A new Plio-Pleistocene viverrid species is described based on two newly discovered maxillae (KW 10141 and KW 10383) from the recent excavations at the hominin-bearing site of Kromdraai (Gauteng, South Africa). This major site allows us to address the conundrum of Paranthropus and Homo origins in South Africa and presents a highly diverse carnivore spectrum (at least 22 species) including herpestids and viverrids. Civettictis braini nov. sp. is a viverrid species comparable in size to the extant African civet Civettictis civetta (Schreber, 1776). However C. braini nov. sp. differs significantly from the extant species in its dental proportions. Its canine and three premolars (P1–P3) are relatively robust, while its carnassials (P4) and two molars (M1, M2) are extremely reduced. This new species supplements our knowledge on carnivore taxonomic diversity and paleoecology in Southern Africa about 2 millions of years ago.
Une nouvelle espèce de viverridé plio-pléistocène est décrite à partir de deux maxillaires (KW 10141 et KW 10383) récemment découverts sur le site à Homininés de Kromdraai (Gauteng, Afrique du Sud). Ce site d’importance nous permet d’aborder la problématique de l’origine des genres Paranthropus et Homo en Afrique du Sud et présente un spectre de Carnivores très diversifié (au moins 22 espèces), parmi lesquels des Herpestidés et des Vivérridés. Civettictis braini nov. sp. est un viverridé de taille comparable à la civette africaine Civettictis civetta (Schreber, 1776). Cependant, C. braini nov. sp. diffère significativement de l’espèce actuelle par ses proportions dentaires. La canine et les trois prémolaires (P1–P3) sont de forte taille et robustes, tandis que la carnassière (P4) et les deux molaires (M1, M2) sont extrêmement réduites. Cette nouvelle espèce complète notre connaissance de la diversité taxinomique des carnivores et de leur paléoécologie en Afrique australe vers 2 millions d’années.
Kromdraai, Plio-Pleistocene, South Africa, Palaeontology, Carnivora, Viverridae, Civettictis braini nov. sp.Kromdraai, Plio-Pléistocène, Afrique du Sud, Paléontologie, Carnivora, Viverridae, Civettictis braini nov. sp.presentedHandled by Lorenzo RookIntroduction
The “Cradle of Humankind” (Gauteng, South Africa) is an important UNESCO area where numerous palaeontological sites preserve Pliocene, Plio-Pleistocene and Pleistocene deposits (e.g., Sterkfontein, Swartkrans, Kromdraai, Gondolin, Malapa, Rising Star). The hominin-bearing and Oldowan-tool bearing assemblages from this area are fundamental for our understanding of the origin of the genus Homo and its relationships with two of its closely affiliated lineages represented in South Africa, Australopithecus sp. and Paranthropus robustus.
Hominin remains and bone tools are abundant in the “Cradle of Humankind” sites but other faunal remains (e.g., carnivores, bovids or non-human primates) usually represent more than 90% of the Plio-Pleistocene assemblages. Therefore, these faunal remains provide important clues to reconstruct palaeoenvironmental conditions and to assess the age of these deposits (biochronology).
The carnivores from the “Cradle of Humankind” have been largely studied from the 1950s up to the 1980s (e.g. Ewer, 1954, Ewer, 1955, Ficcarelli et al., 1984, Hendey, 1973, Hendey, 1974a, Turner, 1984a, Turner, 1984b and Turner, 1986), but more recent studies are less abundant and focused mainly on either a few sites or on particular groups of large carnivores (e.g., Cooper's D inHartstone-Rose et al., 2010; Drimolen inO’Regan and Menter, 2009; Gladysvale inKuhn et al., 2017 and Mutter et al., 2001; Malapa inKuhn et al., 2011, Kuhn et al., 2016, O’Regan et al., 2013 and Reynolds, 2012).
Among the “Cradle of Humankind” assemblages, Kromdraai (26∘00′41′’S, 27∘44′60′’E) appears as a key site for our knowledge of the palaeoenvironmental conditions prevailing during the Plio-Pleistocene transition. It is an unroofed dolomite cave partially shaped by the erosional surface and filled with fossil-bearing deposits, situated approximately 2 km east of Sterkfontein Caves, on the southern side of the Blaauwbank stream. It yielded the type specimen of Paranthropus robustus, TM 1517, the only partial skeleton of this species known thus far (Broom, 1938a, Broom, 1938b, Broom, 1942 and Broom, 1943), and more than 50 other individuals attributed to either Homo or Paranthropus (Braga and Thackeray, 2003, Braga and Thackeray, 2016 and Braga et al., 2017), some representatives of this latter genus appearing more primitive than their congenerics from Swartkrans (Braga et al., 2013).
The Kromdraai carnivores have been largely studied (e.g., Brain, 1981, Ewer, 1954, Ewer, 1955, Ewer, 1956a, Ewer, 1956b, Ewer, 1956c, Hendey, 1973, Turner, 1984a, Turner, 1984b, Turner, 1986 and Vrba, 1981) and data highlight the high richness and specific diversity of the bone accumulations from this site. Five different families (Felidae, Canidae, Hyaenidae, Herpestidae, Viverridae), 18 genera and a minimum of 22 species have been recognized (for details see Fourvel et al., 2016). The first record of carnivores from the previously called “Kromdraai B” locality (see details in Braga et al., 2017) comes in Hendey (1973) and was excavated by Brain from decalcified deposits in 1955–1956 (Brain, 1958). This list is subsequently used in Brain (1975) and Brain (1981) with some minor amendments. The second stage of publications comes from Vrba's excavations and the carnivore material briefly presented and discussed in Vrba (1981). Subsequently, single specimens in one time studies, but no review on the Kromdraai carnivore fauna has been published. There is now a necessity to take into account the very rich and interesting carnivore material uncovered at Kromdraai since 2014 (mainly from the oldest Member 2 – or Member P – breccias) (see Braga and Thackeray, 2016 and Braga et al., 2017 for more details) because it is absent from the most recent review of the African Carnivora (Werdelin and Peigné, 2010).
New excavations initiated in 2014 indicate that the Kromdraai site extends more than 30 m towards the north. During the 2014/2017 period, the Kromdraai Research Project successive field seasons (Braga and Thackeray, 2016 and Braga et al., 2017) led by JB and more occasionally in 2016 and 2017 by the author, have produced over 4600 faunal remains (including hominins) derived mainly from Member 2 (or Member P), an assemblage highlighting a high species diversity covering a wide faunal spectrum including birds, mammals and reptiles (Braga and Thackeray, 2016 and Fourvel et al., 2016). Within that new sample, 459 specimens are related to carnivores (including coprolites) including 6 families, 22 genera and possibly up to 24 species (Fourvel et al., 2016; Table 1). These newly discovered specimens are curated at the Evolutionary Studies Institute of the University of Witwatersrand in Johannesburg.
Part of this important carnivore sample is related to the small herpestids and viverrids, two groups that are rarely studied in details with the exception of Petter, 1963 and Petter, 1987 for eastern Africa and Hendey (1974b) for southern Africa. Four genera of relatively rare viverrids and herpestids were previously identified (Crossarchus, Atilax at “Kromdraai A”, Herpestes, Viverra at “Kromdraai B”, the latter now related to Mellivora; Gommery et al., 2008) including two species (Crossarchus transvaalensis and Atilax mesotes). The newly discovered (since 2014) Kromdraai herpestids and viverrids sample comprises nearly 40 specimens that represent at least 6 genera, including recently two portions of maxilla (KW 10141 and KW 10383) likely related to the genus Civettictis Pocock, 1915 but significantly different from the extant African civet Civettictis civetta (Schreber, 1776). Consequently, here I describe a new species according to the fossil specimens KW 10141 and KW 10383.
Sample and methods
I describe two portions of maxilla (KW 10141 and KW 10383) from the recent excavations (dir. J. Braga and F. Thackeray; discovered respectively in May and September 2017) and then compare it to extant and extinct species in order to discuss their affinities with some particular species, and to distinguish their specific features. The KW 10141 and KW 10383 specimens are compared to small canids: the black-backed jackal Canis mesomelas Schreber, 1775 and the bat-earded fox Otocyon megalotis (Desmaret, 1822). For comparative purposes, I also use morphological and metrical data from a large sample of skulls representing the African civet C. civetta and various herpestids including the marsh mongoose Atilax paludinosus Cuvier, 1829, the slender mongoose Galerella sanguinea Rüppell, 1836, the yellow mongoose Cynictis penicillata Cuvier, 1829 and the meerkat Suricata suricatta (Schreber, 1776) (all specimens curated at the Ditsong National Museum in Pretoria and the Evolutionary Studies Institute of the University of Witwatersrand in Johannesburg; Table 2). The comparative fossil sample is composed of some South African extinct species including Atilax mesotes (Ewer, 1956c) from “Kromdraai A”, and Civettictis cf. civetta from Cooper's D (O’Regan et al., 2013) as well as published data on Viverra leakeyiPetter, 1963 from Langebaanweg (Hendey, 1974b) and eastern African species, including Mungos dietrichiPetter, 1963, V. leakeyi and Pseudocivetta ingensPetter, 1967 from Olduvai and Laetoli (described by Petter, 1963, Petter, 1967, Petter, 1969 and Petter, 1987).
The measurement protocol follows Von Den Driesch (1976). The measurements that I considered as relevant for this study and the results are expressed in millimetres.
Abbreviations list:
C
Canine
for. infra orb.
Infra-orbital foramen
I
Incisor
L
Length
M
Molar
P
Premolar
R. tth
Raw teeth
W
Width
1 to 4
Tooth rank
Results
Order Carnivora Bowdich, 1821
Sub-order Feliformia Kretzoi, 1945
Family Viverridae Gray, 1821
Sub-family Viverrinae Gray, 1821
Genus Civettictis Pocock, 1915
Civettictis braini nov. sp.
Etymology
In honor of C. K. “Bob” Brain who contributed significantly for our understanding of ancient hominins’ behaviour in the “Cradle of Humankind” (including Kromdraai, Swartkrans and Sterkfontein) and hominin-carnivore interactions according to his invaluable work in the field of taphonomy.
Type specimen
Almost complete maxilla (KW 10141), including the complete left one and posterior part of the right one, as well as the left zygoma. Presence of C, P2, M1 and M2 as well as alveoli of I1 to I3, P1, P3, P4.
Paratype
Fragment of right edentate maxilla (KW 10383). Presence of alveoli of C, P1, P2, P3, P4 and M1.
Type locality
Kromdraai Member 2 (or Member P), Gauteng, South Africa (estimated age, Plio-Pleistocene Transition, early Pleistocene).
Diagnosis
Civettictis braini is a large-sized viverrid comparable in size to the extant African civet C. civetta, but differing from the latter in the extreme reduction of the P4 and the upper molars. Civettictis braini has a square-shaped rostrum and an elongated palate. The dental formula includes 3I, 1C, 4P, 2M with large diastema between the canine, the P1 and the P2. The canine and the P2 are well-developed and robust. The same comment applies to the P1 even though if only its alveolus is preserved. The P3 is double rooted (according to its alveoli). Its posterior root is located between the protocone and the parastyle of the P4.
Description
The specimen KW 10141 was found fragmented (post-depositional fragmentation) and was restored. KW 10141 is a partial rostrum composed of two distinct fragments belonging to the same individual. The first and smaller fragment of KW 10141 is a right maxilla preserving the alveoli of P4. The alveoli of the M1 and the M2 are preserved even though they are partially altered in the M2. The second and larger fragment is the almost complete left maxilla with the zygomatic arch (including a large portion of the zygomatic process of the temporal bone), the right premaxilla with the alveoli of the incisors as well as the anterior part of the right maxilla (up to the canine alveolus) (Fig. 1). The left canine, the left P2, the left protocone of the P4 as well as the M1 and M2 are still present.
KW 10141 belongs to an old individual. In viverrids, the M1 and the P4 erupt before the other raw teeth (Slaughter et al., 1974). This suggests that the DP3 and the DP4 (which are replaced by the M1 and the P4) are expulsed before the eruption of the P2, P3 and M2. KW 10141 has alveoli for four premolars and two molars. This is consistent only with an adult individual. Moreover the permanent canine of KW 10141 is heavily worn.
The edentate paratype KW 10383 shows the same features (the alveoli are preserved for C, P1, P2, P3, P4 and M1) and represents at least a young adult (Fig. 2).
KW 10141 belongs to a viverrid with a long and wide rostrum, with metrics clearly comparable to the extant African civet C. civetta or even more robust (Table 3). KW 10383 is slightly smaller than KW 10141 but reflects the size variation in C. braini. In spite of an approximate measure of the palate length of KW 10141, the rostrum appears long and ends clearly behind the M2. The infra-orbital foramen (preserved on the left maxilla of KW 10141 and on KW 10383) is large and oval-shaped, located below the orbit at the basis of the zygomatic arch. This feature is typical of a viverrid and clearly different from a canid one (see below). The zygomatic arch is high and robust. It is well-developed suggesting an opened orbit and a marked zygomatic muscle attachment.
According to the preserved teeth and the alveoli on both specimens, the upper dentition could be described as follow: 3I, 1C, 4P, 2M. The canine and the P1 are separated by an important diastema (4.3 mm length). A similar diastema is observed between the P1 and the P2 (6.8 mm length) as well as between the P2 and the P3 (3.5 mm length). The location of the alveolus of the posterior root of the P3 between the alveoli of the parastyle and the protocone of the P4 suggests that the P3 flanked the P4. This comment applies on both KW 10141 and KW 10383. No diastema has been observed between the P4 and the molars M1 and M2.
The upper canine has an elongated and transversally compressed crown (its length is almost twice its width). In spite of the heavy wear of this tooth, the crown was relatively high. On the lingual surface of the canine and at its crown basis, there is a light groove forming a gutter that extends all along the root. The imprint of this gutter is observed on the canine alveoli (both left and right).
The P1 is missing. However the morphology of the alveoli allows some measurements on both KW 10141 and KW 10383 (W and L; Table 3). The P1 alveolus is oval-shaped and indicates a tooth with a single robust and large root, like in the P2 (see below).
The P2 is preserved on KW 10141. This tooth is an oval-shaped premolar with a partially fused double root. The tooth wear stage is too heavy to allow any detailed description of the crown. The alveolus of the P2 on KW 10383 attests that this tooth is characterized by a partially fused double root.
The P3 is missing on both KW 10141 and KW 10383. However the alveoli are preserved. The morphology of the alveoli indicates a long tooth (longer than wider) with a double root, one supporting the anterior lobe and another below the posterior lobe. As previously described, the P3 is located against the P4. The posterior lobe of the P3 is placed between the parastyle and the protocone of the P4.
A part of the protocone of the left and right P4 of KW 10141 is preserved. On KW 10383 only the alveoli of the P4 are preserved. According to the P4 alveoli, this tooth seems highly reduced, slightly longer than its width. However, the preserved portion of the P4 highlights the protocone position. This cusp is clearly dissociated from the rest of the tooth. The protocone is developed from the paracone and distally-oriented up to the same level than the parastyle.
The molars M1 and M2 are preserved. However the heavy wear stage does not allow any description of their occlusal surface. The M1 and M2 are globally triangular-shaped teeth. Their widths are almost twice their length. However the molars are extremely reduced compared to the general size of the maxilla.
Comparison
KW 10141 is clearly different from the small canids. The maxilla in the black-backed jackal and the bat-eared fox is very wide at the level of the carnassial teeth and reduces progressively up to the canines and the third incisors. In KW 10141, the palate width between the canines (43.9 mm) is almost similar to the palate width between the P1 (43.7 mm) and the palate width between the P2 (46.1 mm).
In canids (C. mesomelas, V. chama and O. megalotis) the infra-orbital foramen is located above the P3 about 1 cm away from the inferior part of the orbit. In KW 10141 and KW 10383, the infra-orbital foramen is located above the P4 below the orbit.
The P2 of KW 10141 is less sharp and more conical than in the extant small canids.
In canids, the diastema between each raw tooth are shorter, the P2 has two separate roots and the molars are well-developed (two in Canis and Vulpes and three in Otocyon) compared to the Kromdraai specimens.
The palate in KW 10141 ends clearly behind the M2; this feature is typical of viverrids and herpestids. While in canids, the palate ends at the level of the M2.
KW 10141 differs from the herpestids. The water mongoose A. paludinosus has a short rostrum, the P2 is double-rooted, the P3 is wide and robust and the P4 is well-developed. In A. paludinosus the M1 and the M2 are triangular-shaped and are proportionally well-developed compared to the other raw teeth. The relative proportions of the P4 and the molars in C. penicillata, S. suricatta and G. sanguinea attest of their high crushing capacities even if these species are smaller than the Kromdraai specimens.
In the fossil context, the species that could be compared to KW 10141 and KW 10383 are rare: A. mesotes from “Kromdraai A”, V. leakeyi from Langebaanweg and P. ingens Laetoli, and Civettictis cf. civetta from Coopers’ D.
Ewer (1956c) described a new herpestid, Herpestes mesotes, from “Kromdraai A” based on an almost complete skull associated with mandibles as well as some post-cranial remains (specimen labelled KA 86). Firstly related to the genus Herpestes, this species has been recently referred to the genus Atilax (Werdelin and Peigné, 2010). A. mesotes from “Kromdraai A” is almost similar to the water mongoose A. paludinosus (for details see Ewer, 1956c). The morphological features recognized on KA 86 differ significantly from KW 10141 and KW 10383 excluding any affinities between these newly discovered specimens and A. mesotes (Fig. 3).
Petter (1963) described the first African fossil related to the Asian genus Viverra. V. leakeyi has been firstly identified in Laetoli (Tanzania). It is a large-sized viverrid sharing clear affinities with the Asian species (Petter, 1963). This extinct species is uncommon, even if its geographical range is wide. V. leakeyi has been recognized in southern Africa (Langebaanweg inHendey, 1974b) as well as in eastern Africa (Petter, 1963, Petter, 1969 and Petter, 1987), up to Morocco (Geraads, 2008). Few fossil specimens have been attributed to this species and most of them are fragmentary. According to the dental morphology and the partially complete skull from Langebaanweg labelled LQ-L.51590 figured by Werdelin and Peigné (2010), KW 10141 is clearly different from V. leakeyi. In spite of the important diastema between C, P1, P2 and P3, the rostrum in V. leakeyi is transversally compressed and the carnassials and the molars are clearly larger than in KW 10141 and KW 10383. The canine L 12863 described by Hendey (1984) differs from KW 10141 by the presence of light grooves on both lingual and buccal surfaces. The P2 and the P3 are similar in V. leakeyi from Langebaanweg (Hendey, 1974). These premolars are elongated and double-rooted. In KW 10141 and KW 10383, the P2 and the P3 appear different according to their alveoli (partially fused double root in the P2, double root in the P3). The P4 from Langebaanweg (L 12863, L 20253 and L 16224) are significantly larger than KW 10141 and KW 10383 (Hendey, 1974: Table 11, p.77; Fig. 4). The same comment applies for the M1 and the M2. The Langebaanweg specimens are larger than the Kromdraai specimens.
P. ingens has been defined by Petter (1967) in Olduvai Bed I (Tanzania). This species is significantly larger than the extant C. civetta and is characterized by square-shaped molars and bunodont teeth (Petter, 1969). KW10141 and KW 10383 have reduced molars (smaller than in the extant C. civetta) and the M1s are triangular-shaped. Thanks to these features, the Kromdraai specimens could not be referred to P. ingens.
Fossil relatives of C. civetta in palaeontological sites of southern Africa are extremely rare. Actually, C. civetta has been identified in only two early Quaternary bone assemblages from the Cradle of Humankind: Cooper's D (O’Regan et al., 2013) and Kromdraai (Fourvel et al., 2016). The upper carnassial belonging to C. cf. civetta from Cooper's D is too large to be considered as similar to KW 10141 and KW 10383.
Even if there are evidence of morphological and metric differences between the Kromdraai specimens (KW 10141 and KW 10383) and the African civet C. civetta, the Kromdraai specimens share affinities with the viverrids and particularly the genus Civettictis (Fig. 3). The rostrum in C. civetta is relatively elongated and has sub-parallel margins from the canines up to the anterior part of the P3. The infra-orbital foramen in C. civetta is located below the orbit at the basis of the zygomatic arch just above the P3. The zygomatic arch is less high in C. civetta than in KW 10141. The frontal process of the arch is also less developed in the extant species. In C. civetta the palate ends about a centimetre behind the M2 like in the Kromdraai specimens. In C. civetta, there are diastema between the C and the P1, the P1 and the P2 as well as between the P2 and the P3 like in KW 10141 and KW 10383. However each diastema is clearly shorter in the extant species than in the Kromdraai specimens. The P3 location according to the P4 is similar in C. civetta than in KW 10141 and KW 10383. However the P2 in the extant species is a double-rooted tooth while the P3 could have up to three roots. The P4 and the molars are well developed in C. civetta while those teeth are very small in KW 10141 and KW 10383.
According to the metrics, KW 10141 and KW 10383 seem to share more affinities with the genus Civettictis than any other viverrids (Table 3, Fig. 4 and Fig. 5). The width between the I3 as well as between the canines are slightly more important than in C. civetta. The same comment applies to the width between the infra-orbital foramen. The palatal length, the length from the anterior margin of the canine up to the posterior margin of the M2, the raw teeth length as well as the length of the premolars (P1 to P4), are nearly similar to the large-sized C. civetta. However the length of the molars (M1 and M2) is extremely reduced and almost similar to the genus Atilax (A. paludinosus and A. mesotes).
According to the metrical analysis, the teeth are the most distinctive element. The canine is a well-developed tooth, which is significantly larger (longer and wider) than the largest C. civetta as well as the extinct large-sized species V. leakeyi (Fig. 4). The P1 in KW 10141 is a very large tooth which is significantly larger than the African civet C. civetta and the large-sized V. leakeyi. In KW 10383, the P1 is similar the largest C. civetta. The P2 is more reduced than the P1 in KW 10141 and KW 10383. The P2 of both specimens are included in the size variation range of the extant African civet C. civetta even if they are less elongated. The P2 are slightly larger than the P2 belonging to Atilax (both A. mesotes and A. paludinosus). In spite of alveolar measurements, the P3 in KW 10141 and KW 10383 are metrically similar to a small-sized civet (even smaller in the case of KW 10383). The P4 in KW 10141 and KW 10383 are of a very small size. The P4 are significantly smaller than the size variation range observed in the extant African civet C. civetta and nearly similar in size than the carnassials in Atilax. The raw teeth reduction in the Kromdraai specimens becomes intensified with the M1 in which the measurements are similar to the large-sized meerkat S. major as described by Hendey (1974b). Like in the M1, the M2 is also very small. The metrics of the M2 are included in A. mesotes variation range as well as in the meerkat S. suricatta.
Consequently, the identification of KW 10141 and KW 10383 to the new species Civettictis braini can be summarized as follows:
the Kromdraai specimens share affinities with the large-sized viverrids according to the dental formula composed of 3I, 1C, 4P, 2M as well as the rostrum morphology and the palate elongation;
the morphological features and the metrics of KW 10141 and KW 10383 are more similar to the genus Civettictis than any other viverrids;
the identification of the Kromdraai specimens as a new civet species Civettictis braini is based on the metric differences in the teeth proportions (highly reduced P4 and molars) with the African civet C. civetta.
Biochronology
We should report here the unsual wealth of the Plio-Pleistocene site of Kromdraai. Since 2014, more than 400 carnivore bone and dental remains have been collected, including about 8% of small specimens (NISP = 38) attributed to herpestids and viverrids. The new species Civettictis braini presented in this paper has been found within the Member 2 (or Member P) at Kromdraai, in association with other carnivore species that support a Plio-Pleistocene transition/early Pleistocene age, including the false dirk-toothed cat Dinofelis barlowi, the extinct weasel Prepoecilogale bolti as well as the extinct bat-earded fox Prototocyon cf. recki. A detailed palaeontological analysis of the Kromdraai Member 2 carnivore community will be published soon. It will be an important contribution because this Member was previously considered as sterile (see Braga et al., 2017 for more details). The study of the bovid remains collected since 2014 will also be of interest. The co-occurrence of Connochaetes taurinus prognu and cf. Connochaetes gentryi is noteworthy because it is reminiscent of the late Pliocene and early Pleistocene of Kenya, as well as the older levels of Olduvai (Harris et al., 1988, Fourvel et al., 2016, Gentry, 2010 and Harris, 1991). Moreover, the presence of a Makapania-like caprine could relate Kromdraai Member 2 (or Member P) to a Plio-Pleistocene transition/early Pleistocene age, since it is abundant at Makapan Limeworks Members 3 and 4 as well as in the basal levels of Olduvai (Fourvel et al., 2016, Gentry, 2010 and Gentry and Gentry, 1978).
Conclusion
Morphologically close to the genus Civettictis, it is reasonable to assign KW 10141and KW 10383 to this genus. However, other features and metric data (particularly tooth size) differ significantly from the extant civet supporting that the Kromdraai specimens represent a new species, Civettictis braini. According to the relative scarcity of the fossil remains from Plio-Pleistocene viverrids (as well as the scarcity of the palaeontological analysis focused on the Plio-Pleistocene herpestids and viverrids) and the edentate specimens from Kromdraai, it is difficult, or even impossible, to assess the phylogenic relations between C. braini, C. civetta as well as the larger form V. leakeyi and P. ingens. Most detailed researches focused on those small carnivores from the African Quaternary have been mainly published by Petter, 1963, Petter, 1969 and Petter, 1987 in eastern Africa and Hendey (1974b) and Ewer (1956c) in southern Africa. Even if there are some recent studies on carnivore remains from the “Cradle of Humankind”, including herpestids and viverrids (Fourvel et al., 2016, Kuhn et al., 2011 and O’Regan et al., 2013), these publications are not useful to assess the phylogenic relations and discuss the evolution of these small carnivores.
It is also important to emphasize the scarcity of the palaeontological samples related to those small species. The small quantity of bone sample is not necessarily a reflection of the scarcity of the species during their living time. This could be the result of a bias due to taphonomical processes (e.g., biological or non-biological processes) or related to the excavation and the bone collecting methods. It seems more reasonable to suggest that the lack of specimen may be related to the excavations methods (e.g., wet and dry sieving of the excavated sediment allowing us to collect more small remains) as well as the difficulties to identify these small bone remains or teeth. The field works currently conducted in the Plio-Pleistocene sites in South Africa (like at Kromdraai) using modern excavations methods (specimen collecting methods, systematic sieving of all sediments, detailed examination of each bone remains) should fill partly this lack of small species.
The identification of a new species such as Civettictis braini is important to emphasize and will help to further describe the palaeoenvironmental conditions that prevailed when Homo and Paranthropus evolved and coexisted in South Africa.
Acknowledgments
This research is supported by the French Ministry of Foreign Affairs, the Institut des Déserts et des Steppes (Paris) and the Institut Picot de Lapeyrouse (Toulouse). The author was supported by the Erasmus Mundus AESOP+ program (Application ES15PD0043). I thank the Cradle Management Authority, the South African Heritage Ressources Agency and M Ryan Lotz for their continuous support. I also thank Stephany Potze, Lazarus Kgasi, Shaw Badenhorst, Wynand Van Zyl (Ditsong Museum, Pretoria) and Bernhard Zipfel (Evolutionary Studies Institute, University of the Witwatersrand, Johannesburg) who gave me access to the modern and fossil samples under their care. I also thank Raphaël Hautefort and Benjamin Lans for their help with the photographs presented in this study. I am grateful to Jean-Philip Brugal (CNRS-UMR7269 LAMPEA) for his help providing data of the extant African civet. I would like to thank A. Hartstone-Rose and an anonymous reviewer for their helpful comments, which improved this paper.
BragaJ.ThackerayJ.F.Early Homo at Kromdraai B: probalistic and morphological analysis of the lower dentitionC. R. Palevol242003269–27910.1016/S1631-0683(03)00044-7BragaJ.ThackerayJ.F.Kromdraai, a Birthplace of Paranthropus in the Cradle of Humankind2016Sun Media Metro113 pp.BragaJ.ThackerayJ.F.DumoncelJ.DescouensD.BruxellesL.LoubesJ.-M.KahnJ.-L.StampanoniM.BamL.HoffmanJ.de BeerF.SpoorF.A new partial temporal bone of a juvenile hominin from the site of Kromdraai B (South Africa)J. Hum. Evol.652013447–45610.1016/j.jhevol.2013.07.013BragaJ.ThackerayJ.F.BruxellesL.DumoncelJ.FourvelJ.-B.Stretching the time span of hominin evolution at Kromdraai (Gauteng, South Africa): recent discoveriesC.R. Palevol161201758–7010.1016/j.crpv.2016.03.003BrainC.K.The Transvaal Ape-Man-bearing cave depositsTransvaal Museum Memoir 111958Transvaal MuseumPretoria131BrainC.K.An interpretation of the Bone assemblage from the Kromdraai Australopithecine site, South AfricaTuttleR.H.Paleoanthropology, morphology and paleoecology1975MoutonParis225–243BrainC.K.The hunters or the hunted?An introduction to African Cave Taphonomy1981The University of Chicago PressChicago376BroomR.The Pleistocene anthropoid apes of South AfricaNature1421938377–379BroomR.Further evidence on the structure of the South African Pleistocene AnthropoidsNature1421938897–899BroomR.The hand of the ape-man, Paranthropus robustusNature1491942513–514BroomR.An ankle-bone of the Ape-man, Paranthropus robustusNature1521943689–690EwerR.F.The fossil carnivores of the Transvaal caves. The Hyaenidae of KromdraaiProc. Zool. Soc. London1241954565–585EwerR.F.The fossil carnivores of the Transvaal caves: MachairodontinaeProc. Zool. Soc. London1251955587–615EwerR.F.The fossil carnivores of the Transvaal caves: CanidaeProc. Zool. Soc. London126195697–119EwerR.F.The fossil carnivores of the Transvaal caves: FelinaeProc. Zool. Soc. London126195683–95EwerR.F.The fossil carnivores of the Transvaal caves: two new viverrids, together with some general considerationsProc. Zool. Soc. London1261956259–274FiccarelliG.TorreD.TurnerA.First evidence for a species of raccoon dog, Nyctereutes Temminck, 1838, in South African Plio-Pleistocene depositsEstr. boll. Soc. Paleont. Italiana2311984125–130FourvelJ.-B.BrinkJ.O’ReganH.BeaudetA.PaviaM.Some preliminary interpretations of the oldest faunal assemblage from KromdraaiBragaJ.ThackerayJ.F.Kromdraai, a birthplace of Paranthropus in the Cradle of Humankind2016Sun Media Metro71–104GentryA.W.BovidaeWerdelinL.SaundersW.J.Cenozoic mammals of Africa2010University of California PressBerkeley741–796GentryA.W.GentryA.Fossil Bovidae (Mammalia) of Olduvai Gorge, TanzaniaPart I. Bull. British Mus. (Natural History)291978290–446GeraadsD.Plio-Pleistocene Carnivora of northwestern Africa: a short reviewC. R. Palevol782008591–59910.1016/j.crpv.2008.09.008GommeryD.ThackerayJ.F.PotzeS.BragaJ.The first recorded occurence of hiney badger of the genus Mellivora (Carnivora: Mustelidae) at Kromdraai BSouth Africa. Ann. Transv. Mus.4512008145–148HarrisJ.M.Family BovidaeHarrisJ.M.Koobi Fora Research Project31991Clarendon PressOxford139–320HarrisJ.M.BrownF.H.LeakeyM.G.Stratigraphy and Paleontology of Pliocene and Pleistocene localities West of Lake Turkana, Kenya3991988Natural History Museum of Los Angeles County1–128(Contributions in Science)Hartstone-RoseA.WerdelinL.De RuiterD.J.BergerL.R.ChurchillS.E.The Plio-Pleistocene ancestor of wild dogs, Lycaon sekowei n. sp.J. Paleont.8422010299–30810.1666/09-124.1HendeyQ.B.Carnivore remains from the Kromdraai Australopithecine site (Mammalia: Carnivora)Ann. Transvaal Museum288197399–112HendeyQ.B.New fossil carnivores from the Swartkrans australopithecine site (Mammalia: Carnivora)Ann. Transvaal Museum29197427–47HendeyQ.B.The Late Cenozoic Carnivora of the South-Western Cape ProvinceAnn. S. Afr. Mus.6319741–369KuhnB.F.WerdelinL.Hartstone-RoseA.LacruzR.S.BergerL.R.Carnivoran Remains from the Malapa Hominin Site, South AfricaPLoS ONE6112011e2694010.1371/journal.pone.0026940KuhnB.F.Hartstone-RoseA.LacruzR.S.HerriesA.I.R.WerdelinL.BamfordM.K.BergerL.R.The carnivore guild circa 1.98 million years: biodiversity and implications for the palaeoenvironment at Malapa, South AfricaPalaeobio. Palaeoenvir.9642016611–61610.1007/s12549-016-0245-0KuhnB.F.WerdelinL.SteiningerC.Fossil Hyaenidae from Cooper's Cave, South Africa, and the palaeoenvironmental implicationsPalaeobio. Palaeoenvir.9722017355–36510.1007/s12549-016-0247-yMutterR.BergerL.R.SchmidP.New evidence of the Giant Hyaena, Pachycrocuta brevirostris (Carnivora, Hyaenidae), from the Gladysvale Cave deposit (Plio-pleistocene, John Nash Nature Reserve, Gauteng, South Africa)Palaeont. Afr.372001103–113O’ReganH.J.MenterC.G.Carnivora from the Plio-Pleistocene hominin site of Drimolen, Gauteng, South AfricaGeobios422009329–35010.1016/j.geobios.2009.03.001O’ReganH.J.CohenB.F.SteiningerC.M.Mustelid and viverrid remains from the Pleistocene site of Cooper's D, Gauteng, South AfricaPalaeont. Afr.48201319–23PetterG.Étude de quelques Viverridés (Mammifères, Carnivores) du Pléistocène inférieur du Tanganyika (Afrique orientale). C. R. sommBull. Soc. geol. France751963265–274PetterG.Petits carnivores villafranchiens du Bed I d’Oldoway (Tanzanie), Problèmes actuels de PaléontologieEvol. Vert.1631967529–538PetterG.Interprétation évolutive des caractères de la denture des Viverridés africainsMammalia3341969607–625PetterG.Small carnivores (Viverridae, Mustelidae, Canidae) from LaetoliLeakeyM.D.HarrisJ.M.Laetoli, a Pliocene site in northern Tanzania1987Clarendon PressOxfordUK, 194–243ReynoldsS.C.Nyctereutes terblanchei: the raccoon dog that never wasSouth Afr. J. S.1081/220121–1010.4102/sajs.v108i1/2.589SlaughterB.H.PineR.H.PineN.E.Eruption of Cheek Teeth in Insectivora and CarnivoraJ. Mammal.5511974115–125TurnerA.Panthera crassidens Broom. The Cat That Never Was?South Afr. J. S.801984227–233TurnerA.The interpretation of variation in fossil specimens of spotted hyaena (Crocuta crocuta Erxleben, 1777) from Sterkfontein Valley Sites (Mammalia: Carnivora)Ann. Transvaal Museum331984399–418TurnerA.Miscellaneous carnivore remains from Plio-Pleistocene deposits in the Sterkfontein Valley (Mammalia: Carnivora)Ann. Transvaal Museum341986203–226Von Den DrieschA.A Guide to the Measurement of Animal Bones from Archaeoloigcal Sites1976Peabody Museum BulletinsBulletin 1, Harvard University137 p.VrbaE.S.The Kromdraai australopithecine site revisited in 1980; recent investigations and resultsAnn. Transvaal Museum33198117–60WerdelinL.PeignéS.CarnivoraWerdelinL.SandersW.J.Cenozoic Mammals of Africa2010University of California Press603–657
Civettictis braini nov. sp. (KW 10141) maxilla. A1, Palate view; A2, palate view of right maxilla (note that the M1 and M2 alveoli are partially altered); B, upper view; C, lateral view. Scale: 50 mm.
Maxillaire (KW 10141) de Civettictis braini nov. sp. A1, Vue palatine ; A2, vue palatine du fragment de maxillaire droit (notons que les alvéoles des M1 et M2 sont partiellement altérés) ; B, vue supérieure ; C, vue latérale. Échelle : 50 mm.
Photo. R. Hautefort.
Civettictis braini nov. sp. (KW 10383) maxilla. A, Palate view; B, upper view; C, lateral view. Scale: 50 mm.
Maxillaire (KW 10383) de Civettictis braini nov. sp. A. Vue palatine ; B, vue supérieure ; C, vue latérale. Échelle : 50 mm.
Photo. B. Lans.
Comparative plate of viverrid and herpestid skulls, upper and lower view. A, Civettictis braini nov. sp. (KW 10141); B, Civettictis civetta (AZ 333); C, Atilax paludinosus (AZ 1025); D, Atilax mesotes (KA 86). Scale: 50 mm.
Comparaison de quelques crânes de viverridés et d’herpestidés. Vues supérieures et inférieures. A, Civettictis braini nov. sp. (KW 10141) ; B, Civettictis civetta (AZ 333) ; C, Atilax paludinosus (AZ 1025) ; D, Atilax mesotes (KA 86). Échelle : 50 mm.
Photo. J.-B. Fourvel et R. Hautefort.
Bivariate plot (width/length ratio) of the raw teeth (C, P1, P2, P3, P4, M1 and M2).
Diagramme bivarié (rapport longueur/largeur) des dents supérieures (C, P1, P2, P3, P4, M1 et M2).
Herpestid and viverrid comparative datasets, labels, curation and references.
Liste des échantillons de comparaison d’herpestidés et de viverridés. Numéro d’inventaire, Institution et Référence.
Table 2SpeciesLabelSiteCurationReferenceCivettictis civettaAZ 731, AZ 333, AZ 2441, AZ 1164, AZ 1729, AZ 2442, AZ 599, AZ 1190, AZ 730(Extant)Ditsong MuseumFourvel, unpublishedAtilax paludinosusAZ 1025, AZ 1938(Extant)Ditsong MuseumFourvel, unpublishedSuricata suricattaAZ 2214, AZ 614, AZ 2285(Extant)Ditsong MuseumFourvel, unpublishedSuricata major8669, 6873, 6893, 16558ElandsfonteinHendey, 1974bCynictis penicilataBPI/C 312, BPI/C 310, BPI/C 314, AZ 1961, AZ 2342, AZ 527,(Extant)Ditsong Museum / ESI WitsFourvel, unpublishedGaleralla sanguineaAZ 743, AZ 1960, AZ 742, AZ 744, AZ 408, AZ 3323, AZ 2286, AZ 745(Extant)Ditsong MuseumFourvel, unpublishedAtilax mesotesKA 86Kromdraai ADitsong MuseumEwer, 1956c; Fourvel, unpublishedHerpestes ichneumon(Extant)Petter, 1963 and Petter, 1987Mungos mungos(Extant)Petter, 1963 and Petter, 1987Mungos dietrichiL.3741, no numberLaetoli / OlduvaiPetter, 1963 and Petter, 1987Pseudocivetta ingensno numberOlduvaiPetter, 1967Viverra leakeyiL. 12836, L. 13907, L.20253, L.16240A, L.16224, L.62240 C, L.15174LangebaanwegHendey, 1974bLIT 59/466LaetoliPetter, 1987
Measurements of KW 10141 and KW 10383, maxilla fragments from Kromdraai Member 2 as well as comparative datasets (minimum–maximum, mean) of extant species C. civetta (n = 9), A. paludinosus (n = 2), S. suricatta (n = 3), C. penicillata (n = 6) and G. sanguinea (n = 8).
Mesures des portions de maxillaire KW 10141 et KW 10383 de Kromdraai membre 2 et données de comparaison (minimum–maximum, moyenne) d’espèces actuelles : C Civettictis civetta (n = 9), Atilax paludinosus (n = 2), Suricata suricatta (n = 3), Cynictis penicillata (n = 6) et Gallerela sanguinea (n = 8).
Table 3Civettictis brainiCivettictis civettaAtilax paludinosusSuricata suricattaCynictis penicillataGalerella sanguineaKW 10141KW 10383Min–maxMeanMin–maxMeanMin–maxMeanMin–maxMeanMin–maxMeanW I3–I326.618.9–21.720.113.1–13.613.48.2–9.48.66.6–8.87.86.7–8.07.1W. C–C43.927.9–32.830.125.6–26.426.013.3–16.314.913.5–16.214.410.4–14.812.0W for infra orb40.233.0–36.734.926.1–26.526.313.2–16.815.415.6–18.216.812.7–17.713.9L Cant–M263.755.1–61.959.038.1–38.838.522.2–24.423.023.3–26.725.420.8–25.422.8L raw tth48.147.5–52.049.729.2–31.129.916.5–19.117.419.9–22.121.117.0–21.018.7L P38.234.334.8–38.537.022.8–25.423.911.0–12.311.815.8–17.917.113.7–17.015.6L M8.013.8–18.315.89.3–11.510.26.0–6.86.34.9–6.55.73.6–5.54.6L Palate98.0a75.7–87.181.554.4–59.156.934.7–37.436.333.1–39.335.928.2–34.932.1C W6.95.0–5.55.24.7–5.24.92.1–2.82.62.3–2.82.51.6–2.62.0C L11.96.2–7.56.66.3–6.96.63.5–4.03.73.1–4.33.62.5–3.73.0P1W4.4b3.2b2.8–3.53.1N/AN/AN/AN/A1.3–1.61.51.1–1.61.3P1 L6.0b4.8b3.8–4.84.2N/AN/AN/AN/A1.3–1.61.41.3–2.11.6P2W4.53.6b4.1–4.84.43.7–4.44.01.7–2.32.01.9–2.52.11.6–2.31.8P2 L5.96.1b6.8–8.27.53.7–5.85.12.8–3.53.23.8–5.04.13.2–4.23.7P3W4.6b3.5b4.7–6.25.85.8–6.16.03.2–3.43.32.9–4.03.42.2–3.22.7P3 L8.2b7.9b7.4–9.78.57.2–7.37.32.3–4.13.53.9–5.04.33.5–4.43.9P4W7.2b6.3b9.1–11.510.28.0–8.38.24.9–6.55.75.1–5.55.33.7–5.34.4P4 L10.4b8.9b11.6–14.012.99.4–10.39.93.9–4.74.35.4–6.36.05.7–7.06.4M1W7.25.3b11.3–14.413.39.2–11.410.05.7–6.56.05.8–7.36.76.2–7.36.4M1 L4.64.5b8.9–10.89.76.4–8.07.33.4–3.63.53.0–3.83.42.6–3.63.1M2W5.38.7–10.28.9N/AN/A4.1–4.94.54.2–5.34.82.7–4.53.5M2 L2.96.2–6.86.4N/AN/A2.0–2.42.21.8–3.12.41.3–1.81.5