Content uploaded by Alberto Valenciano Vaquero
Author content
All content in this area was uploaded by Alberto Valenciano Vaquero on Jul 20, 2023
Content may be subject to copyright.
iScience
Article
Langebaanweg’s sabertooth guild reveals an
African Pliocene evolutionary hotspot for
sabertooths (Carnivora; Felidae)
Qigao Jiangzuo,
Caitlin Rabe, Juan
Abella, Romala
Govender,
Alberto
Valenciano
albvalen@ucm.es
Highlights
Langebaanweg’s (LBW)
sabertooth cat’s guild
includes two new species
Dinofelis from LBW is the
potential ancestor of Plio-
Pleistocene Dinofelis
We confirm a Pan-African
distribution of
Adeilosmilus and
Lokotunjailurus
Comparison with Eurasian
felid guild implies open
environment at LBW
Jiangzuo et al., iScience --,
107212
--,2023ª2023 The
Author(s).
https://doi.org/10.1016/
j.isci.2023.107212
ll
OPEN ACCESS
iScience
Article
Langebaanweg’s sabertooth guild reveals
an African Pliocene evolutionary hotspot
for sabertooths (Carnivora; Felidae)
Qigao Jiangzuo,
1,2,3,4
Caitlin Rabe,
5
Juan Abella,
6,7,8
Romala Govender,
5,9
and Alberto Valenciano
9,10,11,
*
SUMMARY
Here, we describe and revise craniodental material from Langebaanweg ‘E’
Quarry (South Africa, early Pliocene, 5.2 Ma), which represents one of the
largest and best-preserved collections of sabertooth felids from Mio-Pliocene de-
posits of Africa. Four taxa, including two new species, are recognized: Lokotun-
jailurus chinsamyae sp. nov., Adeilosmilus aff. kabir,Yoshi obscura,andDinofelis
werdelini sp. nov. The felid guild composition analyzed herein suggests the pres-
ence of a mosaic environment with open components in the region, and shows a
potential relationship with that of Yuanmou, suggesting a similar environment
and/or dispersal route/event. The reassessment of the rich early Pliocene felids
from Langebaanweg is a step toward understanding the transition and evolution
of the felids in the southern hemisphere during the late Miocene to early Pliocene.
INTRODUCTION
ThePliocenewitnessedglobalclimatechangetowardacolderenvironment.
1,2
Late Miocene fauna expe-
rienced a significant change in Europe around the Mio-Pliocene boundary, documenting a shift from a
more open savannah-like environment to a more closed forest environment.
3
This change contrasts with
eastern and southern Africa where, during the Late Miocene, the environment changed to more open,
arid grassland or desert, which continued into the Pliocene.
4–9
Located on the west coast of South Africa (Figure 1), Langebaanweg ‘E’ Quarry is one of the most important
fossil communities to document the Mio-Pliocene environmental change and animal evolution in southern
Africa.
10,11
Baard’s Quarry and ‘E’ Quarry comprise part of the Langebaanweg fossil locality and are located
13–15 km inland from Saldanha Bay on South Africa’s west coast (Figure 1).
12
They were commercially ex-
ploited as open-cast phosphate mines from 1943 to 1993. Singer and Hooijer
13
reported the first occur-
rence of Tertiary vertebrate fossils in the area, describing the remains of an elephant0s relative from ‘Baard’s
Quarry’ which was later back-filled.
14
Fossils from Baard’s Quarry are fragmentary and show evidence of
rolling. They were thought to originate from a river lag deposit or reworked from deposits below level 2
that may have been contemporaneous. In 1965, mining started at the Varswater mine, ‘E’ Quarry, 2.5 km
west of Baard’s Quarry, producing the rich and diverse fossils from Langebaanweg.
11,12,15
Fieldwork at
‘E’ Quarry began in 1965 and continues until today.
The fossils from ‘E’ Quarry occur in the Varswater Formation that spans the middle Miocene (Langhian) to
early Pliocene (Zanclean).
10,16,17
All the terrestrial carnivoran fossils from ‘E’ Quarry,
12,18–25
including the
new material described herein, come from the Langeberg Quartz Sand Member (LQSM) and Muishond
Fontein Pelletal Phosphorite Member (MPPM). They are concentrated within an ‘abbreviated’ stratigraphic
interval of ca 26–30 m above sea level
10, p. 208
and are associated with numerous transgressive-regressive
episodes.
10,16,17,26
The LQSM was deposited in a lagoonal and/or estuarine setting protected from wave
action but open to the sea.
10,27,28
The MPPM, Beds 3aN and 3aS, was deposited in a shallow embayment
sheltered from the open ocean by granitic islands during the early Pliocene transgression.
29
Both beds are
inferred as very close in age, with Bed 3aS being slightly older.
11
However, the ages of LQSM and MPPM
were estimated by palaeomagnetic data and global sea level reconstructions as 5.15 G0.1 Ma, suggest-
ing that the fossils accumulated at an early stage in the Early Pliocene transgression when the sea level was
30m above present level.
10
Baard’s Quarry
12,14
is interpreted as a much more recent deposit (late Pliocene
or early Pleistocene).
1
Key Laboratory of Orogenic
Belts and Crustal Evolution,
School of Earth and Space
Sciences, Peking University,
Beijing 100044, China
2
Key Laboratory of
Vertebrate Evolution and
Human Origins of Chinese
Academy of Sciences,
Institute of Vertebrate
Paleontology and
Paleoanthropology, Chinese
Academy of Sciences, Beijing
100044, China
3
CAS Center for Excellence in
Life and Paleoenvironment,
Beijing 100044, China
4
Division of Paleontology,
American Museum of Natural
History, New York City, NY
10024, USA
5
Department of Biological
Sciences, University of Cape
Town, Cape Town 7701,
South Africa
6
Departament de Bota
`nica i
Geologia, Grup
d’Investigacio
´en
Paleontologia de Vertebrats
del Cenozoic (PVC-GIUV),
Universitat de Vale
`ncia,
Burjassot, 46100 Vale
`ncia,
Spain
7
Institut Catala
`de
Paleontologia Miquel
Crusafont, Campus
Universitat Auto
`noma de
Barcelona, Cerdanyola del
Valle
`s, 08193 Barcelona,
Spain
8
Instituto Nacional de
Biodiversidad, Quito 170131,
Ecuador
9
Research and Exhibitions,
Iziko Museums of South
Africa, Cape Town 8001,
South Africa
10
Departamento de
Estratigrafı
´a, Geodina
´mica y
Paleontologı
´a, Faculty of
Geological Sciences,
Universidad Complutense de
Madrid, 28040 Madrid, Spain
11
Lead contact
*Correspondence:
albvalen@ucm.es
https://doi.org/10.1016/j.isci.
2023.107212
iScience --, 107212, --, 2023 ª2023 The Author(s).
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1
ll
OPEN ACCESS
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
Langebaanweg ‘E’ Quarry is well-suited because of its temporal position at the Miocene–Pliocene bound-
ary, and its geographic location at the southern tip of Africa.
30
It is extremely fossiliferous, including
more than 230 documented species,
15,17
providing key information for the reconstruction of the environ-
ment during the Mio-Pliocene in southern Africa, and the evolution and dispersal of animals in this
region. The taxonomy, evolution, and diversity of this fauna is therefore very important for a precise
reconstruction of the paleoenvironment. Following the initial systematic study led by Hendey,
27,31–38
continuous taxonomic revision of macro-mammal fossils, including hyaenids,
39
viverrids,
40
mustelids,
23,24
canids,
25
equids,
41
and proboscideans
42
has been undertaken.
Sabertooth cats are also a very important and diverse group at Langebaanweg. Four species of saber-
tooth, including cf. Homotherium sp., Machairodus sp., Dinofelis diastemata,andFelis obscura,were
described by Hendey.
12
Several studies have discussed or mentioned the felid taxa from this fauna,
e.g. cf. Homotherium sp. Werdelin and Sardella,
43
Machairodus sp. Werdelin and Peigne
´,
44
Felis
obscura Turner,
45
and D. diastemata Werdelin and Lewis,
46
but no systematic comparison or study
has focused on any of these species, or the sabertoothed guild as a whole. Recently, knowledge of
the Mio-Pleistocene sabertoothed cats has greatly improved,
46–51
enabling a systematic revision of
this group from Langebaanweg. The current study presents an update of the dentognathic remains
of sabertooth felids from Langebaanweg ‘E’ Quarry, comprising the classical material
12
and numerous
previously unpublished fossils (Table S1). Our main goal is to update our knowledge of these signif-
icant felids from Langebaanweg, composed of Miocene and Pliocene genera, to better understand the
Figure 1. Location of the Langebaanweg ‘E’ Quarry fossil site
(A) Silhouette of Africa, pointing out the site of Langebaanweg (gray star).
(B) simplified geographic map of the West Coast of South Africa. Modified from Valenciano & Govender (2020b). WCFP,
West Coast Fossil Park.
ll
OPEN ACCESS
2iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
taxonomy, paleobiology and ecology of one of the best-preserved and most diverse guilds of saber-
tooths from Africa.
RESULTS
Systematic paleontology
Order Carnivora Bowdich, 1821.
Family Felidae Fischer, 1817.
Subfamily Machairodontinae Gill, 1872.
Tribe Machairodontini Gill, 1872.
Lokotunjailurus Werdelin, 2003.
Type species: Lokotunjailurus emageritus Werdelin, 2003.
Included species: Lokotunjailurus fanonei Bonis et al., 2010 and Lokotunjailurus chinsamyae sp. nov.
Chronology and distribution: Currently known from central, eastern, and southern Africa, late Miocene to
earliest Pliocene.
Amended diagnosis: Machairodontinae of medium to large size. Very large infraorbital foramen. Alisphe-
noid canal present. Glenoid fossa antero-posteriorly elongated, and slightly overhung. Mandibular flange
absent or very weak, horizontal ramus slender, ascending ramus relatively tall for a machairodont cat.
Upper canine moderately laterally compressed, and relatively low-crowned. Serrations present on both
anterior and posterior edges. P2 present and often double rooted. Upper carnassial long and slender
with highly reduced protocone; p2 can be present; p3 smallwithsingleordoubleroot.Lowercarnassial
long, slender, and low. Metaconid-talonid complex absent.
Lokotunjailurus chinsamyae sp. nov.
Machairodus sp. p.149, Hendey, 1974.
Machairodus sp. p.248, Turner, 1990.
Amphimachairodus sp. B., p. 201, Werdelin, 2006.
’Machairodus’ sp. p.73, Sardella and Werdelin, 2007.
Holotype: SAM-PQL20505, fragmented maxilla, mandible, and anterior part of basicranium of the same in-
dividual (Figures 2,3,and4).
Etymology: In honor of Prof. Anusuya Chinsamy, a world renowned South African paleontologist for her
contributions to vertebrate paleontology.
Type locality: Langebaanweg ‘E’ Quarry, LQSM, earliest Pliocene (5.2 Ma).
Assigned material: SAM-PQL12641, right fragmentary hemimandible with i3-c, and p2-3 alveoli, and frag-
mentary p4-m1. SAM-PQL-22193, six thoracic vertebrae, three lumbar vertebrae, two femora, two tibiae,
the left fibula, the right astragalus, the left calcaneum and the second, third and fourth metatarsals; and
SAM-PQL-52061, complete left calcaneum described in Rabe et al.
52
The large size, and its relatively
high femur greater trochanter indicate a large felid affinity. Moreover, itsmorphology does not fit with Am-
phimachairodus (and presumably Adeilosmilus),butisclosertoLokotunjailurus, as has been discussed in
Rabe et al.,
52
and here assigned to L. chinsamyae sp. nov.
Chronology and distribution:KnownfromthetypelocalityLangebaanweg, ‘E’ Quarry (LQSM), South
Africa, early Pliocene (5.2 Ma).
Diagnosis: Medium-sized machairodontine similar to a large jaguar. Mandibular flange indistinct. Symphy-
sis very robust. Upper canine relatively robust and low-crowned, with clear serration. P2 double or single
ll
OPEN ACCESS
iScience --, 107212, --, 2023 3
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
rooted. P3 less than 1/2 length of P4. P4 with large preparastyle. p2 present and single-rooted. p3 relatively
large and double rooted. m1 slender.
Differential diagnosis: Differs from type species, L. emageritus, by having slightly stronger
mandibular flange, less posteriorly extended angular process and less posterior inclined coronoid pro-
cess, retaining of the p2, and less reduced p3; differs from L. fanonei by having smaller size,
longer C-P3 diastema, slightly smaller upper canine, retaining of the p2, less posteriorly extended
angular process; differs from Machairodus and Amphimachairodus by smaller size, especially upper
canine, presence of double rooted P2, proportionally smaller p3/P3 (compared with P4/m1), slenderer
P4; differs from Homotherium by presence of P2, proportionally larger p3/P3 (compared with P4/m1),
weaker mandibula flange, and non-ventrally extended glenoid fossa, and presence of the alisphenoid
canal.
Remarks: The specimens SAM-PQL20505 and SAM-PQL12641 were described in Hendey (1974)
12
as
Machairodus sp. (Figures 2,4,and5;Tables 1 and 2). The maxilla is incomplete but the dentition is
well preserved. The canine is relatively low crowned for a machairodontine. There is strong serration
in both keels of the canine. P2 is present on both sides, but on the left side it is single rooted whereas
on the right side it is double rooted. The P3 is slender, with a mesial accessory cusp located slightly
medial to the main cusp. The distal cingulum is cusp-like. The P4 is very slender and elongated. The
preparastyle is large and in line with the main axis of the tooth. The protocone is highly reduced
(more than that of Amphimachairodus, and similar to the state of early Homotherium), and located
at or slightly distal to the parastyle/paracone notch. The mandible has a complete horizontal ramus,
but the dentition is broken. The symphysis is very thick. The mandible flange is small. There are two
mental foramina. The anterior one is larger. The lower dentition is poorly preserved, but p2 is present,
and the p3 is very small.
There is a small fragment of the basicranium of SAM-PQL20505 not described by him (Figures 3E and 3F). It
preserves several very important anatomical traits that are included here. The glenoid slightly overhangs
the basicranium. It is distinctly antero-posteriorly elongated and faces more ventrally. The oval foramen
is located medial to the glenoid fossa. A short alisphenoid canal is present.
Figure 2. The holotype of Lokotunjailurus chinsamyae sp. nov., SAM-PQL20505 from Langebaanweg ‘E’ Quarry
(A–C)LeftmaxillarywithC,P3-P4.(A)Buccal,(B)Occulsal,and(C)Lingualviews.
(D–F) Right maxillary with P3-4. (D) Buccal, (E) Occlusal, and (F) lingual views. Scale bar equals 3 cm.
ll
OPEN ACCESS
4iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
Adeilosmilus aff. kabir
cf. Machairodus, p.149, Hendey, 1974
cf. Homotherium sp., p.158, Hendey, 1974.
Amphimachairodus sp. indet., p. 125, Werdelin and Sardella, 2006.
Amphimachairodus sp. A., p. 201, Werdelin, 2006.
Material: SAM-PQL11846, a broken left upper canine from LQSM (Figure 5C), and SAM-PQL6386, a
mandibular fragment retaining the symphysis to the p3 part from MPPM (Figures 4Fand4G).
Remarks: The upper canine was initially described in Hendey (1974).
12
It is very large, and the last third of
the crown including the tip is missing and restored. The upper canine has distinct serration in both keels.
The enamel is clearly more extended on its distal side. The mandibular fragment has a moderately sized
flange, an antero-posteriorly narrow symphysis. The mesial border seems to be vertical. Two large mental
foramina are present and located between the canine and the p3. The p2 is not present.
‘‘Metailurini’’
Dinofelis Zdansky, 1924.
Dinofelis werdelini sp. nov.
D. diastemata, p.169, Hendey, 1974.
Dinofelis barlowi, p.250, Turner, 1990.
Figure 3. Basicranial material of sabertoothed cats from Langebaanweg
(A–D) Braincase of Dinofelis werdelini sp. nov. SAM-PQL41573, posterior, ventral, dorsal and lateral views; (E and F)
glenoid and basicranial fragment of Lokotunjailurus chinsamyae sp. nov. SAM-PQL20505, anterior and ventral views. Note
on the presence of elevated glenoid fossa and alisphenoid canal in Lokotunjailurus chinsamyae sp. nov., neither of which
is present in Dinofelis werdelini sp. nov. Scale bar equals 10 cm.
ll
OPEN ACCESS
iScience --, 107212, --, 2023 5
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
D. cf. D. diastemata, p.237, Werdelin et Lewis, 2001.
Dinofelis sp., p. 201, Werdelin, 2006.
Holotype: SAM-PQL-47523, fragmented maxilla (Figures 6A–6G).
Paratype: SAM-PQL-20284, a pair of nearly complete mandibles, with associated upper canine,
Langebaanweg ‘E’ Quarry, LQSM (Figures 7A–7D).
Etymology: In honor of Prof. Lars Werdelin, who has made great contributions to African Palaeontology,
and the systematics and evolution of Carnivora, especially Dinofelis.
Type locality: Langebaanweg ‘E’ Quarry.
Assigned materials: SAM-PQL-20995, a partial mandible with complete toothrow from LQSM; SAM-PQL-
20685, a nearly complete mandible, and associated lower canine, m1 and upper canines from LQSM; SAM-
PQL-41573 braincase of a cranium, unknown member; SAM-PQL-51591, a partial mandible with complete
toothrow, from LQSM; SAM-PQL-50128, a maxilla fragment with P4 and m1 from MPPM; SAM-PQL-20702,
a single upper canine, unknown member; SAM-PQL-12237, a mandibular fragment with p4 and m1, un-
known member; SAM-PQL-41054, pair of partial mandibles of a juvenile individual, with dp3 and dp4,
from MPPM; SAM-PQL-41726, a partial mandible of a juvenile individual, with dp3 and dp4, from MPPM.
Chronology and distribution: Known from the type locality, Langebaanweg ‘E’ Quarry (LQSM and MPPM),
South Africa, early Pliocene (5.2 Ma).
Figure 4. Mandibular fragments of Lokotunjailurus chinsamyae sp. nov and Adeilosmilus aff. kabir from
Langebaanweg ‘E’ Quarry
Lokotunjailurus chinsamyae sp. nov (A–E). (A–C) SAM-PQL-12641, dorsal, lateral and anterior views; (D and E) SAM-PQL-
20505, lateral and dorsal views; Adeilosmilus aff. kabir (F and G) SAM-PQL-6386, lateral and dorsal views. Scale bar equals
5cm.
ll
OPEN ACCESS
6iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
Diagnosis: Medium Dinofelis, similar to a large jaguar. Mandibular flange indistinct, but with clear mental
ridge; mandibular symphysis high, and corpus deep; mastoid process enlarged; Upper canine robust;
cheek teeth relatively small; P2 absent; P3 with large anterior accessory cusp and distinct postero-lingual
convexity. P4 with variable preparastyle.
Differential diagnosis: Differs from all the African Dinofelis, by having proportionally smaller cheek teeth,
less reduced anterior premolars, more robust upper canine, and higher mandibular symphysis; differs
further from Dinofelis petteri,Dinofelis aronoki and Dinofelis piveteaui by having longer rostrum, less elon-
gated P4 without strong reduction of the protocone; differs further D. barlowi by having smaller size, less
reduced P4 protocone; differs further from D. piveteaui by having smaller incisor, without V-shaped acces-
sory cusps in I1 and I2; from D. diastemata by having slightly more robust upper canine, wider mastoid pro-
cesses with distinctly more ventrally extended anterior branch, and deeper mandible; from Dinofelis cris-
tata by having a distinctly smaller body size, shorter rostrum, more distinct angle between the anterior and
ventral borders of symphysis, higher but antero-posteriorly narrower coronoid process, more distinct ante-
rior and posterior ridges of the upper canine, more robust P4 protocone, and less reduced p3; differs from
Dinofelis palaeoonca by having slightly larger size, longer rostrum, more ventrally extended anterior
branch of the mastoid process, deeper mandibular corpus; from Metailurus by having distinctly larger
and deeper mandibular corpus, more distinct P3 anterior accessory cusp, absence of the m1 metaconid-
talonid complex; from ‘‘Metailurus’’ ultimus by larger size, more ventrally extended anterior branch of mas-
toid, deeper mandibular corpus, stronger postero-lingual convexity of the P3; from Paramachaerodus by
having distinct larger size, more ventrally extended anterior branch of mastoid, deeper mandibular corpus,
more distinct P3 anterior accessory cusp; differs from Adelphailurus and Yoshi by having much larger size,
lower-crowned cheek teeth, presence of distinct P3 anterior accessory cuspid, and absence of the m1
metaconid-entoconid complex.
Remarks: The specimens SAM-PQL-20284, SAM-PQL-20685, SAM-PQL-20702, SAM-PQL-12237 were
described in Hendey
12
as D. diastemata. The maxilla SAM-PQL-47523 was selected as the holotype
because the upper dentition of felids is often more diagnostic (Figures 6A–6G). It preserves complete
dentition from I1 to M1. The incisor row is only very weakly procumbent. The incisors are small, especially
the I1 and I2, with both having two distinct posterior accessory cusps. I3 is distinctly larger than the I1 and
I2 and has a distinct medial accessory cusp. The canine is broken. P3 has a strong anterior accessory
cusp, and a weak additional cusplet behind it. The posterolingual corner is convex, and the posterior
cingulum is well developed and laterally extended. The P4 has a strongly undulated buccal border.
Figure 5. Upper canine of sabertoothed cat from from Langebaanweg ‘E’ Quarry
(A) Dinofelis werdelini sp. nov., SAM-PQL20284;
(B) Lokotunjailurus chinsamyae sp. nov., SAM-PQL20505; (C) Adeilosmilus aff. kabir, SAM-PQL11846. Scale bar equals
5cm.
ll
OPEN ACCESS
iScience --, 107212, --, 2023 7
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
The preparastyle is large, and oriented in a strong angle with the parastyle. The protocone is not
reduced and is located at the level between the parastyle and paracone. The M1 is small, transversely
shortened.
The maxillary fragment SAM-PQL-50128 (Figures 6H and 6I) has P4 and M1 preserved. There is only an
indistinct preparastyle present, and the buccal border is less undulating. M1 is similar/same as that of
SAM-PQL-47523.
The braincase SAM-PQL-41573 (Figures 3A–3D) has a well-developed sagittal crest and its widest point
is at the level of the glenoid. The occipital plane is posteriorly inclined. In lateral view, both mastoid
and paroccipital processes are well developed. The mastoid process is distinctly anteroventrally
extended and does not cover the ventral border of the auditory bulla. The anterior branch of the mas-
toid extends to a similar level as the posterior branch, and the posterior branch is strongly posteriorly
inclined. The paroccipital process is posteriorly extended and is detached from the bulla. The auditory
bulla is inflated and has its highest point at its anterior part. In the ventral view, the glenoid fossa does
not overhang the basicranium, and is not antero-posteriorly elongated. It faces more anteriorly than
that of Lokotunjailurus (Figures 3E and 3F). The oval foramen is located medial to the glenoid fossa
and was partially covered by a bone bridge. The auditory bulla is enlarged. The medial border is nearly
straight. The opening of the stylomastoid foramen and hyoid fossa share a common opening. The mas-
toid is laterally expanded, located at the buccal side of the common opening mentioned above. The
paroccipital process is located slightly medial to this opening. The posterior lacerated foramen is
rather narrow and separated from the hypoglossal foramen. In posterior view, the nuchal crest is
wide, robust, and well developed dorsally. The muscle scar in the occipital plane is strongly developed,
especially near the edge of the lambdoidal crest.
Table 1. Measurements and ratios of the upper dentition of sabertoothed cats in this study
Taxon Dinofelis werdelini
?Dinofelis
werdelini Yoshi obscura
Lokotunjailurus
chinsamyae
Catalog L20284 L47523 L L47523 R L20685 L L20685 R L50128 L20702 L2674 L10100 L20505 L L20505 R
CL 20.17 18.69 17.55 17.88 20.84 18.61 23.23
CW 13.24 14.53 10.52 10.73 13.25 9.01 11.22
CH 50.04 41.96 42.58 37.30 49.40
CW/L 0.66 0.78 0.60 0.60 0.64 0.48 0.48
CH/L 2.48 2.39 2.38 1.79 2.13
P3L 19.80 20.83 19.35 14.00 15.55
P3W 11.63 11.97 9.04 6.30 6.84
P3W/L 0.59 0.58 0.47 0.45 0.44
P4L 33.45 33.29 30.49 23.80 37.71 37.19
P4W 17.82 17.03 14.38 14.18 12.20 12.37 13.25
P4BW 11.92 12.33 9.84 9.74 10.69 10.85
P4W/L 0.53 0.51 0.47 0.51 0.33 0.36
P4BW/L 0.36 0.37 0.32 0.28 0.29
M1L 5.62 4.23 5.60 4.48
M1W 10.05 9.04 7.40 6.26
M1W/L 1.79 2.14 1.32 1.40
CL/P4L 0.56 0.55 0.63
P3L/P4L 0.59 0.63 0.59 0.41
M1W/P4L 0.30 0.31 0.17
LT 87.52 59.65 88.16
LDP 9.80 5.20 6.39 7.39 7.61
LDP/LT 0.11 0.11 0.09
ll
OPEN ACCESS
8iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
There are several mandibles with well-preserved dentition (SAM-PQL-20284, SAM-PQL-20685, SAM-PQL-
20995, SAM-PQL-51591), but unfortunately none of them are complete. The horizontal ramus is generally
deep, especially in the region of the symphysis. The coronid process is large, and the angular process is
ventrally extended. There are a variable number of mental foramina. The lower canine is large, with a
sharp distal keel. The c-p3 diastema is elongated. The p3 is much lower crowned than the p4. The mesial
border of the p4 and the m1 is posteriorly inclined. The m1 is robust, without a metaconid-talonid
complex.
The juvenile mandible SAM-PQL-41054 (Figures 7M and 7N) has the deciduous dp3 and dp4 in place, and
the permanent canine and m1 in an initial stage of eruption. The m1 morphology is the same as other spec-
imens of the m1 in Langebaanweg and supports its assignment to this taxon. The high symphysis is already
Table 2. Measurements and ratios of the lower dentition of sabertoothed cats in this study
Taxon Dinofelis werdelini
Lokotunjailurus
chinsamyae
catalog L20995 L20284 L20284 L51591 L12237 L20685 L41054 L L41054 R L41726 L12641 L-20505
cL 15.20 15.79 16.41 12.92
cW 9.92 10.56 12.07 11.43
cH 23.77 25.25 20.58 22.29
cW/L 0.65 0.67 0.74 0.89
cH/L 1.56 1.60 1.25 1.73
p3L 14.03 14.54 13.99 15.25 12.30
p3W 6.47 7.22 7.41 8.84 5.50
p3W/L 0.46 0.50 0.53 0.58 0.45
p4L 20.09 22.38 22.45 22.64 20.47 20.60 20.00
p4W 9.35 9.25 9.95 10.67 8.81 8.70 9.50
p4W/L 0.47 0.41 0.44 0.47 0.43 0.42 0.48
m1L 22.70 23.78 23.78 25.63 23.87 23.00 28.00 29.00
m1W 10.70 11.59 10.66 12.77 10.22 10.20 11.00 11.20
m1W/L 0.47 0.49 0.45 0.50 0.43 0.44 0.39 0.39
dcL 8.81
dcW 3.71
dp3L 13.59 13.45 12.68
dp3W 4.33 4.39 4.70
dp4L 17.83 17.81 18.80
dp4W 7.23 7.86 6.19
c/m1 0.64 0.66 0.64 0.56
p3/m1 0.62 0.61 0.59 0.60 0.44
p4/m1 0.89 0.94 0.94 0.88 0.86 0.90 0.71
LT 89.49 98.71 97.93 101.10 86.26 111.83
LDP 26.04 25.15 24.16 26.27 20.51 25.34
LM 55.90 59.42 59.12 61.33 56.48 66.05
H1 32.15 44.13 39.90 40.44 36.01 38.92
W1 19.98 26.25 24.10 23.95 20.96 31.56
H2 29.10 31.82 32.52 32.02 25.23 34.09
H3 37.53 38.70 37.27 32.88 37.01 39.14
W3 17.87 18.53 24.63 16.47 17.84 17.58
LDP/LT 0.47 0.42 0.41 0.43 0.36 0.38
H1/H3 1.18 1.03 1.09 1.10 1.05
m1L/H3 0.63 0.61 0.69 0.70 0.76 0.74
ll
OPEN ACCESS
iScience --, 107212, --, 2023 9
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
developed. The dp3 has sharp main cuspid and anterior and posterior accessory cuspids, and a cusp-like
posterior cingulid. The dp4 has well developed metaconid and talonid. Its lingual border is strongly
concave. The juvenile mandible SAM-PQL-41726 (Figures 7O and 7P) belongs to a slightly younger individ-
ual than SAM-PQL-41054, with both the canine and m1 unerupted. The overall morphology is similar, but
the posterior cingulid of the dp3 in SAM-PQL-41726 is less developed.
?D. werdelini.
D. diastemata, p.169, Hendey, 1974.
Material: SAM-PQL-2674, a fragmented maxilla with canine, and fragments of maxilla with P3-M1,
from MPPM.
Remarks: This material was initially described and figured as D. diastemata by Hendey.
12
ItsP3issimilarto
the P3 of SAM-PQL-47523, with a distinct anterior accessory cusp, but its P4 buccal border is much less un-
dulating, and the protocone is distinctly smaller. The canine of this individual is much narrower, and differs
from the other canine assigned to D. werdelini. It is unclear whether this individual represents another spe-
cies or a variant of D. werdelini.
Yoshi Spaasov et Gerrads, 2014.
Yoshi obscura (Hendey, 1974)
Felis obscura, p.164, Hendey, 1974.
Figure 6. Maxillary fragments of Dinofelis werdelini sp. nov. from Langebaanweg ‘E’ Quarry
(A–G) Holotype: SAM-PQL-47523, (A–D) lateral, medial, anterior and ventral views of right side, (E–G) occlusal, medial and
lateral views of left side; (H and I) SAM-PQL-50128, occlusal and lateral views. Scale bar equals 5 cm.
ll
OPEN ACCESS
10 iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
Adelphailurus obscura, p.250, Turner, 1999.
?Megantereon obscura, p.43, Morales et al., 2005.
Metailurus obcurus, p. 201, Werdelin, 2006.
Metailurus obscurus, p.646, Werdelin et Peigne
´, 2010.
Holotype: SAM-PQL10100 from Langebaanweg ‘E’ Quarry, MPPM (Figure 8).
Chronology and distribution: Known from the type locality, South Africa, earliest Pliocene.
Emended diagnosis: Small metailurine, size of a female leopard. Infraorbital foramen moderate in size.
Small canine. Short C-P3 diastema. P3 high crowned without anterior accessory cusp. P4 slender, with large
protocone, without distinct preparastyle. M1 small, button-like and transversely short.
Differential diagnosis: Differs from all other machairodonts except Yoshi spp. in having smaller size and
straight P4 buccal border; differs from all other species of Yoshi by having less high crowned P3, and reduc-
tion of the M1 inner lobe.
Figure 7. Mandibles of Dinofelis werdelini sp. nov. from Langebaanweg ‘E’ Quarry
(A–D) paratype: SAM-PQL-20284.
(E and F) SAM-PQL-20685.
(G and H) SAM-PQL-20995.
(I and J) SAM-PQL-12237.
(K and L) SAM-PQL-51591.
(M and N) juvenile, SAM-PQL-41054.
(O and P) juvenile, SAM-PQL-41726. All lateral and dorsal views. Scale bar equals 10 cm.
ll
OPEN ACCESS
iScience --, 107212, --, 2023 11
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
Remarks: The maxillary fragments contain complete dentition from the upper canine to the M1. The in-
fraorbital foramen is not enlarged and located at the level of the P3. The upper canine is moderately robust
with a posterior keel. There is no P2. The P3 is slender without a mesial accessory cusp but the distal acces-
sory and cingulum are well developed. The P4 is also slender with a straight buccal border, and no prepar-
astyle. The protocone is large and located mesial to the parastyle/paracone notch. The M1 is transversely
shortened.
DISCUSSION
Phylogenetic analysis
The evolutionary position of Langebaanweg felids has been previously discussed, but never tested by
phylogenetic analysis.
12,30,43
Our detailed description of both old and new materials enables such an anal-
ysis for the first time. We use the matrix and the character descriptions and states from the recent work on
sabertooth phylogeny by Jiangzuo et al.,
53
and added three newly revised taxa here (appendix nexus file,
Table S2). The coding of Adeilosmilus is still based on previous material, as the specimens from Lange-
baanweg did not introduce new knowledge.
Figure 8. Holotype of Yoshi obscura, SAM-PQL10100, from Langebaanweg ‘E’ Quarry
(A) Bucal, (B) lingual and (C) Oclussal views. Scale bar equals 5 cm.
ll
OPEN ACCESS
12 iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
The overall topology of the phylogeny (Figure 9) is very close to the one in Jiangzuo et al.
53
The smaller Lan-
gebaanweg Machairodontini is supported as a member of Lokotunjailurus, sister to the two previo usly known
species. The new species D. werdelini is supported as the earliest definitively defined species of Dinofelis.
Y. obscura is sister to recently described Yoshi yongdengensis. These two species have similar size and share
the derived trait of transversely shortening of the M1 compared with other species of this genus.
The affinity of two machairodontini
The tribe Machairodontini flourished during the Late Miocene, with numerous speciesfrom Africa to North
America.
54
The two most common genera are Machairodus and Amphimachairodus.
54,55
Amphimachairo-
dus was viewed as invalid as the species within this genus shows a continuous evolution from Machairodus
aphanistus,
56–58
but a recent paper suggest there is significant ecological change from Machairodus to Am-
phimachairodus,andAmphimachairodus evolved early in the border of the Tibetan Plateau,
59
therefore
Amphimachairodus is accepted as a valid genus in this study.
In Africa, two distinct body sizes are recognised for machairodont felids, the larger one, referred to Ma-
chairodus or Amphimachairodus, and recently Adeilosmilus, and the smaller Lokotunjailurus.
49,60
Lokotun-
jailurus is previously known only from Central and East Africa.
44,49,60
Machairodus kabir, the large machairodont, was established based on mandible and humerus material
from the Late Miocene hominid locality of TM 266, Toros-Menalla, Chad.
60
Later, a complete mandible
and a partial cranium preserving the braincase and surrounding parts, from Sahabi of Libya were assigned
to this species and generically changed to Amphimachairodus.
61
A lower canine of an indeterminate saber-
tooth has been found in the early Pliocene of West Africa (Tobe
`ne fauna, Senegal) and could be related
to this lineage.
62
Recently, a phylogenetic analysis supported this species belonging to its own genus
Adeilosmilus.
63
Our new analyses also support this view (Figure 9). This species has a large body size
and dentition. It is characterized by a derived dentition comprising a more reduced p3 and p4 compared
with those of Amphimachairodus. The p3 is no more than half of the m1 length, and p4/m1 length ratio is
smaller than 0.75.
56,60,61,64–67
This genus is intermediate between Amphimachairodus and the more
derived Plio-Pleistocene Homotherium. The glenoid fossa from the partial cranium from Sahabi is distinctly
Figure 9. Bayes inferences tree using tip-dating methods and constraints on Felidae and Machairodontinae
The position of the four sabertooth from Langebaanweg are in bold. The outgroup is Canis lupus. Numbers in the nodes represent the posterior probablility.
It was implemented using the software Mrbayes 3.2.7 (https://nbisweden.github.io/MrBayes/download.html).
ll
OPEN ACCESS
iScience --, 107212, --, 2023 13
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
antero-posteriorly wide and overhangs the basicranial region, similar to Homotherium and is more derived
than Amphimachairodus.
L. emageritus was established based on a partial skeleton from Lothagam, Nawata Formation, Kenya.
49
This species is characterized by highly derived dentition comprising a P4 with a large preparastyle, and
a very reduced protocone. The lower premolars are reduced and similar to Adeilosmilus kabir.The
mandible is slender, without a distinct mandibular flange, and the coronoid process is well developed,
larger than that of Amphimachairodus.L. fanonei was later identified from Toros Menalla, Chad.
50
The
maxilla suggests that P2 is present and double-rooted, in contrast to its highly reduced premolars. This
genus shows an interesting combination of derived dentition, and conservative mandibular (weak flange,
large coronoid process), and dental traits (retaining of the double rooted P2). The Sahabi mandible as-
signed to Dinofelis sp. by Rook and Sardella
68
has a reduced and double-rooted p3. This specimen fits
L. fanonei in morphology and probably represents the same species.
At Langebaanweg (‘E’ Quarry), South Africa, two Machairodontini species have been identified at the genus
level. Hendey
12
assigned the smaller one to Machairodus sp., and the larger one to cf.Homotherium sp. A
detailed analysis of the canine, together with some postcranial bones, especially the astragalus and calcaneum,
supported that this species should be better assigned to Amphimachairodus.
43
This material does not provide
clues for a more specific taxonomic assignment.The new mandibular fragment SAM-PQL-6386 (Figures 4Fand
4G) probably represents this species. It is slightly larger than that of Lokotunjailurus from the same locality, and
more importantly, has a more expanded symphysis, with a distinct mandibular flange, which is absent in Loko-
tunjailurus. The mandibular heights of SAM-PQL-6386 behind the canine and in frontof p3 are similar to those of
the small individuals of Amphimachairodus coloradensis and Amphimachairodus horribilis. The p3 is much
smaller than any individuals of A. coloradensis-horribilis, suggesting a reduction of this tooth. This is character-
istic of Ad. kabir from Toros-Menalla, Chad
60
and Sahabi, Libya,
61
which has a distinctly larger overall size and
stronger mandibular flange. We suggest that the large machairodont from Langebaanweg represents a species
with the closest affinity to Ad. kabir and is herein ascribed to Adeilosmilus aff. kabir.
The potential presence of species belonging to or closely related to Lokotunjailurus in Langebaanweg has
recently been proposed based on postcranial bone morphology and limb proportions.
52
This is in accor-
dance with our study on craniodental material. The original and the new material of Machairodus sp. from
Langebaanweg is represented by several fragments of a skull, and a partial mandible (Figures 2 and 3E, 3F,
4A–4E, 5B, and 10 [as L. chinsamyae sp. nov.]). Sardella and Werdelin
61
suggested that this species is close
to Lokotunjailurus in many aspects but differs in many other aspects. We agree with this view. The Machair-
odus sp. material from Langebaanweg shows characteristic traits of Lokotunjailurus such as a relatively low
and robust canine with distinct serration, reduction in size of the anterior premolars, P4 elongated with
straight buccal border, large preparastyle, long metacone, and a very reduced protocone. This combina-
tion is not seen in other machairodonts except Lokotunjailurus.Amphimachairodus spp. are generally
large, with larger mandibular flange, higher and narrower upper canine, proportionally larger anterior pre-
molars, undulating P4 buccal border, with larger protocone but shorter metacone.
56,57,64,65,69
Homothe-
rium spp. and Ischyrosmilus-Xenosmilus are also generally larger, with larger mandibular flange and
distinctly deeper mandibular corpus, higher and narrower upper canine, proportionally smaller anterior
premolars, more reduced P4 preparastyle, and protocone.
63
Machairodus spp. have proportionally larger
anterior premolars, undulating P4 buccal border, with smaller and angled preparastyle, distinctly larger
protocone, and shorter metacone. The material from Langebaanweg therefore should be assigned to
Lokotunjailurus. It differs from both known species as indicated in the differential diagnosis, and merits
a new species name, L. chinsamyae. Our phylogenetic analysis (Figure 9)pointstoL. chinsamyae as the sis-
ter group of the two known species of the genus (L. emageritus and L. fanonei).
The material from Langebaanweg provides many important craniodental traits for this bizarre machairo-
dont. The P2 is present and double rooted on one side, as in L. fanonei, suggesting this trait is not a vari-
ation in the latter species, but probably common to the genus. The p2 is also present, which seems to be
lost in L. fanonei. We list this trait in diagnosis, but note that this tooth could be variably present in some
species, e.g., Paramachaerodus ogygia.
48
The glenoid fossa is derived in morphology, being antero-pos-
teriorly widened, and faces ventrally, like in Homotherium, but only slightly overhangs the basicranium,
similar to Amphimachairodus. Especially interesting is the presence of the alisphenoid canal. This structure
is absent in all living felids and derived machairodonts, e.g., Homotherium,Smilodon,Megantereon,and
ll
OPEN ACCESS
14 iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
all metailurine, but present in more primitive species, e. g. Proailurus,Hyperailurictis.
54,70–73
It is generally
absent in Amphimachairodus,
64
though occasionally is remnant in some individuals.
57
The situation in
M. aphanistus is not mentioned by Anto
´netal.,
47
butitispresentinNorthAmericanNimravides (Q. Jiang-
zuo, personal observation).
In summary, Lokotunjailurus shows a mosaic of craniodental traits, with highly derived cheek tooth
morphology, and glenoid fossa, but retaining primitive mandibular morphology (no or very weak
mandibular flange, large coronoid process), and being conservative in retaining P2, p2 (sometime spe-
cies), and alisphenoid canal. This combination precludes a judgment of its position within Machairo-
dontini, but it is likely to be a unique lineage, converging with Amphimachairodus-Homotherium line-
age in dental traits. Lokotunjailurus was described as having an appendicular skeleton relatively
slender, lacking extreme machairodont features, which reveals it as a cursorial felid.
49,74
Thus, three
species of Lokotunjailurus roamed Africa during the late Miocene and early Pliocene. During the
Late Miocene L. emageritus occurred in East Africa (Lothagam, Kenya),
49
and L. fanonei in Central
and North Africa (Toros Menalla, Chad and probably in Sahabi, Lybia)
50,68
and at the beginning of
Figure 10. Comparative measurements and ratios of Machairodontini from Langebaanweg ‘E’ Quarry and other
Neogene sites compared to the similar-sized metailurine Dinofelis werdelini from Langebaanweg ‘E’ Quarry
(A) C length vs width/length ratio; (B) P3 length vs width/length ratio; (C) P4 length vs width/length ratio; (D) P4 length vs
P3/P4 length ratio; (E) upper tooth row (C-P4) length vs (C-P3) diastema/tooth row length ratio; (F) p3 length vs width/
length ratio; (G) p4 length vs width/length ratio; (H) m1 length vs width/length ratio; (I) m1 length vs p3/m1 length ratio;
(J) m1 length vs p4/m1 length ratio; (K) lower tooth row (c-m1) length versus (c-p3) diastema/tooth row length ratio. Biplot
measurements/ratios were made using ggplot2 (https://cran.r-project.org/web/packages/ggplot2/index.html)ofR
(https://www.r-project.org/).
ll
OPEN ACCESS
iScience --, 107212, --, 2023 15
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
the Pliocene L. chinsamyae lived in the South of the continent (Langebaanweg). This confirms the high
mobility of this taxon as a Pan-African sabertooth felid.
The postcranial bones described by Rabe et al.
52
show similarity to L. emageritus, and are here assigned to
L. chinsamyae. The size of the postcranial is slightly larger than the corresponding bone of L. emageritus
(e.g., tibia), but such degree of differences does not support a distinct body size discrepancy of the two
species. It is interesting to note that the maxilla and mandible of L. chinsamyae are only slightly larger
than those of D. werdelini (see Figure 10), but the postcranial bones are distinctly larger, suggesting the
different body plane of these machairodonts.
The affinity of two metailurine
Several small to medium-sized sabertoothed genera, assigned to metailurine, were known from the Late
Miocene in the Old and New World. Metailurus was established on a well-preserved skull from Baode,
northern China.
66
The type species Metailurus major is of medium size, with moderately developed upper
canine, high-domed cranium, and wide forehead. A smaller species, M. minor was also found in Baode.
66
This small species is probably similar or identical to ‘‘Machairodus’’ parvulus from Pikermi,
75
but as Spassov
and Geraads
51
suggested, the type of latter species is too poorly preserved, and the species should be
treated as nomen dubium. This species is craniodentally different from M. major, and was erected as a
separate genus Yoshi by Spassov and Geraads.
51
Paramachaerodus is another medium-sized sabertooth,
with a similar body size to that of Metailurus.
48
This genus is thought to be or close to the ancestor of the
Plio-Pleistocene Smilodon and Megantereon, and was referred to Smilodontini.
54
The overall morphology,
however, is not significantly different from Metailurus, with a lower cranium and narrower forehead. Some
authors support separating the Vallesian species P. ogygia to a separate genus Promegantereon,
48,76
whereas some other authors support the two genera are indistinguishable.
77
Dinofelis is the largest and
most widely spread member of the metailurines. It was first established based on a skull from the Quater-
nary site of Mianchi, Henan, central China, as Dinofelis abeli.
66
Several new species were later recognized
from Africa, Europe, and North America see a review by.
46
The North American medium-sized saber-
toothed cats are represented by Adelphailurus kansensis and Pratifelis martini.
78
The validation of the
former genus is generally accepted, whereas the latter is largely omitted by most authors. Turner and An-
to
´n
54
assigned Adelphailurus to metailurini, whereas Martin
79
hypothesized this genus evolved from North
American Pseudaelurus (now viewed as the distinct genus Hyperailurictis).
80
Adelphailurus is generally
similar to Metailurus but differs in two important traits: the presence of the P2, and the lack of the anterior
accessory cusp in the P3. The m1 talonid of Pratifelis is well-developed, unlike any other metailurine, and its
position is unclear. Two genera were known from the Late Miocene (Turolian) of Spain, Stenailurus
81
and
Fortunictis.
82
Both genera were assigned to metailurini by Salesa et al.
48
Stenailurus is peculiar in having
the P2, whereas Fortunictis is generally not mentioned by most authors.
The medium-sized sabertoothed cat from Langebaanweg (Figures 6 and 7) shows affinity to Dinofelis,Para-
machaerodus,andMetailurus. It is distinctly larger than Yoshi and Adelphailurus, in addition to many traits
(e. g. presence of enlarged P3 anterior accessory cusp, which is absent or very weak in Yoshi and Adelphai-
lurus). The Langebaanweg medium-sized sabertoothed cat was first assigned to D. diastemata by Hen-
dey.
12
Hendey
12
compared this material to D. barlowi,D. piveteaui and D. diastemata, and concluded
that the Langebaanweg material is closer to the latter; we agree. Hendey,
12
however, did not give a direct
comparison of the Langebaanweg material to D. diastemata, nor the Asian species D. cristata. Turner
45
as-
signed the Langebaanweg material to D. barlowi without clear evidence. Werdelin and Lewis
46
suggested
that the Langebaanweg material is metrically different from D. barlowi and does not likely belong to this
species. The general morphology of this Langebaanweg felid is close to that of D. diastemata, but Werdelin
and Lewis
46
realized that this similarity is represented by primitive traits, which provides little information in
inferring its phylogenetic relationships.
The new maxillary material (Figure 6) provides important information for this machairodont. The unique P3,
P4 morphology is unlike any known species of Dinofelis.ItisclosetoD. diastemata in general proportion,
and protocone size, but the preparastyle seems to be stronger, and the P4 buccal contour hasmore undu-
lations. All other species of Dinofelis, including Asian D. cristata, have a more reduced protocone. The
other African species of Dinofelis have more reduced P3, without a strong postero-lingual convexity,
and less developed preparastyle. The incisors of the maxilla are enlarged and more arched in arrangement
compared with two Asian species of Dinofelis, and comparable to that of other African Dinofelis.The
ll
OPEN ACCESS
16 iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
mandible of the Langebaanweg machairodont is mainly characterized by its well-developed symphysis and
deep corpus. All other species of African Dinofelis have a shallow symphysis, with the height of the
mandible behind the canine being lower than that behind the m1. In contrast, for all specimens from Lan-
gebaanweg, the opposite is true (Figure 11M). In this aspect, the Langebaanweg machairodont resembles
Eurasian D. diastemata and D. cristata, with a high symphysis. Both Eurasian species, however, have a
rather elongated c-p3 diastema, with LDP/LT ratio of mandible clearly higher than that of Langebaanweg
machairodont and other African species (see Figure 11L but see 11e that the differences in the DP/LT ratio
of the maxilla are not distinct). The mandibles of the two Eurasian species are also not as deep. The lower
dentition generally shows fewer differences, and the Langebaanweg machairodont has unreduced p3
compared with most other species of Dinofelis. The upper canine is more robust than any other species,
other than D. cristata, and overlaps with D. barlowi. The dental sizes place the Langebaanweg machairo-
dont among the smallest Dinofelis, however, the toothrow length (C-P4 and c-m1 length) is only distinctly
smaller than that of D. cristata, similar to that of D. diastemata and D. barlowi, and slightly larger than that
of D. petteri and D. piveteaui.ThissuggeststhatDinofelis largely retained a similar body size, but the
derived African species developed proportionally larger cheek teeth, reaching its peak in D. aronoki-piv-
eteaui. The North American species D. palaeoonca is the smallest Dinofelis, except for a recently discov-
ered mandible from North Africa.
83
This species has a very short rostrum, and conservative dental traits,
and seems to be a unique lineage of Dinofelis. The braincase (Figure 3D) shows a clear Dinofelis affinity,
e.g., the glenoid fossa is not antero-posteriorly widened and not overhung, and the alisphenoid canal is
totally absent. The mastoid process is well developed, especially its anterior branch, which reaches a level
more than that in D. diastemata and D. palaeoonca.
Apart from Dinofelis, the Langebaanweg machairodont also shows some similarities to Metailurus and Par-
amachaerodus. Both genera are distinctly smaller, and generally have a shorter rostrum, but the dentitions
are similar. In both M. major and P. transasiaticus, the P3 has distinct postero-lingual convexity, and P4 has
undulating buccal border. The lower p3 and p4 are also similar, especially the p4 lateral view.
66,77
However,
both M. major and P. transasiaticus have m1 metaconid-talonid complex and have a somewhat trapezoid
occlusal shape (buccal border not smooth). Dinofelis, therefore, probably originated from either of these
two genera, but the Langebaanweg machairodont is already too derived to be placed in these two genera.
A new species, Metailurus ultimus was named by Li
84
from the Mazegou Formation of the Yushe Basin, Late
Pliocene. Two unpublished crania of this species from Shanxi housed in AMNH, suggest the cranium of this
species is low and the forehead is narrow. This contrasts withthat of Metailurus, but better fits with Dinofelis
as assigned by Turner and Anto
´n.
54
This species shows greater similarity to the Langebaanweg machairo-
dont, in the m1 morphology. We add the metric data of this species to our analysis, and it can be seen that
the Langebaanweg machairodont is larger and has a longer rostrum (Figure 11). Moreover, the incisors of
these taxa are also smaller and less arched in the arrangement as seen in the Langebaanweg machairodont
(Figure 12). A mandible and several other fragments of Dinofelis were reported from the Middle Awash by
Haile-Selassie and Howell.
85
This taxon, which is slightly earlier than Langebaanweg, shows and overlap in
dental size with that of the Langebaanweg felid (Figure 10), but the c-p3 diastema is shorter, and the corpus
depth (behind the m1, 24.4 mm) is distinctly smaller than that of the South African one (behind the m1,
32.88–38.70 mm) (Figures 6 and 7). The Dinofelis from the Middle Awash seems to better fit the trait of
Metailurus or Paramachaerodus and is distinct from the Langebaanweg felid.
In summary, the Langebaanweg machairodont should be assigned to Dinofelis, but is different from all
known species of this genus, and is erected as a new species, that we named D. we rdelini sp.nov.(Figures 6,
7and 9). In general, our analyses support the view of Werdelin and Lewis,
46
that this species is closer to
D. diastemata in morphology, but such similarity does not necessarily support a close phylogenetic rela-
tionship. D. diastemata seems to share a longer c-p3 diastema, and a similar mandibular morphology
with the larger and more derived D. cristata. The younger African species seem to share the further reduc-
tion of both the anterior premolars, and P4 protocone. D. petteri,D. aronoki and D. piveteaui also share the
shortening of the rostrum, and highly elongated P4. D. werdelini shares the enlarged and slightly arched
incisors with the younger African species.
The smaller sabertoothed cat from Langebaanweg (Figure 8), only known by a fragmentary maxilla, has had
a controversial systematic position since its description. It was first assigned to Felis obscura sensu lato by
Hendey.
12
This is probably because of its relatively small canine, which is broken and has an unclear crown
height. Hendey
12
thought the species may be correlated with Sivapanthera becauseofitssmallsnout.The
ll
OPEN ACCESS
iScience --, 107212, --, 2023 17
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
P2, however, is absent in this species, which is more common with Machairodontinae. In fact, the posterior
ridge of the canine is clearly present, a key trait that distinguishes Machairodontinae from Felinae. Turner
et al.
86
assigned this species to the North American genus Adelphailurus without stating their evidence.
Morales et al.
87
recognized the machairodontine affinity of this species and hypothesized it could becorre-
lated with Megantereon.However,Megantereon is a more derived genus, with much larger canine, and
smaller P4 protocone. The loss of the P2 is also not unique to Megantereon among machairodontine.
Figure 11. Comparative measurements and ratios of metailurine from Langebaanweg ‘E’ Quarry and other
selected metailurine
(A) C length vs width/length ratio; (B) P3 length vs width/length ratio; (C) P4 length vs width/length ratio; (D) P4 length vs P3/P4
length ratio; (E) upper tooth row (C-P4) length vs (C-P3) diastema/tooth row length ratio; (F) width of rostrum across P4/M1 vs
width ratio of rostrum across C and across P4/M1; (G) p3 length vs width/length ratio; (H) p4 length vs width/length ratio;
(I) m1 length vs width/length ratio; (J) m1 length vs p3/m1 length ratio; (K) m1 length vs p4/m1 length ratio; (L) lower tooth row
(c-m1) lengthvs (c-p3) diastema/toothrow length ratio; (M) mandibular height behindthe m1 vs mandibular height ratiobehind
C/behind m1. Biplot measurements/ratios were made using ggplot2 (https://cran.r-project.org/web/packages/ggplot2/
index.html)ofR(https://www.r-project.org/).
ll
OPEN ACCESS
18 iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
Finally, Werdelin and Peigne
´
44
assigned this species to Metailurus, but did not mention the trait for this
assignment.
Metrically, the body size of ‘‘Felis’’ obscura is closest to that of Yoshi spp., within its variation range,
though near the upper boundary, and is distinctly smaller than Metailurus,Paramachaerodus,andDi-
nofelis (Figure 11). Morphologically, ‘‘Felis’’ obscura is also closest to Yoshi. The shared traits include
the relatively high-crowned P3 (judging from the preserved part), without an anterior accessory cusp,
P4 with a slender blade, no or weakly undulating buccal border, and relatively large protocone. The
anterior accessory cusp of the P3 is present in most sabertooths, including derived Paramachaerodus
(Pa. transasiaticus), M. major, and Machairodontini. Notably, the North American Adelphailurus also
lacks this accessory cusp (probably the reasonwhyTurneretal.assigned‘‘Felis’’ obscura to this genus).
Yoshi generally lacks this accessory cusp, but in some cases, it is present but very small and less
defined. The P4 of Paramachaerodus and Metailurus shows a distinctly undulating buccal border,
48,66,77
which is very weak or absent in ‘‘Felis’’ obscura and Yoshi. However, there are two differences between
‘‘Felis’’ obscura and Yoshi. Most importantly, the M1 of ‘‘Felis’’ obscura is button-like and transversely
short. In Yoshi, the M1 is invariably transversely elongated, and there is weak buccal concavity between
the connected paracone and metacone. The morphology in the latter case represents the primitive
stage seen in Pseudaelurus-Hyperailurictis. Another difference is the relatively longer C-P3 diastema.
A very small-sized sabertoothed cat, Tchadailurus adei was erected from TM 112, Toros Menalla,
Figure 12. Comparison of maxilla of Dinofelis and related taxa
(A) Dinofelis werdelini sp. nov. SAM-PQL-47523, right side mirrored; (B) Dinofelis diastemata, Perpignan, cast; (C)
Dinofelis cristata, Siwaliks (photos courtesy of L. Werdelin); (D) Dinofelis barlowi UCMP88766, cast, Bolt’s Farm; (E)
Dinofelis piveteaui Ka 61, Kromdraai, from MorphoSource. A; (F) Yoshi minor PMU M3835, Baode (photos courtesy of
X. Wang); (G) Paramachaerodus transasiaticus IVPP V20106, Shilei, Linxia Basin; (H) Metailurus major PMU M3841, Baode
(photos courtesy of X. Wang). Adjust to the same size to show the different dental proportion.
ll
OPEN ACCESS
iScience --, 107212, --, 2023 19
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
Chad, 7Ma.
88
Bonis et al.
88
noticed the similarity of their new genus to Yoshi and stated that Tcha-
dailurus is different from Yoshi in smaller size, proportionally higher canine, and longer diastema. In
other aspects, this genus is much like Yoshi and unlike other metailurine, in having a high crowned
P3, without an anterior accessory cusp and a transversely elongated M1. This genus is very close, if
not identical, to Yoshi.‘‘Felis’’ obscura is also close to this species in general morphology mentioned
above, and in addition, having long C-P3 diastema.
In brief, ‘‘Felis’’ obscura is close to both Yoshi and Tchadailurus, and these two genera are closer to each
other than to ‘‘Felis’’ obscura. This, therefore, implies two hypotheses of classification. The first one, is to
include ‘‘Felis’’ obscura and T. adei in the genus Yoshi. This implies that the first two represent unique mem-
bers of the genus with elongated C-P3 diastema, and Y. obscura an independently reduced M1 inner lobe.
The second one, validates the distinct genus Tchadailurus,andthat‘‘Felis’’ obscura should be viewed as a
distinct genus too. In order to not oversplit in genera taxonomy, and that the material is incomplete, we
adopt the first classification, which includes both ‘‘Felis’’ obscura and T. adei in Yoshi. Thus, Y. obscura
emerges so far as the youngest known and most unique species of the genus. Our phylogenetic analysis
(Figure 9) confirms its assignation in the genus Yoshi, and it emerges as the sister taxon of
Y. yongdengensis from the latest Miocene of China.
53
Unique combination of felid guild and its palaeoecological implications
Among the six species of Felidae at Langebaanweg ‘E’ Quarry, two of them are small to medium felids
(Felis-like and Caracal-like forms; under study) and four are sabertooths. Among them, there are two
Machairodontini (Adeilosmilus aff. kabir,L. chinsamyae sp. nov.), and two metailurine (D. werdelini
sp. nov., Y. obscura). The two Machairodontini species are larger than the two metailurine species,
but Lokotunjailurus is probably only slightly larger than Dinofelis. Even though the postcranial bones
from this locality have not been studied in detail (e.g., Hendey, 1974
12
;Rabeetal.
52
), it is generally
thought that Machairodontini are more adapted to open environments, as most known members of
this tribe have certain cursorial adaptation.
54,89
The known species of metailurine shows no
46
or only
moderate
90
cursorial adaptation. The composition of different tribes of machairodonts in the fauna,
therefore, gives some clues about the paleoenvironment. Overall, the guild of sabertooths from Lange-
baanweg ‘E’ Quarry is composed of typical Miocene Machairodontini (Adeilosmilus,Lokotunjailurus)
and metailurine (Yoshi), which represent one of the last records of these genera in the fossil record,
as well as one of the earliest well-known Pliocene records of Dinofelis. Previously, significant differences
between the carnivoran faunas of the MPPM and LQSM have been highlighted,
12,15,17,20–22,30,33,91
and
interpreted as because of temporal differences and faunal replacement. More specifically, Hendey,
12
Werdelin, and Sardella,
43
pointed out the differential occurrence of Amphimachairodus,Lokotunjailu-
rus and Dinofelis (determined as Homotherium,Amphimachairudus and Machairodus) in LQSM, and
Yoshi (determined as Felis and Metailurus) in MPPM. The updated distribution of these taxa in the
members of ‘E’ Quarry indicate that Lokotunjailurus occurred in LQSM (NMI = 2), Ad. Aff. Kabir and
Dinofelis in both members (NMI of Adeilosmilus = 1 in both members; NMI of Dinofelis =5inboth
members), and Yoshi in MPPM (NMI = 1). The re-study of plenty of new fossils of mustelids, canids
and sabertooths from the site
23–25,52
suggests that the previously observed differences may be pro-
duced by sedimentation (estuarine/marine/fluvial deposition) or sampling biases, instead of temporal
replacement of the carnivoran guild.
23–25
Moreover, this is supported by a same relative dating
of both members (5.15 G0.1 Ma) by sedimentological, petrographical and geochemical
evidence.
10,92,93
We compared the felid guild of Langebaanweg with six other well-established Late Miocene faunas:
Lothagam in eastern Africa,
49
Toros Menalla in central Africa,
50,60,88,94
Samos-PikermiinGreece,
95
Las Casiones in Spain,
96
BaodeinnorthernChina,
66
and Yuanmou in southern China,
97
as well as seven
faunas from the Pliocene and early Pleistocene of Africa: Laetoli bed Upper Unit in Tanzania (3.7–4.46
Ma),
98–100
OMO Shungura in Ethiopia (2.91–2.53 Ma),
44,101
Limeworks Makapansgat in South Africa (3–
2.6 Ma),
102
Malapa in South Africa (1.98Ma),
103,104
Cooper’s Cave System (1–0.6 Ma) in South Africa,
105
Swartkrans Member 1 in South Africa (1.8–1.36 Ma),
106,107
and Koobi for a, Okote member in Kenya (1.5
Ma).
108,109
The body sizes of these felids (which play a key role in their ecology), were classified into six
grades based on cranial and P4 length (Table S3). Our multivariate analyses (hierarchical cluster and
principal component analyses) based on taxonomic composition (five basic measurements: number
ll
OPEN ACCESS
20 iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
of Machairodontini, metailurine, Smilodontini, Felini and Pantherini) and mean body size (averaged for
grade, see Table S3 for details, and the highlight rows are used in cluster and PCA analyses), suggest
that the Late Miocene faunas have an assemblage of felids comprising both Machairodontini and meta-
ilurine. The metailurine are generally more diverse, as they have a smaller body size and occupy a lower
position in the ecosystem. In terms of Machairodontini proportion Langebaanweg has a relatively high
value, only lower than that of Toros Menalla and Yuanmou, suggesting the environment was at least
partially open. The presence of Lokotunjailurus, a relatively cursorial Machairodontini
74
also support
this hypothesis. The number of Dinofelis (metailurine) specimens at Langebaanwegexceeds that
of other felids, suggesting the forest environment is most likely present, revealing a complex environ-
ment with a mosaic of ecosystems in Langebaanweg. Isotopic analyses of the herbivorous fossil fauna
from Langebaanweg
110
suggest a Mediterranean climate, consisting in an open C3-dominated
ecosystem where C4 grasses had not yet extended. Additional evidence from the site such as pollen
and the occurrence of grazing taxa
10,17,111,112
indicate an open subtropical C3 vegetation, with a
weighty contribution by sclerophytic fynbos pointing to a cooler and more seasonal climate than in
the Miocene.
In terms of mean body size grade, it is interesting that among the Mio-Pliocene sites, African felid guilds
generally have a large body size compared with those in Eurasia (Table S3). This is because of its generally
higher Machairodontini proportions, and larger body size (Dinofelis) among metailurines. In general, the
large-bodied carnivorans are more abundant in an open environment, as seen in the modern eastern
African savanna. The body size grade therefore supports that the African environment is generally open
during the Late Miocene.
Comparisons of the evolution of the felid guild from the Miocene-Pliocene to the Plio-Pleistocene in
southern Africa (Figure 13) are intricate due to the c. 3 million years
113
of information hiatus between
Langebaanweg and the Cradle of Humankind fossils (northern part of South Africa), where multiple taxa
of sabertooth cats have been found. The proportionally higher composition of Machairodontini,
including Lokotunjailurus, and larger body size metailurines (D. werdelini) in Langebaanweg,
contrasts with the increase in proportions of both Smilodontini e.g., Megantereon spp., and felines
(Acinonyx,Caracal,Leptailurus,Panthera spp., and Felis) in the Cradle of Humankind fossil record dur-
ing the late Pliocene (e.g., Limeworks Makapansgat, 3–2.6 Ma) and Pleistocene (e.g., Malapa, 1.98 Ma;
Swartkrans Member 1, 1.8–1.36Ma; Cooper’s Cave System, 1–0.6 Ma). Overall, the Pliocene and
Pleistocene African localities are not closely nested with the Mio-Pliocene localities. In contrast, in
PCA plots, the faunas from different ages are clearly located close to each other, and in PC1
from the most negative to the most positives values, the age tends to be younger. This suggests
that the felid guild is more chronological than geographic and supports a high dispersal ability of
this group.
Of interest, Langebaanweg shows the same felid composition to Yuanmou, a middle Late Miocene fauna
from Yunnan Province, southern China. In hierarchical cluster and PCA analyses (Figure 13), Langebaanweg
is close to Yuanmou in both analyses. A recent study of Yuanmou felids
97
reveals that these felids are
unique in many respects, and one new form Longchuansmilus, potentially has a close relationship with
Lokotunjailurus. This suggests that the environment of the two regions could be similar, but also suggests
the presence of a potential dispersal route between these two remote areas, though current evidence is
insufficient to give a more comprehensive analysis.
During the Mio-Pliocene boundary, most sabertoothed cats in Eurasia went extinct, and during the
Pliocene, new forms of machairodonts appeared.
54,114
This is not the case in Africa. Most Pliocene ma-
chairodonts in Africa evolved directly from their late Miocene ancestors, including Dinofelis spp., and
Ad. kabir-Homotherium. As already mentioned, Africa experienced a more continuous change to more
open and arid grassland or desert since the late Miocene which continued into the Pliocene, in contrast
to Eurasia. Our analysis shows that the African late Miocene-early Pliocene felid guild structure already
showed adaptation to open environments compared with Eurasia, where there is no distinct disruption
in environmental change. Eurasia experienced a profound environmental change during this boundary,
and as a result, most saber-toothed cats, together with many other carnivores, e.g., very diversified hy-
ena guild, went extinct. The continuous aridification throughout the Mio-Pliocene, with the spread of
ll
OPEN ACCESS
iScience --, 107212, --, 2023 21
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
open environments, could be an important trigger on the bipedalism of hominids,
115,116
and the saber-
tooth guild in Langebaanweg and its environmental and paleobiogeographic implications provide
background for future discussion on hominid origination and evolution.
Limitations of the study
In this study, we have not compared our specimens with many African sabertooths from other localities,
e.g., Adeilosmilus,Lokotunjailrus, in person, so our coding on these species is based on literature. The
material for some species, especially Adeilosmilus aff. kabir, is too poor to give a good evaluation on its
affinity, so more material will be needed to clarify this large sabertooth cat.
Figure 13. Felid guild composition (species number) and mean body size of selected faunas from Eurasia and
Africa
(A) Hierarchical cluster (Euclidean distance).
(B) PCA scoring plot. Localities in orange are from Africa, in green from Asia and in blue from Europe. Those analyses were
performed using software PAST 4.0.3 (https://past.en.lo4d.com/windows).
ll
OPEN ACCESS
22 iScience --, 107212, --, 2023
Please cite this article in press as: Jiangzuo et al., Langebaanweg’s sabertooth guild reveals an African Pliocene evolutionary hotspot for saber-
tooths (Carnivora; Felidae), iScience (2023), https://doi.org/10.1016/j.isci.2023.107212
iScienc
e
Article
STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:
dKEY RESOURCES TABLE
dRESOURCE AVAILABILITY
BLead contact
BMaterials availability
BData and code availability
dEXPERIMENTAL MODEL AND SUBJECT DETAILS
dMETHOD DETAILS
d