Cyonarctos dessei n.

Bonis, Louis De, 2013, Ursidae (Mammalia, Carnivora) from the Late Oligocene of the “ Phosphorites du Quercy ” (France) and a reappraisal of the genus Cephalogale Geoffroy, 1862, Geodiversitas 35 (4), pp. 787-814 : 790-795

publication ID	https://doi.org/ 10.5252/g2013n4a4
DOI	https://doi.org/10.5281/zenodo.4818332
persistent identifier	https://treatment.plazi.org/id/03AE87A4-700B-FFC7-FCC1-FB74FE35F90D
treatment provided by	Felipe
scientific name	Cyonarctos dessei n.
status	n.

Treatment

Cyonarctos dessei n. sp.

Cephalogale cf. minor – Bonis 1974: 28. Non Cephalogale minor Filhol, 1882 .

Cephalogale minor – Remy et al. 1987: 187. — Biochrom’ 97: 793.

Cephalogale cadurcensis – Remy et al. 1987: 187. Non Canis cadurcensis Filhol, 1877 .

Cephalogale sp. – Remy et al. 1987: 187.

HOLOTYPE. — Left hemi-mandible with p2-m2 ( PD 477 ) from Pech Desse.

PARATYPES. — Left hemi-mandible with broken p2, p3-m3 ( PD 475), left hemi-mandible with c, p2-m2 ( PD 273), fragment of right maxilla with P2-M1, ( PD 130), frag-

Dimension 1; Eigenvalue:,00031 (52,34% of Inertia)

ment of right maxilla with P4-M2 ( PD 507), left M2 ( PD 514), right M1 ( PD 137) and other labelled material from Pech Desse which are identified as the components of the original type series ( ICZN: art. 73 D).

LOCALITY. — Pech Desse near Mouillac, Quercy , France.

REFERRED SPECIMENS. — See material from Pech du Fraysse.

DATING. — MP 28 (about 24.9-24.5 Ma).

DISTRIBUTION. — Quercy (South western France).

ETYMOLOGY. — Dedicated to my colleague Jean Desse who gave his name to the locality.

DIAGNOSIS. — Small,(minimum length of m1 = 14.5 mm), to medium-sized (maximum length of m1 = 18.2 mm) Cephalogalini n. tr. P2-P3 without posterior accessory cusp (pac); P2 slightly asymmetrical; P3 symmetrical; P2-3 with tall and acute crowns; P4 with moderately distally displaced large protocone, moderate cingulum, very small parastyle, short metastyle blade; triangular M1with paracone higher than metacone, V shaped protocone of which the anterior arm joins the mesial cingulum while the posterior one turns distally and thus opening the trigone basin; oval M2 with metacone smaller than or similar to the paracone and close to or separated from the paracone; small one-rooted p1; p1-p3 asymmetrical without posterior accessory cuspid (pacd); p4 more symmetrical with a developed pacd; p2 as tall as p3 and p4 taller than the m1 paraconid; m1 with moderately reduced metaconid and a “ Cephalogale like” talonid; m2 with a small but clear paraconid and a relatively short talonid; small m3 with a clear metaconid and a distinct talonid.

BIOMETRY OF CYONARCTOS DESSEI N. GEN., N. SP.

Ŋe morphological characters of the material from Pech du Fraysse and Pech Desse correspond to those of the tribe Cephalogalini n. tr. It can be divided in each locality in two populations clearly separated by the size but attributed to one species.

Metric analysis of the dental material included within the genus Cyonarctos n. gen. in the two studied localities, and especially m1 measurements, is used to decide if the fossils belong to a single species or not. Although the m1 length would be “the poorest predictor of body weight” ( Van Valkenburgh 1990), it generally gives a good indication on the sexual dimorphism in extant species of carnivorans. A principal components analysis (PCA) based on m1 dimensions ( Fig. 1A View FIG ) shows a group of small specimens clearly separated from a group of larger specimens. Ŋe morphological characters of these groups are quite similar and the problem is to know if there are two size separated species or only one dimorphic species, generally carnivoran males being larger than females with some exceptions. A correspondence analysis (CA) shows that the two groups are mixed ( Fig. 1B View FIG ). Ŋe CA being based on the differences in proportions ( Bonis & Lebeau 1974), it indicates that both groups, small and large individuals, have the same proportions. Ŋe difference in size between the two groups is tested by comparisons with extant and fossil species.

Metric comparisons with extant or fossil taxa were made from published data ( Del Campana 1913; Spahni 1955; Bonifay 1971, 1972; Poplin 1972; Bonifay 1975; Davis 1977; Klein 1986; Berta 1988; Legendre & Roth 1988; Van Valkenburgh 1990; Dayan et al. 1992; Kieser & Kroeneveld 1992a, b; Koufos 1992; Fistani & Crégut-Bonhoure 1993; Okarma & Buchalczyk 1993; Argand 1995; Crégut-Bonnoure 1996; Rook & Torre 1996; Baryshnikov 1996, 2012; Gittlleman & Van Valkenburg 1997; Szuma 2000; Dayan et al. 2002; Palmqvist et al. 2002; Abramov & Puzachenko 2005; Baryshnikov & Tsukala 2010; Castel et al. 2010; Pérez-Rippol et al. 2010; Woldrich 1879) and personal data.

A sexual dimorphism index (SDI 1) can be computed for m1 by the difference between male and female m1 length means × 100 divided by the sum of both means:

(Lm 1a – Lm 1b) × 100

Lm 1a + Lm 1b

Ŋis ratio gives the difference between male and female specimens relatively to the size of m1. It has been computed for 152 extant populations of carnivorans analysed from data in the literature. Ŋe resulting histogram is J shaped ( Fig. 2 View FIG ), the lowest values, (similar of slightly different sex sizes), are close to the vertical axis and the highest values (large difference between both sexes) are far from it. Most of the species indices are situated between the values zero to four (116 species or populations). Ŋe SDI 1 for the two Quercy localities (7.2) is between four and eight values with 26 extant species, eight species having higher indices. Ŋe largest differences may occur in large species as Panthera leo (Linnaeus, 1758) , Panthera tigris (Linnaeus., 1758) or Ursus arctos Linnaeus, 1758 but also in medium sized as Gulo gulo Linnaeus, 1758 or Felis temmincki Vigors & Horsfield, 1827 and small species as Mustela erminea Linnaeus, 1758 . Indices can vary within a same genus and even within the same species depending on the origin of the sample (0.16 to 4.8 for Canis latrans Say, 1823 , 2 to 10 for Ursus arctos , 3.25 to 10.4 for Panthera leo , 1.5 to 5.8 for Nasua nasua Linnaeus, 1766 and 0.3 to 5.1 for Vulpes vulpes (Linnaeus, 1758)) . Ŋe SDI 1 of the Pech Desse-Pech du Fraysse sample is compatible with a single species hypothesis.

Ŋe observed range of the m1 length within a group of specimens corresponds to another index of sexual dimorphism (SDI 2): “(Lm1 of the smallest specimen/Lm1 of the largest specimen) × 100”. Here we compare the group of small specimens (females?) and the group of large specimens (males?) of the Fig. 1 View FIG . In a sample of m1, if the difference in length between the two extreme specimens is low, the ratio will be close to 100, but if the difference is significant, the ratio will be lower. When the ratio of the difference in size of the m1s is computed in extant carnivorans whose sexes are known, we observe that this index, computed in large populations, is higher for each sex than when the two morphs are mixed, indicating the sexual dimorphism ( Table 3 View TABLE ). For example, in Meles meles Linnaeus, 1758 from Transcaucasia, the index is 76 for the males, 88 for the females but only 75 when the two sexes are mixed. Ŋat index reaches 96 for the group of small specimens (14.49/15.05) and 90 for the group of large specimens (16.46/18.22) of Pech Desse and Pech du Fraysse. It reaches only 79.5 when the two groups are mixed (14.49/18.22). Ŋe last value is close to those which are generally found in extant and fossil carnivorans when both sexes are mixed. Even if this index is partially depending on the number of specimens, it gives a quite good idea on the sex difference in a sample of extant or fossil carnivorans. In the case of the studied fossils, we conclude that the numbers are compatible with a single species hypothesis.

Ŋus each index within both groups, small and large, of the Quercy material seems to be too high to belong to two different species but they are convenient with a sexual dimorphism. Ŋus, on these criteria, the whole material from Pech Desse and Pech du Fraysse seems to be within the range of a single species.

DESCRIPTION OF CYONARCTOS DESSEI N. GEN., N. SP. Mandible

Ŋere are four quite complete mandibles from Pech Desse, one from Pech du Fraysse ( Fig.3 View FIG ) and several fragments from Pech Desse and Pech du Fraysse. Ŋe corpus is quite robust (PD 273) or more slender (PD 475, 476, 477, PF 420) but it is always anteriorly tapering. Ŋe ventral border is gently convex in all specimens and the maximum height of the mandible is under m1. Ŋere is one mental foramen beneath the mesial root of p2 and another one beneath p3. Ŋe coronoid process is high with a large masseteric fossa. Ŋe condyle, moderately extended laterally, is situated at the level of the apex of m1. Ŋe symphyseal plate has a defined area of smooth bone above the rugose portion ( Figs 3B View FIG , 4B View FIG ). It is interpreted as corresponding to the presence of a fibrocartilage pad as in the wolf and allowing a quite flexible mandible symphysis and acting as a cushion during the displacements of the two branches of the mandible during chew- ing ( Scapino 1965, 1981; Hunt 2009). Below the smooth plate, there are bony rugosities forming quite horizontal ridges converging forward. Two small foramina are ventrally situated beneath the canine for entry of nerves and blood vessels into this area ( Scapino 1965: fig. 12). Ŋe smooth surface of the symphyseal plate is a horizontal rectangle in the large mandible PD 273 and a smaller elongated oblique oval in PD 477.

Lower dentition ( Table 1 View TABLE ; Figs 3-6 View FIG View FIG View FIG View FIG )

Ŋe incisors are missing in all the specimens but two very small alveoli, corresponding to i2 and i3 are present in PD 315 and PD 477 mesio-buccally of the canine; the former is mesio-distally elongated and narrow, the latter is rounded and just a little enlarged. Ŋere is no place for a first incisor alveolus. Ŋe canine, whose crown is partially broken in PD 273, has a robust root but must have been quite low. Ŋe single rooted p1 is lacking in all the specimens. Ŋe other premolars increase in size from p2 to p4. Ŋe former is low, slightly elongated, without any accessory cusp and with a distal half longer than the mesial one and a faint distal crest from apex to the base. Separated by a short diastema (1 to 4 mm), p3 is quite similar to p2 but the difference in length between mesial and distal halves of the crown is lesser and there is possibly a very minute posterior accessory cuspid (pacd) and a mesial basal cuspid (PD 273); a slight crest runs distally from the apex to the base. Ŋe crown of p4 is more symmetrical, with both mesial and distal basal cuspids and with a well developed posterior accessory cuspid in the distal crest. Ŋe premolars are quite tall, p3 being as tall as the paraconid of m1 and p4 being taller than it. In m1, the protoconid is taller than the paraconid; the metaconid is reduced and posteriorly displaced, clearly separated from the paraconid. Ŋe talonid has an elongated and dominant hypoconid prolonged by a crest which encloses the narrow talonid basin and joins the metaconid without any notch. Ŋe metaconid of m2 is the taller cuspid of this tooth; the paraconid is present and a small mesio-buccal crest encloses the small trigonid basin; the talonid is similar t in shape to that of m1. Ŋe m3 is oval, the metaconid higher than the protoconid and the paraconid is very small, a small hypoconid is present in the buccal crest. Ŋe cingulids are present but weak in premolars and molars.

Upper dentition ( Table 2 View TABLE ; Fig. 7 View FIG )

A piece of maxilla with P3 and part of the mesial alveoli of P4 (PF 46), another one with P3-P4 (PF 37) and a third with P2-M1 (PD 130) are the witnesses of ante-carnassial premolars; their crowns are simple, acute, without accessory cusp, slightly elongated, lower than the paracone of P4 and the mesial portion is shorter than the distal one, especially in P3; there are two small crests, a mesial one from the apex to the mesio-lingual corner and a distal one from the apex to the distal border; a weak cingulum surrounds the base of the crown of P3. Ŋe carnassial is short relatively to the molars with a relatively short and quite robust metastyle; the protocone is a large semi-circular basined platform surrounded by a moderate cingulum and situated quite anteriorly for a hemicyonine; the moderate cingulum is surrounding the whole crown; a mesial crest runs from the apex of the paracone to the cingulum which is turning up as a very small parastyle and separated from the protocone by a shallow notch. In M1, the V shaped protocone is anteriorly prolonged by the preprotocrista to the base of the mesial surface of the paracone and posteriorly by a very tiny, quite indistinct metaconule, the large trigone basin being distally open; the paracone is larger and taller than the metacone and its mesial width is larger than its distal one; a distinct cingulum surrounds the buccal face ending mesially by a small bump corresponding to a tiny parastyle. Ŋe lingual cingulum is developed around the protocone but more distolingually. M2 has an oval occlusal outline; the paracone generally is larger and slightly taller than the metacone; the other characters are similar to those of M1 although the V is more open, its distal crest being almost parallel to the sagittal plane, and the cingulum better spread around the protocone although posteriorly thicker.