridge between the horn cores. Otherwise the two taxa are similar (Fig. 6), and one would probably be justified reassigning Tragoportax abyssinicus to Miotragocerus. I hesitate to do this here, however, given continuing uncertainty on how to diagnose and differentiate Tragoportax and Miotragocerus, and the resulting unclear benefits of such a generic distinction. From the Siwaliks, Sivaceros vedicus is also a close match for T. abyssinicus, particularly the horn core of S. cf. vedicus illustrated by Pilgrim . Likewise, Tragoportax sp. ‘A’ from Lothagam  also also compares well to Sivaceros vedicus.
Figure 6. Type specimens of Tragoportax abyssinicus (AMH-VP-1/1, top) and Miotragocerus monacensis (cast in Natural History Museum, London, bottom) in left lateral and anterior views, showing the relative similarities between these two taxa.
Scale bar equals 10 cm total.doi:10.1371/journal.pone.0016688.g006
Other Tragoportax species are represented in Africa by Tragoportax cyrenaicus (originally Miotragocerus cyrenaicus) from Sahabi , T. aff. cyrenaicus from Lower Nawata, Upper Nawata, and Apak members at Lothagam , Tragoportax sp. ‘large’ from the Asa Koma Member , T. acrae (originally Mesembriportax acrae) from Langebaanweg , and a calvarium referred to Tragoportax sp. from the Namurungule Formation . Tragoportax cyrenaicus is only otherwise recorded from the Baynunah Formation  while T. acrae is known only from its type locality.
The Lothagam Tragoportax aff. cyrenaicus, while close to the Baynunah and Sahabi Tragoportax, also merits comparison to Siwaliks forms (Fig. 7). KNM-LT 23149, a partial frontal with left and right horn cores from the Lower Nawata, finds a good match in AMNH 101260, a left horn core from the Dhok Pathan deposits of the Siwaliks labelled as T. punjabicus (junior synonym of T. rugosifrons  or senior synonym of T. browni and T. curvicornis , ). In AMNH 101260 and KNM-LT 23149, the horn cores lack significant posterior curvature and possess marked torsion of about 90° total, such that the medial surface of the horn core clearly comes to face anteriorly in its distal sections. These characters are absent in other AMNH material of T. rugosifrons, T. browni, and T. curvicornis that I was able to see, and call for further comparison of African and Siwaliks tragoportacins.
Figure 7. Tragoportax aff. cyrenaicus (KNM-LT 23149) in anterior view and AMNH 101260, a left horn core, in anterior and lateral views.
From Lothagam and the Siwaliks, respectively, these two specimens show great morphological similarities and call for further comparison of African and Asian Tragoportacini. Scale bar equals 10 cm total.doi:10.1371/journal.pone.0016688.g007
The high taxonomic and morphological diversity that exists within and among species of Tragoportax and Miotragocerus means it is difficult to interpret with any confidence the similarities mentioned above between certain African and Eurasian tragoportacins. Further work is needed to adequately reconstruct the phylogeny of tragoportacins from different assemblages in Asia, Europe, and Africa. For the moment, it is sufficient to point out that the similarity of African tragoportacins to specific tragoportacins from Europe and southern Asia provides evidence for a greater degree of biotic continuity between Africa and Eurasia during the late Miocene than is found later in time. This echoes the record presented by Prostrepsiceros vinayaki and Kobus porrecticornis. Unlike antilopins and reduncins, tragoportacins experience a global extinction in the earliest Pliocene, vanishing completely and mysteriously from the European, Asian, and African records at around or just after 5 Ma.
Alcelaphini, Hippotragini, “Neotragini.”.
One notable occurrence of bovid taxa of possibly African origin in late Miocene Europe comes from the early late Miocene Grosseto lignites of Tuscany. These include the alcelaphin-like Maremmia haupti, and Tyrrhenotragus gracillimus, a “neotragin” .
Otherwise, Alcelaphini, along with Hippotragini, are restricted for the duration of their histories to Africa and Arabia, with the exception of short-lived dispersals to the Indian subcontinent between around 3 Ma and 2.5 Ma . Though today extinct in North Africa and the Levant, Alcelaphus buselaphus is recorded from late Pleistocene sites in the Levant, and was widespread in North Africa until the early 20th Century , , .
III. Development of the Ethiopian Biogeographic Realm as evidenced by the Bovid Fossil Record
Thomas , , following Gentry , discussed the development of African biogeography through the bovid fossil record of the middle Miocene and the early late Miocene, indicating significant Eurasian influences and the appearance of what he termed a ‘proto-Ethiopian’ phase during this time. The Namurungule fauna, dated to around 9.5 Ma , bears strong Eurasian affinities and can also be included in the early late Miocene proto-Ethiopian phase. I here outline three further phases in the development of modern African biogeography since the latter half of the late Miocene, or from around 7 Ma (Fig. 8). These three phases are also generally recognizable in other aspects of the African large mammal fossil record as well (though the micromammalian record often offers a different view , ). The first two phases are based on the review presented in this work, while the third phase is based mainly on the recent work of Geraads .
Figure 8. A summary of the record of sub-Saharan fossil bovids indicating faunal exchange with Eurasia, and the corresponding history of the Ethiopian biogeographic realm.
The Plio-Pleistocene record of African caprins is spotty, and it is not evident whether the majority of its taxa represent Eurasian immigrants. Abbreviations: A, Antilope aff. subtorta; Bc, Budorcas churcheri; Kp, Kobus porrecticornis; Mb, Makapania broomi; Pv, Prostrepsiceros cf. vinayaki.doi:10.1371/journal.pone.0016688.g008
I. By 7 Ma and up to around 5 Ma: The main faunal characteristics of the Ethiopian biogeographic realm are by this time already developed and recognizable. The Ethiopian realm covers the entirety of Africa and Arabia, with relatively porous boundaries permitting exchange with Eurasian faunas.
During this time all over Africa, fossil bovid tribal makeup is characteristically African in nature, but taxa common to Eurasia, namely the Tragoportacini, are present and diverse. Thomas [40: 251] noted that the “emergence of the true Ethiopian fauna” had occurred by 7 Ma. This is reflected in that, by this time, African bovid faunas were dominated by newly emerged African tribes (Tragelaphini, Alcelaphini, Hippotragini, Aepycerotini) and lineages (of Reduncini and Bovini). However, while not abundant, Tragoportacini remain present and ubiquitous in Africa, a strong reminder of faunal continuity with much of Eurasia. Similarly, the wide ranges of Kobus porrecticornis and Prostrepsiceros vinayaki provide further evidence for cross-continental faunal exchange. Additional records of admixture of African and Eurasian faunas during this time come from Spain , Italy , Libya , and the Arabian Peninsula . Faunal exchange between sub-Saharan Africa, North Africa, and Arabia at this time is fluid , , , , and the balance of taxa unites these regions under a single Ethiopian biogeographic realm. Regional bioprovinciality is however present, as evidenced by the Chado-Libyan biogeographical province that united the Lake Chad Basin to Sirt Basin in Libya , .
II. ~5 Ma to the late Pleistocene: The Ethiopian realm continues to cover the entirety of Africa and Arabia, though barriers to faunal exchange with Eurasia are now significant.
Among bovids at this time, there is an almost total dominance of tribes and lineages of African origin, with only rare occurrences of taxa shared with any Eurasian sites. The extinction of Tragoportacini around 5 Ma leaves African tribes (Tragelaphini, Alcelaphini, Hippotragini, Aepycerotini, Cephalophini) and lineages (of Bovini, Reduncini, and Antilopini) in near total faunal dominance. Sub-Saharan African fossil assemblages such as those of the Haradaso Member of the Sagantole Formation at about 4.9 Ma (personal observations), Lower Aramis Member at 4.4 Ma ,  or Kanapoi at 4.1 Ma  comprise bovid faunas dominated by species and clades of African origin. A similar pattern of post-Miocene endemism has been documented among Carnivora . During the late Pliocene and Pleistocene, any indications among Bovidae of Eurasian influence into sub-Saharan Africa are almost entirely restricted to rare records of Caprini (e.g. Budorcas churcheri, Makapania broomi), and the Shungura Antilope. The short-lived appearance of some alcelaphins and hippotragins in the Indian subcontinent in the late Pliocene provides evidence for dispersals out of Africa. The carnivore fossil record also provides evidence for greater migration out of Africa than into Africa at this time . North African faunas from this time continue to show a dominance of African, rather than Eurasian, taxa , .
III. Late Pleistocene to Recent: The northern limit of the Ethiopian realm shifts south towards its present configuration about the Tropic of Cancer.
In North Africa, the immigration of Palaearctic taxa and the loss of faunal elements in common with sub-Saharan Africa changes the makeup and continental affinities of this region's fauna . This leads to the modern-day classification of North Africa in the Palaearctic realm and the redrawing of the Palaearctic-Ethiopian realm boundary along the Saharo-Arabian desert belt. Records of Eurasian bovids in sub-Saharan Africa are absent or rare. Occasional dispersals of bovids and other large mammals from sub-Saharan African into North Africa and the Levant take place into the late Holocene .
The Ethiopian biogeographic realm appears to have had a distinct history of assembly through the Neogene, reflecting a pattern of ever-increasing isolation of African faunas since the late Miocene. The ‘isolating barriers’ of Wallace, defining the northern boundary of the Ethiopian realm, were in place by 7 Ma, though the geographic location of this boundary, and its permeability to African and Eurasian taxa, has changed over time. The presence of Prostrepsiceros cf. vinayaki and a possible caprin in the ASKM and KUSM, taken with the remainder of the bovid record reviewed above, highlights a greater rate of interchange of faunal elements between Eurasia and Africa in the late Miocene and up to around 5 Ma, than found later in time. The majority of the African Pliocene and Pleistocene record is exceptionally poor in Eurasian elements and an African-endemic fauna dominates. Of interest at this time is what appears to be a greater rate of faunal dispersal out of Africa than into it, documented also among other taxa. Throughout most of the last 7myr, the Ethiopian realm covered the entirety of Africa and Arabia. The distinction of North Africa from Sub-Saharan Africa, and the delineation of the northern limits of the modern Ethiopian realm along the Saharo-Arabian desert belt, would come only in the late Pleistocene, presumably on account of increased sub-tropical aridification.
Materials and Methods
Fossils were studied in collections housed at the National Museum of Ethiopia, National Museums of Kenya, the American Museum of Natural History, the Abu Dhabi Authority for Culture and Heritage, and the Natural History Museum, London. Fossil specimens were measured using digital calipers, angle measure, and metric tape.
Thanks to the Middle Awash Research Project and the National Museum of Ethiopia, National Museums of Kenya, and the American Museum of Natural History. I am grateful also to the following individuals for access to collections, advice, and reviews: B. Asfaw, A. Ademassu, M. Beech, J.-R. Boisserie, J. Barry, J. Galkin, J. Gauthier, A. Gentry, Y. Haile-Selassie, A. Hill, D. Kostopoulos, E. Mbua, E. Vrba, T. White, as well as two anonymous reviewers.
Conceived and designed the experiments: FB. Performed the experiments: FB. Analyzed the data: FB. Wrote the paper: FB.
- 1. Wallace AR (1876) The Geographical Distribution of Animals. New York: Harper & Brothers. 607 p.
- 2. Sclater PL (1858) On the general geographic distribution of the members of the class Aves. Journal of the Proceedings of the Linnean Society of London, Zoology 2: 130–136.
- 3. Haile-Selassie Y, WoldeGabriel G, editors. (2009) Ardipithecus kadabba: Late Miocene Evidence from the Middle Awash, Ethiopia. Berkeley: University of California Press. 641 p.
- 4. Haile-Selassie Y (2001) Late Miocene hominids from the Middle Awash, Ethiopia. Nature 412: 178–181.
- 5. Pilgrim GE (1939) The fossil Bovidae of India. Palaeontologia Indica NS 26: 1–356.
- 6. WoldeGabriel G, Haile SY, Renne PR, Hart WK, Ambrose SH, et al. (2001) Geology and palaeontology of the Late Miocene Middle Awash valley, Afar rift, Ethiopia. Nature 412: 175–178.
- 7. Thomas H (1984) Les Bovidés anté-hipparions des Siwaliks inférieurs (Plateau du Potwar, Pakistan). Mémoires de la Société Géologique de France, Nouvelle Série 145: 1–68.
- 8. Gentry AW (1999) Fossil pecorans from the Baynunah Formation, Emirate of Abu Dhabi, United Arab Emirates. In: Whybrow PJ, Hill A, editors. Fossil Vertebrates of Arabia. New Haven: Yale University Press. pp. 290–316.
- 9. Kostopoulos DS (2009) Contribution to the systematics and phylogeny of Prostrepsiceros vallesiensis Bouvrain, 1982 (Mammalia, Bovidae). Geodiversitas 31: 879–891.
- 10. Kostopoulos DS, Bernor RL Geodiversitas. In press. The Marageh bovids (Mammalia, Artiodactyla): Systematic revision and biostratigraphic-biozoogeographic interpretation.
- 11. Badgley C, Barry JC, Morgan ME, Nelson SV, Behrensmeyer AK, et al. (2008) Ecological changes in Miocene mammalian record show impact of prolonged climatic forcing. Proceedings of the National Academy of Sciences 105: 12145–12149.
- 12. Whybrow PJ, Hill A, editors. (1999) Fossil Vertebrates of Arabia, with Emphasis on the Late Miocene Faunas, Geology, and Palaeoenvironments of the Emirate of Abu Dhabi, United Arab Emirates. New Haven: Yale University Press.
- 13. Bibi F, Shabel AB, Kraatz BP, Stidham TA (2006) New fossil ratite (Aves: Palaeognathae) eggshell discoveries from the Late Miocene Baynunah Formation of the United Arab Emirates, Arabian Peninsula. Palaeontologia Electronica 9: 2A: 1–13.
- 14. Brunet M, Heintz E, Battail B (1984) Molayan (Afghanistan) and the Khaur Siwaliks of Pakistan; an example of biogeographic isolation of late Miocene mammalian faunas. Geologie en Mijnbouw 63: 31–38.
- 15. Sen S (1998) The age of the Molayan mammal locality, Afghanistan. Géobios 31: 385–391.
- 16. Renne P, Morgan LE, WoldeGabriel G, Hart WK, Haile-Selassie Y (2009) Geochronology. In: Haile-Selassie Y, WoldeGabriel G, editors. Ardipithecus kadabba: Late Miocene Evidence from the Middle Awash, Ethiopia. Berkeley: University of California Press. pp. 93–104.
- 17. Thomas H, Bernor R, Jaeger JJ (1982) Origines du peuplement mammalien en Afrique du Nord durant le Miocène terminal. Géobios 15: 283–297.
- 18. Lehmann U, Thomas H (1987) Fossil Bovidae (Mammalia) from the Mio-Pliocene of Sahabi, Libya. In: Boaz NT, El-Arnauti A, Gaziry AW, de Heinzelin J, Boaz DD, editors. Neogene Paleontology and Geology of Sahabi. New York: Alan R. Liss. pp. 323–335.
- 19. Bouvrain G, de Bonis L (2007) Ruminants (Mammalia, Artiodactyla: Tragulidae, Cervidae, Bovidae) des gisements du Miocène supérieur (Turolien) de Dytiko (Grèce). Annales de Paléontologie 93: 121–147.
- 20. Haile-Selassie Y, Vrba ES, Bibi F (2009) Bovidae. In: Haile-Selassie Y, WoldeGabriel G, editors. Ardipithecus kadabba: Late Miocene Evidence from the Middle Awash, Ethiopia. Berkeley: University of California Press. pp. 277–330.
- 21. Geraads D, Blondel C, Likius A, Mackaye HT, Vignaud P, et al. (2008) New Hippotragini (Bovidae, Mammalia) from the late Miocene of Toros-Menalla (Chad). Journal of Vertebrate Paleontology 28: 231–242.
- 22. Gentry AW (1971) The earliest goats and other antelopes from the Samos Hipparion fauna. Bulletin of the British Museum (Natural History) Geology Series 20: 229–296.
- 23. Kostopoulos DS (2009) The late Miocene mammal faunas of the Mytilinii Basin, Samos Island, Greece: New Collection. 14. Bovidae. Beiträge zur Paläeontologie 31: 345–389.
- 24. Vrba ES (1997) New fossils of Alcelaphini and Caprinae (Bovidae; Mammalia) from Awash, Ethiopia, and phylogenetic analysis of Alcelaphini. Palaeontologia Africana 34: 127–198.
- 25. Nakaya H, Pickford M, Yasui K, Nakano Y (1987) Additional large mammalian fauna from the Namurungule Formation, Samburu Hills, northern Kenya. African Study Monographs Suppl 5: 79–129.
- 26. Nakaya H, Pickford M, Nakano Y, Ishida H (1984) The Late Miocene large mammal fauna from the Namurungule Formation, Samburu Hills, northern Kenya. African Study Monographs Suppl 2: 87–131.
- 27. Tsujikawa H (2005) The updated late Miocene large mammal fauna from Samburu Hills, northern Kenya. African Study Monographs Suppl 32: 1–50.
- 28. Hassanin A, Pasquet E, Vigne J-D (1998) Molecular systematics of the subfamily Caprinae (Artiodactyle, Bovidae) as determined from Cytochrome b sequences. Journal of Mammalian Evolution 5: 217–236.
- 29. Ropiquet A, Hassanin A (2005) Molecular phylogeny of caprines (Bovidae, Antilopinae): the question of their origin and diversification during the Miocene. Journal of Zoological Systematics and Evolutionary Research 43: 49–60.
- 30. Hassanin A, Douzery EJP (1999) The tribal radiation of the family Bovidae (Artiodactyla) and the evolution of the mitochondrial cytochrome b gene. Molecular Phylogenetics and Evolution 13: 227–243.
- 31. Gentry AW (2000) Caprinae and Hippotragini (Bovidae, Mammalia) in the Upper Miocene. In: Vrba ES, Schaller G, editors. Antelopes, Deer, and Relatives. New Haven: Yale University Press. pp. 65–83.
- 32. Bosscha Erdbrink DP (1988) Protoryx from three localities east of Maragheh, N.W. Iran. Proceedings Koninklijke Nederlandse Akademie Van Wetenschappen Ser B 91: 101–159.
- 33. Gilbert WH (2008) Bovidae. In: Gilbert WH, Asfaw B, editors. Homo erectus: Pleistocene Evidence from the Middle Awash, Ethiopia. Berkeley: Univ. of California Press. pp. 45–94.
- 34. Farke AA (2010) Evolution and functional morphology of the frontal sinuses in Bovidae (Mammalia: Artiodactyla), and implications for the evolution of cranial pneumaticity. Zoological Journal of the Linnean Society 159: 988–1014.
- 35. Bernor RL, Rook L, Haile-Selassie Y (2009) Paleobiogeography. In: Haile-Selassie Y, WoldeGabriel G, editors. Ardipithecus kadabba: Late Miocene Evidence from the Middle Awash, Ethiopia. Berkeley: University of California Press. pp. 549–563.
- 36. Haile-Selassie Y, White T, Bernor RL, Rook L, Vrba ES (2009) Biochronology, faunal turnover, and evolution. In: Haile-Selassie Y, WoldeGabriel G, editors. Ardipithecus kadabba: Late Miocene Evidence from the Middle Awash, Ethiopia. Berkeley: University of California Press. pp. 565–583.
- 37. Gentry AW (2006) A new bovine (Bovidae, Artiodactyla) from the Hadar Formation, Ethiopia. Transactions of the Royal Society of South Africa 61: 41–50.
- 38. Cooke HBS, Coryndon SC (1970) Pleistocene mammals from the Kaiso Formation and other related deposits in Uganda. Fossil Vertebrates of Africa, Vol 2. London: Academic Press. pp. 107–224.
- 39. Harris JM (2003) Bovidae from the Lothagam succession. In: Leakey MG, Harris JM, editors. Lothagam: The Dawn of Humanity in Eastern Africa. New York: Columbia University Press. pp. 531–579.
- 40. Thomas H (1984) Les Bovidae (Artiodactyla; Mammalia) du Miocène du sous-continent indien, de la péninsule arabique et de l'Afrique: biostratigraphie, biogéographie et écologie. Palaeogeography, Palaeoclimatology, Palaeoecology 45: 251–299.
- 41. Gentry AW (1985) The Bovidae of the Omo Group deposits, Ethiopia (French and American collections). In: Coppens Y, Howell FC, editors. Les faunes Plio-Pléistocènes de la basse Vallée de l'Omo (Ethiopie); I: Perissodactyles-Artiodactyles (Bovidae). Paris, France: CNRS. pp. 119–191.
- 42. Pilgrim GE (1937) Siwalik antelopes and oxen in the American Museum of Natural History. Bulletin of the Ameican Museum of Natural History 72: 729–874.
- 43. Sawada Y, Pickford M, Itaya T, Makinouchi T, Tateishi M, et al. (1998) K-Ar ages of Miocene Hominoidea (Kenyapithecus and Samburupithecus) from Samburu Hills, Northern Kenya. Comptes Rendus de l'Académie des Sciences-Series IIA-Earth and Planetary Science 326: 445–451.
- 44. Nakaya H (1994) Faunal change of Late Miocene Africa and Eurasia: mammalian fauna from the Namurungule Formation, Samburu Hills, northern Kenya. African Study Monographs Suppl 20: 1–112.
- 45. Bouvrain G, De Bonis L (1986) Ouzocerus gracilis n.g., n. sp., Bovidae (Artiodactyla, Mammalia) du Vallésien (Miocène supérieur) de Macédoine (Gréce). Geobios 19: 661–667.
- 46. Gentry AW (1996) A fossil Budorcas (Mammalia, Bovidae) from Africa. In: Stewart KM, Seymour KL, editors. Palaeoecology and Palaeoenvironments of Late Cenozoic Mammals: Univ of Toronto Press. pp. 571–587.
- 47. Gentry AW (1980) Fossil Bovidae (Mammalia) from Langebaanweg, South Africa. Annals of the South African Museum 79: 213–337.
- 48. Wells LH, Cooke HBS (1956) Fossil Bovidae from the Limeworks quarry, Makapansgat, Potgietersrus, Johannesburg, South Africa. Palaeontologia Africana 4: 1–55.
- 49. Brink JS (1999) Preliminary report on a caprine from the Cape mountains, South Africa. Archaeozoologia 10: 11–25.
- 50. Berger L, Lacruz R, De Ruiter D (2002) Revised age estimates of Australopithecus-bearing deposits at Sterkfontein, South Africa. American Journal of Physical Anthropology 119: 192–197.
- 51. Harris JM, Brown FH, Leakey MG (1988) Stratigraphy and paleontology of Pliocene and Pleistocene localities west of Lake Turkana, Kenya. Contributions in Science 399: 1–128.
- 52. Harris JM (1991) Family Bovidae. In: Harris JM, editor. Koobi Fora Research Project Volume III. Oxford: Clarendon Press. pp. 139–320.
- 53. Vrba ES (1995) The fossil record of African antelopes (Mammalia, Bovidae) in relation to human evolution and paleoclimate. In: Vrba Elisabeth S, Denton George H, Partridge Timothy C, Burckle Lloyd H, editors. Paleoclimate and Evolution, with Emphasis on Human Origins. New Haven, CT, United States: Yale University Press. pp. 385–424.
- 54. Gentry AW (1987) Pliocene Bovidae from Laetoli. In: Leakey MD, Harris JM, editors. Laetoli; a Pliocene site in northern Tanzania. Oxford, United Kingdom: Clarendon Press. pp. 378–408.
- 55. Bibi F (2009) Evolution, Systematics, and Paleoecology of Bovinae (Mammalia: Artiodactyla) from the Late Miocene to the Recent [Ph.D. thesis]. New Haven: Yale University. 469 p.
- 56. Vrba ES, Gatesy J (1994) New antelope fossils from Awash, Ethiopia, and phylogenetic analysis of Hippotragini (Bovidae, Mammalia). Palaeontologia Africana 31: 55–72.
- 57. Thomas H (1980) Les bovidés du Miocène supérieur des couches de Mpesida et de la formation de Lukeino (district de Baringo, Kenya). In: Leakey REF, Ogot BA, editors. Proceedings of the 8th Panafrican Congress of Prehistory and Quaternary Studies (Nairobi 1977). pp. 82–91.
- 58. Franz-Odendaal T, Kaiser T, Bernor R (2003) Systematics and dietary evaluation of a fossil equid from South Africa. South African Journal of Science 99: 453–464.
- 59. Gentry AW, Gentry A (1978) Fossil Bovidae (Mammalia) of Olduvai Gorge, Tanzania; Part I. Bulletin of the British Museum Natural History Geology Series 29: 289–446.
- 60. Thomas H, Geraads D, Janjou D, Vaslet D, Memesh A, et al. (1998) First Pleistocene faunas from the Arabian Peninsula: An Nafud desert, Saudi Arabia. Comptes Rendus de l'Académie des Sciences, Série II, Sciences de la Terre et des Planètes 326: 145–152.
- 61. Geraads D, Blondel C, Mackaye HT, Likius A, Vignaud P, et al. (2009) Bovidae (Mammalia) from the Lower Pliocene of Chad. Journal of Vertebrate Paleontology 29: 923–933.
- 62. Bibi F, Hill A, Beech M, Yasin W Late Miocene fossils from the Baynunah Formation, United Arab Emirates: Summary of a decade of new work. In: Flynn LJ, Wang X, Fortelius M, editors. Neogene Terrestrial Mammalian Biostratigraphy and Chronology in Asia: Columbia Univ Press. In press.
- 63. Bibi F, Bukhsianidze M, Gentry AW, Geraads D, Kostopoulos DS, et al. (2009) The fossil record and evolution of Bovidae: State of the field. 12. Palaeontologia Electronica:
- 64. Stromer E (1928) Wirbeltiere im obermiozänen Flinz Münchens. Abhandlungen der bayerischen Akademie der Wissenschaften, mathematisch-naturwissenshaftliche Klasse 32: 1–71.
- 65. Thomas H (1979) Miotragocerus cyrenaicus sp. nov. (Bovidae, Artiodactyla, Mammalia) du Miocene superieur de Sahabi (Libye) et ses rapports avec les autres Miotragocerus. Géobios 12: 267–282.
- 66. Spassov N, Geraads D (2004) Tragoportax Pilgrim, 1937 and Miotragocerus Stromer, 1928 (Mammalia, Bovidae) from the Turolian of Hadjidimovo, Bulgaria, and a revision of the late Miocene Mediterranean Boselaphini. Geodiversitas 26: 339–370.
- 67. Moyà-Solà S (1983) Los Boselaphini (Bovidae Mammalia) del Neogeno de la Península Ibérica. Publicaciones de Geología, Universidad Autònoma de Barcelona 18: 1–236.
- 68. Thomas H (1984) Les origines africaines des Bovidae (Artiodactyla, Mammalia) miocènes des lignites de Grosseto (Toscane, Italie). Bulletin du Muséum National d'Histoire Naturelle, Section C 6: 81–101.
- 69. Cheylan G (1991) Patterns of Pleistocene turnover, current distribution and speciation among Mediterranean mammals. In: Groves RH, Di Castri F, editors. Biogeography of Mediterranean Invasions. Cambridge: Cambridge University Press. pp. 227–262.
- 70. Clutton-Brock J (1970) The fossil fauna from an Upper Pleistocene site in Jordan. Journal of Zoology 162: 19–29.
- 71. Mallon D, Kingswood S, editors. (2001) Antelopes. Part 4: North Africa, the Middle East, and Asia. 261 p. Global Survey and Regional Action Plans: IUCN, Gland, Switzerland.
- 72. Thomas H (1981) Les Bovidés miocènes de la formation de Ngorora du Bassin de Baringo (Rift Valley, Kenya). Proceedings Koninklijke Nederlandse Akademie Van Wetenschappen Ser B 84: 335–410.
- 73. Gentry AW (1970) The Bovidae (Mammalia) of the Fort Ternan fossil fauna. Fossil Vertebrates of Africa, Vol 2. London: Academic Press. pp. 243–323.
- 74. Geraads D (2010) Biogeographic relationships of Pliocene and Pleistocene North-western African mammals. Quaternary International 212: 159–168.
- 75. Pickford M, Morales J (1994) Biostratigraphy and palaeobiogeography of East Africa and the Iberian peninsula. Palaeogeography, Palaeoclimatology, Palaeoecology 112: 297–322.
- 76. Ferretti MP, Rook L, Torre D (2003) Stegotetrabelodon (Proboscidea, Elephantidae) from the late Miocene of Southern Italy. Journal of Vertebrate Paleontology 23: 659–666.
- 77. Bernor Raymond L, Rook L (2008) A current view of As Sahabi large mammal biogeographic relationships. Garyounis Scientific Bulletin Special Issue 5: 283–290.
- 78. Geraads D (1989) Vertébrés fossiles du Miocène supérieur du Djebel Krechem el Artsouma (Tunisie Centrale). Comparaisons biostratigraphiques. Géobios 22: 777–801.
- 79. Lihoreau F, Boisserie JR, Viriot L, Coppens Y, Likius A, et al. (2006) Anthracothere dental anatomy reveals a late Miocene Chado-Libyan bioprovince. Proceedings of the National Academy of Sciences 103: 8763–8767.
- 80. WoldeGabriel G, White TD, Suwa G, Renne P, de HJ, et al. (1994) Ecological and temporal placement of early Pliocene hominids at Aramis, Ethiopia. Nature 371: 330–333.
- 81. White TD, Ambrose SH, Suwa G, Su DF, DeGusta D, et al. (2009) Macrovertebrate Paleontology and the Pliocene Habitat of Ardipithecus ramidus. Science 326: 50–56.
- 82. Harris JM, Leakey MG, Cerling CE, Winkler AJ (2003) Early Pliocene tetrapod remains from Kanapoi, Lake Turkana Basin, Kenya. Contributions in Science 498: 39–114.
- 83. Werdelin L (2003) Carnivora from the Kanapoi hominid site, Turkana Basin, northern Kenya. Contributions in Science 498: 115–132.
- 84. Werdelin L, Peigné S (2010) Carnivora. In: Werdelin L, Sanders WJ, editors. Cenozoic Mammals of Africa. Berkeley: University of California Press. pp. 609–663.
- 85. Geraads D (2006) The late Pliocene locality of Ahl al Oughlam, Morocco: Vertebrate fauna and interpretation. Transactions of the Royal Society of South Africa 61: 97–101.
- 86. Thomas H (1979) La faune Quaternaire d'Algerie. Archeologia 134: 61–71.
- 87. Vrba ES (2006) A possible ancestor of the living waterbuck and lechwes: Kobus basilcookei sp. nov. (Reduncini, Bovidae, Artiodactyla) from the Early Pliocene of the Middle Awash, Ethiopia. Transactions of the Royal Society of South Africa 61: 63–74.