Sunteți pe pagina 1din 6

C. R. Acad. Sci.

Paris, Sciences de la Terre et des planètes / Earth and Planetary Sciences 330 (2000) 653–658
© 2000 Académie des sciences / Éditions scientifiques et médicales Elsevier SAS. Tous droits réservés
S1251805000001993/FLA

Palaeontology / Paléontologie
(Vertebrate Palaeontology / Paléontologie des Vertébrés)

A new species of Eofelis (Carnivora: Nimravidae)


from the Phosphorites of Quercy, France
Stéphane Peigné*
Laboratoire de géobiologie, biochronologie, et paléontologie humaine, EP 1596 CNRS, faculté des sciences fondamentales et
appliquées, université de Poitiers, 40, avenue du Recteur-Pineau, 86022 Poitiers, France
Received 20 February 2000; accepted 27 March 2000
Communicated by Yves Coppens

Abstract – For the first time, a great part of the material assigned to the small nimravid
Eofelis is present in a single work. A revised diagnosis is proposed and two species are
distinguished in the old collections from the Phosphorites of Quercy: E. edwardsii which
is the type species and the most abundant in collections; and Eofelis giganteus n.sp., a
very large species since it is about twice as large as E. edwardsii. Unfortunately, the nature
of the old collections from Quercy does not allow us to establish an ancestry-to-
descendant relationship between these taxa. © 2000 Académie des sciences / Éditions
scientifiques et médicales Elsevier SAS

Eofelis / Nimravidae / Carnivora / Paleogene / phosphorites of Quercy / France

Résumé – Une nouvelle espèce d’Eofelis (Carnivora : Nimravidae) dans les Phos-
phorites du Quercy, France. Pour la première fois, une grande partie du matériel attri-
bué au petit Nimravidae Eofelis est rassemblée dans une seule étude. Une diagnose
générique révisée est proposée et deux espèces sont distinguées dans les anciennes
collections des Phosphorites du Quercy : Eofelis edwardsii, l’espèce type, qui est la plus
abondante dans les collections, et Eofelis giganteus n.sp., une espèce de très grande
taille, puisqu’elle fait environ deux fois la taille d’E. edwardsii. Malheureusement, la
nature des anciennes collections du Quercy ne permet pas d’établir une relation d’ancê-
tre à descendant entre ces deux espèces. © 2000 Académie des sciences / Éditions
scientifiques et médicales Elsevier SAS
Eofelis / Nimravidae / Carnivora / Paléogène / phosphorites du Quercy / France

Version abrégée noter la présence du genre [2, 3]. Ginsburg [5] a, le


premier, conclu qu’une seule espèce, E. edwardsii, était
1. Introduction présente dans le Quercy. L’examen d’une grande partie
du matériel attribué au genre, incluant de nombreux
Le genre européen Eofelis appartient aux Nimravinae spécimens inédits, permet de proposer une diagnose
(Carnivora : Nimravidae), répartis dans l’hémisphère révisée d’Eofelis. L’étude présentée ici permet aussi de
nord, de l’Éocène supérieur jusqu’au milieu de l’Oligo- distinguer un spécimen de forte taille présentant les
cène. L’espèce type, E. edwardsii, créée pour une man- caractéristiques du genre et de créer une nouvelle espèce,
dibule provenant du Quercy [4], a d’abord été rappor- E. giganteus.
tée au genre Pseudaelurus. Elle fut ensuite placée dans
les genres Aelurogale [12], puis Nimravus [11]. Le genre 2. Paléontologie systématique
Eofelis a été créé par Kretzoï [7] pour distinguer le petit
Nimravidae du Quercy de ces taxons, avec lesquels il a Diagnose révisée du genre Eofelis (liste du matériel
été si souvent confondu. Au cours des 30 dernières plus loin) : Nimravidae dont la taille est comprise entre
années, seul le gisement de Villebramar a permis de celle de Lynx rufus et celle de Panthera pardus ; carac-

* Correspondence and reprints: geo.bio@campus.univ-poitiers.fr

653
S. Peigné / C. R. Acad. Sci. Paris, Sciences de la Terre et des planètes / Earth and Planetary Sciences 330 (2000) 653–658

tères machairodontes peu marqués ; P1-2 de petites nettement supérieure à celle du plus grand des spéci-
tailles ; P4/p4 plus hautes que P3/p3 ; pas d’apophyse mens d’Eofelis connus jusqu’alors (figure 2, tableau). La
mentonnière ; P3 triradiculée ; P4 sans parastyle, avec morphologie des dents et de la mandibule est assez
un protocône distinct situé dans l’angle mésio-lingual ; proche de ce qui est connu chez E. edwardsii.
M1 allongée transversalement, avec un protocône lin-
gual développé et nettement séparé du paracône ; cus- 3. Discussion et conclusion
pide accessoire distale des prémolaires inférieures plus Eofelis se rapproche du genre européen Dinailurictis
développée que la cuspide mésiale (quand elle est pré- et des genres américains Dinictis et Pogonodon. Une
sente), contrairement à ce que l’on observe chez Nimra- étude préliminaire [9] place Eofelis aux côtés de ces
vus ; m1 plus large que celle de Nimravus, dépourvue genres, au sein d’un clade incluant également Dinailu-
de métaconide ; talonide de m1 court (entre 15 et 23 % rictis et Quercylurus, des genres européens. Les carac-
de la longueur totale de m1), plus large et moins cou- tères liés à la machairodontie sont peu prononcés chez
pant que celui des m1 rapportées à Nimravus ; proto- ce petit Nimravidae. Il possède néanmoins des canines
conide et paraconide distincts sur m2. supérieures crénelées et compressées transversalement,
Eofelis edwardsii est une espèce bien connue, bien une denture coupante avec des prémolaires antérieures
qu’aucun crâne n’ait jamais été attribué à cette espèce. et des molaires postérieures réduites, et un menton carré,
La mandibule est forte, avec un menton carré (figure 1). mais dépourvu de toute apophyse mentonnière. La
De la denture supérieure, nous connaissons la canine, réduction des P3/p3 par rapport aux P4/p4, un denti-
peu aplatie latéralement, crénelée sur son bord distal. cule accessoire distal de taille supérieure au denticule
Le bourrelet mésio-lingual ne porte pas de crénulations mésial sur p3–p4, sont quelques-uns des caractères qui
visibles. P1 et P2 sont réduites, mais la seconde est distinguent Eofelis de Nimravus. Exclusivement euro-
biradiculée et porte un simple tranchant. Un petit dias- péen jusqu’à présent, le genre pourrait être originaire
tème sépare P2 de P3, qui est une dent de taille nette- d’Asie. En effet, un fragment de mandibule portant m1
ment plus importante puisqu’elle atteint presque la hau- (AMNH 21674) et provenant de Mongolie présente des
teur du paracône de P4. De plus, P3 porte un tubercule caractères qui rappellent l’espèce européenne. Ce spé-
accessoire distal coupant et allongé. Le plus remarqua- cimen a pourtant été attribué à Proailurus, un Felidae
ble est la présence d’une troisième racine, linguale, [6, 8]. Cependant, la carnassière est dépourvue de méta-
comme chez Dinailurictis. P4 est dépourvue de paras- conide, une cuspide toujours présente chez Proailurus,
tyle et porte un protocône développé, projeté loin dans et possède un talonide peu coupant, dont les propor-
l’angle mésio-lingual. Un seul spécimen (UM-ACQ 5261) tions et la taille rapprochent davantage cette pièce de
conserve une M1 complète. La présence d’un protocône ce que l’on connaît chez Eofelis.
en position très linguale rapproche Eofelis des genres Malgré une denture semblable, l’holotype d’E. gigan-
Dinictis et Dinailurictis. La denture inférieure est mieux teus est d’une taille beaucoup plus importante que celle
représentée dans les collections. Les prémolaires anté- des plus gros spécimens attribués à l’espèce type du
rieures sont très réduites. p1 est absente sur certains genre. Une distinction spécifique reste l’hypothèse la
spécimens. p2 est une dent biradiculée de petite taille. plus probable pour rendre compte de la réalité. Mal-
p3 est nettement plus basse que p4. Toutes deux por- heureusement, la nature des collections dont les spéci-
tent des denticules accessoires dont le distal est le plus mens sont issus ne permet pas de se prononcer sur les
développé. p4 est souvent plus haute que le paraco- relations qui lient ces deux espèces et sur leur réparti-
nide de m1. Celle-ci est une dent plus large que chez tion stratigraphique.
Nimravus. Comme chez ce dernier genre, le métaco- Cette étude permet de faire le point sur un genre
nide est absent. En revanche, le talonide est moins tran- souvent ignoré dans les révisions systématiques qui ont
chant, plus large que ce que l’on trouve chez Nimravus déjà été proposées pour les Nimravidae du Paléogène.
et il est relativement long (environ 18 % de la longueur Pourtant, Eofelis est un genre parfaitement valide. De
totale de m1). Quand elle est présente, m2 est une dent plus, la mise en évidence d’une nouvelle espèce sug-
uniradiculée au simple tranchant. gère que l’étude des Nimravidae européens, en cours
La nouvelle espèce E. giganteus est représentée par actuellement, permettra certainement d’avoir une
un seul spécimen, un fragment de mandibule portant meilleure idée de la diversité de ces prédateurs et du
p2, p4–m1 (figure 3). Elle se différencie par une taille rôle qu’ils ont joué dans les écosystèmes du Paléogène.

1. Introduction species, E. edwardsii, was created for a mandible from


Quercy which was initially assigned to the felid genus
The European genus Eofelis belongs to the Nimravi- Pseudaelurus [4]. The species was successively assigned
nae (Carnivora: Nimravidae), a subfamily of impressive to Aelurogale (e.g. [12]), Nimravus (e.g. [11]), and finally
sabre-tooth predators which lived in northern continents to the new genus Eofelis [7], which is still valid. Indeed,
from Late Eocene to Mid-Oligocene. Originally, the type throughout its history the type species E. edwardsii has

654
S. Peigné / C. R. Acad. Sci. Paris, Sciences de la Terre et des planètes / Earth and Planetary Sciences 330 (2000) 653–658

often been assigned to Nimravus (or to its junior syn- the total length), and broader than in Nimravus, hence
onym Aelurogale [10]). Ginsburg [5] clarified the sys- less trenchant. m2 usually present, with distinct proto-
tematics of Eofelis and judiciously concluded that a single conid and paraconid cusps as in Dinictis, the most abun-
species, Eofelis edwardsii, was present in Quercy. Dur- dant genus in North American collections.
ing the recent field campaigns, the species has been Distribution. Unknown, from the Quercy district,
noticed only in Villebramar [2, 3]. Many features differ- France.
entiating this genus lead us to review the diagnosis of Eofelis edwardsii (Filhol, 1872)
Eofelis in this contribution. Holotype. MNHN-QU 9539; left mandible with c,
Numerous specimens are assigned to Eofelis, which is p3–m1 (figure 1).
one of the most represented nimravid genera in Europe. Type locality. Unknown, phosphorites of Quercy,
A great part of these has been studied in this work, includ- France.
ing many unpublished specimens. This allows us to give Other referred material (see figure 2 and table for mea-
a precise diagnosis of the genus Eofelis and to review surements). Old collections from Quercy: MNHN-QU
that of Ginsburg [5]. A new species, E. giganteus, is 9420 (fragmentary left maxillary with P1, P3–P4),
created in this note, based on a very large size mandible MNHN-QU 9501 (fragmentary left mandible with c,
which was not previously mentioned. The holotype and p4–m1), MNHN-QU 9502 (fragmentary right mandible
single specimen comes from the old collections of Quercy with m1), MNHN-QU 9503 (fragmentary right mandible
which are housed in the Museum of Natural History of with p3–m1), MNHN-QU 9504 (fragmentary left man-
Basel. dible with broken m1), MNHN-QU 9505 (left mandible
Institutional abbreviations: AMNH, Department of Ver- with p3–m1), MNHN-QU 9506 (fragmentary right man-
tebrate Paleontology, American Museum of Natural His- dible with c, p2–m1), MNHN-QU 9507 (fragmentary
tory, New York; BMNH, British Museum of Natural His- right mandible with p4-m1), MNHN-QU 9508 (fragmen-
tory, London; FSL, faculté des sciences, université de tary left mandible with c, p3–p4), MNHN-QU 9509 (frag-
Lyon; MCZ, Museum of Comparative Zoology, Harvard; mentary left mandible with p3–m1), MNHN-QU 9510
MHNG, Museum d’histoire naturelle de Gaillac; MHNT- (fragmentary left mandible with c, p2–m1), MNHN-QU
PHQ, collections from Quercy, Museum d’histoire 9512 (fragmentary left mandible with m1), MNHN-QU
naturelle de Toulouse; MNHN-QU, collections from 9513 (fragmentary right mandible with p2–m2),
Quercy, Museum national d’histoire naturelle, Paris; MNHN-QU 9514 (fragmentary left maxillary with C,
MA-PHQ, collections from Quercy, Museum d’histoire P2-M1), MNHN-QU 9540 (fragmentary left mandible
naturelle de Montauban; NMB, collections from Quercy, with p4-m1). AMNH 55562 (cast of MCZ 8939, frag-
Naturhistorische Museum, Basel; PLV, Institut catholique
de Leuven; SMF, Forschungsinstitut und Naturmuseum
Senckenberg, Francfort; UM-ACQ, collections from
Quercy, Université de Montpellier.

2. Systematic palaeontology

Feliformia Kretzoi, 1945


Nimravidae Cope, 1880
Nimravinae Cope, 1880
Eofelis Kretzoi, 1938
Type species. Eofelis edwardsii (Filhol, 1872).
Included species. Eofelis edwardsii, Eofelis giganteus
n.sp.
Revised diagnosis. Size from that of Lynx rufus to that
of Panthera pardus, little developed machairodont fea-
tures. Premolar dentition frequently complete. Very small,
probably rarely used P1–2. P4/p4 clearly taller than
P3/p3. Robust mandible lacking flange. Three-rooted P3.
No parastyle but well defined protocone on P4, which
is projected farther into the mesiolingual corner. Trans-
Figure 1. Eofelis edwardsii, MNHN-QU 9539. Holotype, left man-
versely elongated M1 with a developed lingual proto-
dible with c, p3–m1. Scale = 1 cm. a: labial view; b: lingual view;
cone. No M2. Distal accessory cusp larger than the mesial c: occlusal view.
one on lower premolars, in contrast to the condition in Figure 1. Eofelis edwardsii, MNHN-QU 9539. Holotype, mandi-
Nimravus. m1 wider than in Nimravus, and without meta- bule gauche avec c, p3–m1. Échelle = 1 cm. a : vue labiale ; b :
conid. Short talonid on m1 (between 15 and 23 % of vue linguale ; c : vue occlusale.

655
S. Peigné / C. R. Acad. Sci. Paris, Sciences de la Terre et des planètes / Earth and Planetary Sciences 330 (2000) 653–658

Figure 2. Bivariate plot of the measurements of m1 assigned to


Eofelis (in mm).
Figure 2. Diagramme bivarié des mesures des m1 attribuées à
Eofelis (en mm).

mentary right mandible with p2–m1). BMNH 1358 (frag- Mandible. The mandible shape is similar to that of
mentary left mandible with c, p2-p4), BMNH 2376 (frag- Dinictis, except that the European genus lacks the flange.
mentary left mandible with p3–m1), BMNH 2377 In lateral view, the wide mandible has a nearly straight
(fragmentary right mandible with m1), BMNH 7431 (frag- ventral outline. It lacks the anteroventral fossa for the
mentary left mandible with m1), BMNH 7489 (left man- digastric muscle which is visible on specimens assigned
dible with p4–m1), BMNH 9633 (right mandible with to Nimravus. The mandibular chin is clearly angular.
p2–m1). FSL 7285 (fragmentary right mandible with There is a varying number of mental foramina, but usu-
p4–m1), FSL 7286 (fragmentary right mandible with c, ally no more than two are present. The coronoid pro-
p3–m1), FSL 7287 (right P4). MA-PHQ without collec- cess is similar in development to that of extant felids
tion number (fragmentary left mandible with i3, p1–m1, though it is less backwardly oriented. The short, robust
fragmentary left mandible with p4-m1, fragmentary right angular process has a large ventral crest which defines
mandible with p3–m1). MHNG without collection num- the fossa for the medial pterygoideus muscle.
ber (fragmentary left mandible with p4–m1). MHNT- Dentition. Upper incisors are not preserved. The
PHQ 225 (fragmentary right mandible with m1), MHNT- canine is not very transversely flattened (length/width
PHQ 233 (fragmentary right mandible with p3–m1),
MHNT-PHQ 234 (fragmentary left mandible with p3–m1), Table. Comparison between measurements of the type specimen
MHNT-PHQ 235 (right mandible with c, p3–m1), MHNT- of E. giganteus n.sp. and those of specimens assigned to E. edward-
PHQ 359 (fragmentary right mandible with p3–m1. sii (in mm). L = length, W = width, LT = length of m1 trigonid, N =
NMB-QB 420 (fragmentary right mandible with p4–m1), sample size, min. = minimum, max. = maximum, SD = standard
NMB-QB 433 (fragmentary left mandible with p4–m1), deviation, nc = not conserved.
Tableau. Comparaison entre les mesures de l’holotype d’E. gigan-
NMB-QB 460 (fragmentary right mandible with p3–m1),
teus n.sp. et les mesures des specimens attribués à E. edwardsii
NMB-QB 530 (right m1), NMB-QB 725 (right m1), (en mm). L = longueur, W = largeur, LT = longueur du trigonide
NMB-QB 719 (right m1), NMB-QB 767 (fragmentary left de m1, N = nombre de spécimens, min. = minimum, max. = maxi-
mandible with p3–m1), NMB-QC 306 (fragmentary right mum, SD = écart type, nc = non conservé.
mandible with p4–m1), NMB-QH 646 (fragmentary right
Eofelis giganteus Eofelis edwardsii
mandible with p2, p4), NMB-QJ 235 (fragmentary right
mandible with p3–m1), NMB-QU 324 (fragmentary right N mean min. max. SD
mandible with p4-m1), NMB-QU 338 (fragmentary left Lc1 nc 7 5.90 5.50 6.80 0.44
maxillary with DP3-DP4), NMB-QW 41 (fragmentary left Wc1 nc 7 4.06 3.60 4.70 0.37
mandible with p3–m1). PLV 198 (fragmentary left man- Lp1 nc 2 2.05 1.70 2.40 0.49
Wp1 nc 2 1.25 1.20 1.30 0.07
dible with p1-m2). SMF 5578 (left C). UM-ACQ 5249 Lp2 6.90 7 3.31 2.40 3.70 0.42
(fragmentary right mandible with p4–m1), UM-ACQ 5259 Wp2 3.10 7 1.93 1.30 2.20 0.30
(fragmentary left mandible with p4–m1), UM-ACQ 5261 Lp3 nc 20 7.35 5.60 9.20 1.04
(fragmentary left maxillary with P4–M1). Wp3 nc 19 3.41 2.70 4.10 0.42
Revised diagnosis. Species which differs from E. gigan- Lp4 13.70 33 9.71 8.10 12.00 1.09
Wp4 6.60 33 4.31 3.40 5.50 0.55
teus n.sp. by a much smaller size, from that of Lynx
Lm1 17.00 36 12.19 10.50 14.50 1.17
rufus to that of a small Neofelis nebulosa. Wm1 7.20 38 4.86 4.00 6.00 0.63
Distribution. Unknown localities in Quercy; ?Villebra- LTm1 14.10 31 10.00 8.60 12.00 0.64
mar, France, MP22 (European mammal biozone, Early % trigo 83 31 81.20 77.20 85.50 2.10
Oligocene). L/Wm1 2.36 35 2.52 2.23 2.82 0.15
Lm2 ? 2 2.15 1.70 2.60 0.35
Description (figure 1). No skull can be referred to Eofe-
Wm2 ? 2 1.85 1.60 2.10 0.90
lis.

656
S. Peigné / C. R. Acad. Sci. Paris, Sciences de la Terre et des planètes / Earth and Planetary Sciences 330 (2000) 653–658

ratio = 1.5) that is correlated to the plesiomorphic con-


dition of the sabertooth features in Eofelis. The distal rim
of the canine is nearly vertical and clearly serrated. A
prominent and serrated ridge is present on the mesiolin-
gual face. When present, P1 is very reduced and single-
rooted. P2 is biradiculated; this is a small but trenchant,
asymmetrical tooth with a small distal cusp. There is a
short diastema between P2 and P3, like in Nimravus
and Dinictis. P3 is a much larger tooth; its main cusp is
a little lower than the paracone of P4. There is no mesial
cusp whereas the distal one is trenchant, elongated, asso-
ciated with a basal cingulum. The lingual third root of
P3 might be partly fused to the labial root. P4 is similar
to that of Dinictis regarding its proportion, the lack of
parastyle and the well-developed protocone which is
mesiolingually projected. A single specimen (UM-ACQ
5261) preserved a complete M1. It has a well devel-
oped, lingual protocone which is distinctly separated from
the paracone, that is similar to the condition in Dinictis
and Pogonodon from North America. Figure 3. Eofelis giganteus n. sp., NMB QB 799. Holotype, frag-
Lower incisors are not preserved in the material. The ment of left mandible with p2, p4–m1. Scale = 1 cm. a: lingual
canine is similar to that of Dinictis. It has a planar lin- view; b: labial view; c: occlusal view.
gual face with two lingual and distolingual prominent Figure 3. Eofelis giganteus n. sp., NMB QB 799. Holotype, frag-
ment de mandibule gauche avec p2, p4–m1. Échelle = 1 cm. a :
ridges. p1 is absent on MNHN-QU 9539, MNHN-QU
vue linguale ; b : vue labiale ; c : vue occlusale.
9504, FSL 7286 but it is present on the left side only of
FSL 7285. When preserved (PLV 198), p1 is single-
rooted, tiny and probably non-functional. p2 is always
Mandible. In spite of its great size, the morphology of
present but rarely preserved. This is usually a double-
the mandible is similar to that of E. edwardsii. The speci-
rooted and very reduced tooth. p3 is lower than p4; its
men lacks the anterior and posteriormost parts. A single
mesial cingular cusp, if any, is smaller than the distal
mental foramen is present and level with p2.
one. The latter is wide, trenchant, separated from the
Dentition. The canine and the single-rooted p1 are
main cusp by a shallow notch. p4 is taller than the para-
not preserved. p2 is an asymmetrical tooth and has a
conid of m1 and, in lateral view, the distal rim of the
long, low distal cingulum. p3 is not preserved on the
main cusp is slightly convex. The mesial cingular cusp
specimen. The top of p4 is broken but the crown is
is trenchant, lingually oriented whereas the distal cusp
likely to reach the height of m1. The mesial cingular
is larger, with a slight cingulum. The lower carnassial is
cusp is small whereas the distal one, much larger, is
more robust, wider than that of Nimravus. It lacks the
trenchant, associated with a developed distolingual cin-
metaconid in contrast to the condition found in Dinictis.
gulum. m1 is a wide tooth, similar to that of E. edward-
The trigonid is trenchant and its distal border is nearly
sii. The distal face of the tall, sharp protoconid is slightly
vertical. The talonid is not a single, narrow trenchant as
oblique in lateral view. The talonid length is similar to
in Nimravus, but it is composed of a wide, sometimes
that of E. edwardsii. It is separated from the protoconid
prominent hypoconid crest, associated with a distal cin-
by a small notch and composed by a short cutting crest,
gulum. The trigonid is shorter than in any nimravid, i.e.,
associated with a cingulum which is distolingually dis-
about 82 % of the total length (see table). m2, if present,
tinct. The preservation of the mandibule prevents us from
is single-rooted; the protoconid and paraconid may be
determining the presence of m2.
distinguished on its mere crown.
Eofelis giganteus n.sp.
Holotype: NMB-QB 799; fragment of left mandible 3. Discussion
with p2, p4–m1 (figure 3).
Type locality: unknown, phosphorites of Quercy, Eofelis is well represented in the ‘old collections’ from
France. Quercy. It shows many similarities with genera such as
Referred material: only the holotype. Dinailurictis from Europe, Pogonodon and Dinictis from
Diagnosis: very large species of Eofelis, similar in size North America. These relationships have been con-
to a small Panthera pardus (∼ 40 kg). See table and fig- firmed by a preliminary analysis [9] of the Paleogene
ure 2 for a comparison with E. edwardsii. nimravid genera. This work distinguishes Dinaelurus and
Etymology: from the latin ‘giganteus’, very large. Nimravus from other genera. Among the latter, a clade
Distribution: unknown locality, Quercy district, France. involves the European genera Eofelis, Dinailurictis and
Description (figure 3). Quercylurus. Eofelis is one of the least sabertooth nim-

657
S. Peigné / C. R. Acad. Sci. Paris, Sciences de la Terre et des planètes / Earth and Planetary Sciences 330 (2000) 653–658

ravine even though it is clearly trending toward this adap- described or even recovered, since the fossils from
tation, in having a transversely compressed, serrated Quercy were widely spread throughout the world.
upper canine, a trenchant dentition in which the ante- Despite a similar dental and mandibular morphology,
rior premolars and the posterior molars are reduced, and the great difference in size between the largest speci-
an angular chin on the mandible. It has also some derived mens assigned to E. edwardsii and the holotype of E.
features which make it different from Nimravus (and giganteus suggests a probable specific distinction. Unfor-
Dinaelurus from North America): the reduction of P3/p3 tunately, the unknown stratigraphical origin of speci-
vs P4/p4, a great size of the distal accessory cusp on mens prevents a discussion about their relationships.
p3-p4 which are synapomorphies acquired by other nim-
ravines. Up to now, the genus was only noticed in Europe.
However, an Asian specimen from Mongolia housed in 4. Conclusions
New York (AMNH 21674) might be close to Eofelis. This
is a fragmentary right mandible with m1 and alveolus of
m2, which has been assigned to Proailurus [6, 8]. How- The genus Eofelis was almost never mentioned in North
ever, this specimen lacks metaconid on m1, a cuspid American studies about Nimravidae but this paper con-
which is always present in the European Proailurus. This firms the validity of this genus and its distinction from
individual (length of m1 = 10.2 mm) is much smaller than Nimravus. Study of many specimens assigned to Eofelis
the Proailurus specimens from Oligocene (Quercy and edwardsii allows us to review the generic diagnosis, dem-
Coderet), and similar in size to the smallest specimen onstrates the specific variation of this species and con-
from the Miocene St-Gérand Basin, Allier [1]. Most firms the synonymy given by Ginsburg [5] who stated
importantly, the morphology of the Mongolian specimen that Aelurogale acutata, Aelurogale minor and Eofelis
is similar to that of Eofelis, in which m1 has no meta- edwardsii correspond to a similar form of which the lat-
conid and a little trenchant talonid. The m1 trigonid of ter is the valid name. A new species is created in this
AMNH 21674 is about 80.4 % of the total length of the contribution. Although the material assigned to this spe-
tooth that is similar to its proportion in Eofelis (see table) cies is fragmentary, a specific distinction is likely. This
but distinctly shorter than the condition known in Proailu- suggests that study of European nimravids, which is cur-
rus [1]. rently managed by the author, will contribute to a better
The distinction of a new species in the collections is knowledge of the diversity and the role of the Nim-
not surprising. Many specimens discovered during the ravidae in the European ecosystem during the Paleo-
two last decades of the 19th century remain to be gene.

Acknowledgements. Many thanks to all those who gave me access to the collections under their care. I also thank the anonymous reviewer
for valuable criticism of this manuscript. I am grateful to S. Riffaud who made illustrations of specimens and to Prof. L. de Bonis for helpful
corrections of my work.

References [6] Hunt R.M. Jr., Evolution of the Aeluroid Carnivora: diversity of
the earliest Aeluroids from Eurasia (Quercy, Hsanda-Gol) and the origin
of Felids, Am. Mus. Novit. 3252 (1998) 1–65.
[1] Bonis L. de, Peigné S., Hugueney M., Carnivores féloïdes de l’Oli- [7] Kretzoi M., Die Raubtiere von Gombaszög nebst einer Übersicht
gocène supérieur de Coderet-Bransat (Allier, France), Bull. Soc. géol. der Gesamtfauna, Ann. Mus. Nat. Hung. Budapest 31 (1938) 88–157.
France 170 (1999) 939–949. [8] Mellett J.S., The Oligocene Hsanda Gol Formation, Mongolia. A
[2] Brunet M., Les Mammifères et le problème de la limite Éocè- revised fauna list, Am. Mus. Novit. 2318 (1968) 1–16.
ne–Oligocène en Europe, Geobios, mém. sp. 1 (1977) 11–27. [9] Peigné S., A first systematic review of the Paleogene Nimravidae
[3] Brunet M., Les grands Mammifères chefs de file de l’immigration (Carnivora, Mammalia), J. Vertebrate Paleontol. 19 (Suppl. 3) (1999)
oligocène et le problème de la limite Éocène–Oligocène en Europe, 68A.
Fondation Singer-Polignac, Paris, 1979. [10] Peigné S., Bonis L. de, Le premier crâne de Nimravus (Mamma-
[4] Filhol H., Sur les Carnassiers et les Chéiroptères dont on trouve lia, Carnivora) d’Eurasie et ses relations avec N. brachyops d’Amérique
les débris fossiles dans les gisements de phosphorites de Caylux, Fre- du Nord, Rev. Paléobiol. Genève 18 (1999) 57–67.
gols, Concots, C. R. Acad. Sci. Paris 75 (1872) 92–94. [11] Piveteau J., Les chats des phosphorites du Quercy, Ann. Paléon-
[5] Ginsburg L., Révision taxonomique des Nimravini (Carnivora Feli- tol. Paris 20 (1931) 107–163.
dae) de l’Oligocène des Phosphorites du Quercy, Bull. Mus. Natl. Hist. [12] Teilhard de Chardin P., Les carnassiers des phosphorites du
Nat. Paris, 4e série 1 (1979) 35–49. Quercy, Ann. Paléontol. Paris 9 (1915) 103–192.

658

S-ar putea să vă placă și