Valentinella vitrollense, Tabuce & Vianey-Liaud & Garcia, 2004

Tabuce, Rodolphe, Vianey-Liaud, Monique & Garcia, Géraldine, 2004, A eutherian mammal in the latest Cretaceous of Vitrolles, southern France, Acta Palaeontologica Polonica 49 (3), pp. 347-356 : 348-352

Valentinella vitrollense sp. nov.

Figs. 2 , 3.

Holotype: ISEM /VLP−2, damaged right dentary with p3–m3.

Referred material.—ISEM/VLP−4: fissured fragment of a right dentary with a damaged?p4, associated with two very fragmentary adjacent teeth. ISEM/VLP−3: a fractured maxillary fragment with roots of right and left canines and a right?P2.

Diagnosis.— Valentinella presents the classic “zhelestid” features: the teeth are characterized by developed crushing function, the paraconid is lingually or sublingually positioned with a clear appression to the metaconid, the entoconid and the hypoconulid are twinned, and the talonid is expanded labiolingually. Valentinella differs from the Asian “zhelestids” by the molarization of the ultimate premolar, and is distinct from Avitotherium and Gallolestes by its simple, bulbous p3 and by its larger metaconid and entoconid on lower molars, respectively. Valentinella differs from Labes and Lainodon in having a mesio−distally compressed trigonid.

Etymology.—From Vitrolles, name of the town near the fossiliferous site.

Measurements.—(in mm): ISEM/VLP−2: p3 (L = 2.3; W = 1.5); p4 (or dp4) (L = 2.3; W = 2); m1 (L = 3.4; W = 2.35; Ltri = 2; Ltal = 1.4; Wtri = 2; Wtal = 2.35); m2 (L>3.5; W = 2.9); m3 (L = 3.1; W>2). ISEM/VLP−4:?p4 (Ltri = 1.5; Ltal = 1.2; Wtri = 1.35; Wtal = 1.9). ISEM/VLP−3: right canine (H>2.5; diameter of the root = 1.7); left canine (H>2.3; diameter of the root = 1.6);?P2 (L = 2.5; W = 2.0)

Description.—During diagenesis and compaction of the sediment, the teeth were distorted along parallel cracks caused by plant roots, and/or between talonids and trigonids, and between teeth. These movements have produced oblique deformations.

On ISEM/VLP−2 ( Fig. 2 ), the dentary is almost an outline of a jaw fissured in the marl, while both dentine and enamel are rather better conserved. The five teeth exhibit two robust roots.

The occlusal outline of the p3 is oval. This is the only well−preserved tooth. The crown has a single, large, bulbous cusp (protoconid). Distally there is a small, low cingulid and a distinct concavity between the disto−lingual flank of the protoconid and the distal heel. On the distal side of the protoconid, although this cusp is damaged, it appears that no crest is present; the distal side appears swollen.

The p4 (or dp4?) is smaller than the molars, and the wear pattern suggests that it is either a molarized premolar or a deciduous premolar; according to its small mesio−distal dimension, we favor a p4 status for this tooth. Taking into account the crushing, the only morphological traits of the crown that can be described are the labially rounded sides of both protoconid and hypoconid. The labial position of the protoconid suggests the presence of a distinct metaconid.

On m1, we can distinguish the protoconid and the hypoconid. The trigonid is badly damaged lingually. The cusps show considerable dental attrition; dentine is well exposed in both the distal part of the trigonid and the whole talonid. The worn occlusal surfaces are flat and the trigonid and talonid are of the same height. On the labial side, an incipient exaenodonty can be seen. The talonid appears to be shorter and wider than the trigonid. The quadrate shape of the mesial part of the crown suggests that the paraconid—if it is present—is not mesially projected but strongly appressed to the metaconid. This trait, which is also found in m2–m3, will need to be verified on unworn specimens.

The m2 is longer and wider than the m1. Its trigonid is very worn and damaged. On the talonid, three cusps can be identified with some difficulty. The hypoconid is not larger than the entoconid; and according to their rounded bases these cusps were likely more bulbous than sharp. A fragment of the hypoconulid is preserved disto−labially to the entoconid and is positioned near it (“twinning”).

The m3 is shorter and appears to be narrower than the two other molars. Only the labial half of the crown is not crushed. The tooth is strongly worn, suggesting, like the occlusal surfaces observed in m1–2, a developed crushing function of the teeth during dental attrition. On the talonid, which is shorter than the trigonid, the remaining area of the postcristid does not suggest a strong distal development of a hypoconulid.

ISEM/VLP−4 is a very damaged specimen which displays fragments of both?p3 and?m1, plus a better preserved?p4. Only the lingual wall of the?p3 is preserved, the length of this tooth being superior to 1.5 mm, as roughly estimated from the two roots. The mesio−distal axis of the?p4 shows a 90° labial rotation relative to the?p3. The tooth is probably two−rooted, although the roots are not well−preserved on the specimen. The crown appears to be high (about 2 mm, on the disto−lingual side), but because of poor enamel preservation, it is difficult to identify the position of the neck. On the occlusal surface, the morphology is badly preserved, the talonid seems shorter and wider than the trigonid. As for the posterior tooth, the?m1, only fragments of enamel remain.

On ISEM/VLP−3 ( Fig. 3), a fragment of bone bears two parallel and elongated vestiges of dentine and enamel. We tentatively interpret these dental remains as right and left upper canines. Distally, some fragments of the maxilla are preserved; they are followed by a two−rooted tooth, that could be P2. This unicuspid upper premolar is the best preserved tooth of the material studied here. The paracone is centrally positioned and laterally narrow; this cusp exhibits a sharp mesial crest, is slightly expanded on its lingual margin, but does not bear an incipient protocone. The mesial root strongly slants distally before joining the crown.

Comparisons and discussion.—The large and three−cusped talonid of Valentinella suggests a therian, and the entoconid−hypoconulid twinning is evocative of the metatherians. Considering the dental formula of the lower jaw ISEM/VLP−2, however, with only three molars bearing a reduced paraconid, we conclude that Valentinella is certainly a eutherian. Some Late Cretaceous eutherians resemble our specimens in their wear pattern; it is particularly true for Gypsonictops (see Lillegraven 1969: 52), which also exhibits a reduced paraconid. This genus clearly differs from Valentinella in having a very reduced p2, a premolarifom p3 with a bicuspidate talonid, a true molariform p4 and by a non−reduced m3. Moreover, as in most Late Cretaceous eutherian mammals, Gypsonictops has a more centrally positioned hypoconulid on the postcristid. According to Setoguchi et al. (1999), the entoconid−hypoconulid twinning (observed in Valentinella ) is typical of the families Zalambdalestidae and “ Zhelestidae ”. In the zalambdalestids, Archibald and Averianov (2003) argued however that, at least in Kulbeckia , the entoconid and hypoconulid are very close to each other, but not twinned.

Within the zalambdalestids, the Mongolian Zalambdalestes and Barunlestes ( Kielan−Jaworowska 1969; Kielan−Jaworowska and Trofimov 1980) also share with Valentinella a reduced m3; but they differ by a vestigial P2, a longer p4, and less swollen cusps on molars. Kulbeckia , the oldest known zalambdalestid, is more evocative of our specimens in the short molariform p4 and by the sub−rectangular occlusal pattern of the teeth. Kulbeckia , however, has a long m3 and a very small P2.

According to the diagnosis of the Ungulatomorpha (“zhelestids” plus ungulates) ( Archibald 1996; Nessov et al. 1998), Valentinella is more reminiscent of this supergrandorder by the lingual or sublingual position of the paraconid with a clear appression to the metaconid, the entoconid and the hypoconulid twinned, the expanded talonid, and by the indication of a quite similar height of both trigonid and talonid. The wear pattern, typical of a crushing function of the teeth, and their relatively large size are also characters of genera included in the “zhelestids”.

The dental formula of the Asian Turonian–Coniacian representatives of this paraphyletic family is characterized by the occurrence of five premolars. Nessov et al. (1998) have equivocally suggested that the small premolar at position three (p3 or permanent dp2) is lost in more derived “zhelestids”. If we admit this assumption, the p3–4 of Valentinella must be compared respectively with the p4–5 of Aspanlestes and Sorlestes , the only known primitive “zhelestids” documented by their lower premolars. These genera differ from Valentinella by their long, labio−lingually compressed, and premolariform p5 without metaconid; the p4 of Aspanlestes is more evocative of the p3 of ISEM/VLP−2 by its mesial protoconid and by the small distal heel without cusp, but clearly differs by a sharp cristid running from the apex of the protoconid to the distal margin of the tooth. As for the anterior upper teeth, ISEM/VLP−3 can be compared only with the holotype of Zhelestes temirkazyk . In this species, the large canine alveolus suggests that this tooth is longer and more acute than in Valentinella ; the two−rooted unicuspidate P2 are more similar by an asymmetrical profile, showing a small mesial shelf, whereas the distal side is steeper. Thus, the main difference between Valentinella and the Asian “zhelestids” concern the molarization of the ultimate premolar, this character is derived and can be explained by the chronological gap between both groups.

According to Nessov et al. (1998), Gallolestes , Alostera , and Avitotherium are the North American Late Cretaceous representatives of the “zhelestids”; these genera share with Valentinella the “zhelestid” molars traits listed above. The p3 of Avitotherium (Ciffelli 1990) obviously differs by its more slender protoconid, which bears both mesial and distal cristids. Moreover a lingual cingulid is present in Avitotherium , and the talonid is broad with two well−defined cusps enclosing a small talonid basin. As for the p3 of Gallolestes , only the distal part of the crown is known ( Lillegraven 1976) and the simplicity of the construction evokes the p3 of Valentinella . Both genera are also similar in that the molariform p4 bears a distinct metaconid and two strongly developed roots. Butler (1977, 1990) suggested that the fourth premolar of Gallolestes is a dp4 (see also Nessov et al. 1998); in the same way we also do not exclude a dp4 status for the ultimate premolar of Valentinella on ISEM/VLP−2. The molars are also similar in the reduction of the m3, their strong roots, their quadrate mesial occlusal outline, and by their robust aspect with rather bulbously−constructed cusps. Gallolestes differs in its smaller metaconid and entoconid relative to protoconid and hypoconid respectively.

The Late Cretaceous European genera Lainodon from Laño and Labes from Champ−Garimond and Quintanilla del Coco have been related to “zhelestids” by Gheerbrant and Astibia (1994, 1999). After that, Nessov et al. (1998) included these genera in the definition of the “zhelestids”. Morphologically, Lainodon and Labes share with Valentinella the “zhelestid” characters, and a robust general morphology including massive bunodont cusps and strong roots, an incipient exaenodonty, and similar dimensions as well ( Valentinella being slightly bigger). All the characters listed by Gheerbrant and Astibia (1999) to differentiate Lainodon from Labes cannot be observed on our specimens. With caution, the shorter talonid of Valentinella seems to indicate more affinities with Lainodon . Lainodon and Labes differ from Valentinella in having a more anteriorly positioned paraconid and by an hypoconulid stronger and more distally positioned on lower molars. A p2? or p3?, tentatively referred to Lainodon by Gheerbrant and Astibia (1999), is distinct to the p3 of Valentinella by its less robust morphology (not bulbous), its narrower occlusal outline, and by the occurrence of a crest on the distal flank of the protoconid.

Recently, Averianov et al. (2003) advocated a “zhelestid” status for the alleged marsupial from the Cretaceous of Madagascar described by Krause (2001). This fragmentary lower molar does not permit close comparisons with Valentinella . The greater trigonid angle of the Malagasy form (a primitive trait also observed in Labes , Lainodon and in both early metatherians and eutherians) suggests that Valentinella is more reminiscent of the older Asian “zhelestids” in this character.

To conclude, Valentinella is a eutherian, and could be a derived “zhelestid”, characterized by a probable molariform p4 (or dp4) and a slightly reduced m3; it could be therefore compared with some early “condylarths”. The crushing specialization, the entoconid−hypoconulid twinning, and the bulbous protoconid on p3 evoke the Mioclaenidae . The genera Promioclaenus , Mioclaenus and some species of Tiuclaenus present in particular a slightly to considerably reduced m3. Moreover, as Valentinella , Mioclaenus and Tiuclaenus robustus are also characterized by obsolete to absent ectocingulid ( Archibald 1998; de Muizon and Cifelli 2000). Valentinella is nevertheless distinct from mioclaenids by its molariform p4 (or dp4). In order to discuss the systematic position of Valentinella , notably its relationships with the “condylarths”, we carried out the analysis of the enamel microstructure of Valentinella .

Figures

Fig. 2. Valentinella vitrollense gen. et sp. nov., stereophotographs and explanatory drawings of ISEM/VLP−2 (holotype) in lingual (A, C) and occlusal (B, D) views.