Pseudocricetodon incertus (Schlosser, 1884)

Vianey-Liaud, Monique, Comte, Bernard, Marandat, Bernard, Peigné, Stéphane, Rage, Jean-Claude & Sudre, Jean, 2014, A new early Late Oligocene (MP 26) continental vertebrate fauna from Saint-Privat-des-Vieux (Alès Basin, Gard, Southern France), Geodiversitas 36 (4), pp. 565-622 : 588-597

Pseudocricetodon incertus (Schlosser, 1884)

( Fig. 10 )

Cricetodon incertum Schlosser, 1884 : pl. VIII, fig. 19a. — Schaub 1925: 45-49, pl. II, fig. 1.

Pseudocricetodon incertus – Hugueney 1980: 60, figs 7-9.

Allocricetodon incertus – Freudenthal 1994: 18, pl. 4, figs 7-12.

DIAGNOSIS. — Comte 1985: 42, 43.

HOLOTYPE. — Left dentary, bearing m2-m3, no. 1879. XV.171a (coll. BSPG, Munich), Schlosser 1884.

TYPE LOCALITY. — Mouillac (Quercy, Old Collections).

DISTRIBUTION. — Late Oligocene: MP 26 to MP 29 (cf. e.g., Comte 1985)

MATERIAL AND MEASUREMENTS ( Tables 3 , 4 ). — Ŋe normality of distributions was tested by the coefficient of variation of Pearson (100 Sd / average) distributed normally between four and ten for mammals (e.g., Simpson et al. 1960). In cases where this ratio was less than four, the normality of the samples was checked by the Shapiro-Wilk test (L.M3 / W = 0.855> W 0.05; w M / 1: W = 0.9543> W 0.05, LM / 2: W = 0.9488> W0.05) .

DESCRIPTION

Upper molars

M1. Ŋe prelobe is always larger than half the tooth width, and on most of the teeth (33/37) there is a sharp angle between its lingual side and the mesial flank of the protocone. Ŋis angle is less defined on four specimens.

3.70 4.70 On most teeth, the anterocone is simple and more or less transversely stretched (32/37). A tendency to split occurs (5/37). A spur behind the anterocone is present on 13 of the 35 teeth: it ends (10/13) in the prelobe basin. It bends to reach the lingual cingulum (2/13); on a single specimen (SPV 30) it connects to the anterior arm of the protocone. On most of the other teeth (33/35), this arm, which can be bent to the paracone, ends freely in the anterosinus. A crest is frequently observed in front of the paracone (18/35): it can be short (2/15) or bend to reach the labial cingulum (1/15), but in most cases (15/18) it joins the anterocone (= anterior ectoloph, according to the nomenclature of Freudenthal et al. 1994: 61). Ŋe lingual anteroloph joins the front side of the protocone in 31 teeth out of 35, closing the protosinus. More often interrupted (12/35), the labial anteroloph closes the anterosinus in 23 of 35 teeth.

Ŋe platform of the protocone (“triangle lingually to the anteroloph and in front of the protocone”; Freudenthal 1994: 9) is observed over about half of M1 (15/35); it is well established on the 2/3 of them (10/35). Ŋe protolophule is widely retroverted i.e. connected to the back of the protocone or to the endoloph (38/39), and in one case, it is linked to the protocone (transverse). A single tooth (SPV 54) also shows the connection of the anterior arm of the protocone with the paracone: therefore two junctions between the anterior cusps are present. Ŋe sinus is strongly proverse (it frequently ends between the protocone and the paracone: 31/39), just proverse in other cases (8/39). Ŋe lingual cingulum is rarely complete, and in this cases (2/39), completely closing the opening of the sinus.Ŋis cingulum is absent on about half of the teeth (21/39). An entostyle, more or less developed, is visible in 10 teeth (10/39); it can be isolated (3/39) or associated with an incomplete lingual cingulum (7/39).

Ŋe mesoloph is absent on SPV 50 (1/38), medium in length on about 23/ 38 specimens, sometimes it is a little shorter (6/38), and more rarely 3/4 length (2/38). It connects the labial edge of 6/ 38 specimens with an interruption on one of them. Ŋe mesostyle is absent on 4/37 teeth, in other cases (22/37) it forms a cingular crest bearing (6/37) or not a cusp, the cusp being isolated (11/37). An entomesoloph is visible on SPV47 . Ŋe metalophule is always proverse (36/36). On about 2/3 of the teeth (22/33), a more or less developed crest is observed on the distal flank of the paracone ; occasionnally oblique to the labial edge, it can be connected to the mesostyle. A short ridge in front of the metacone is visible on two teeth ( SPV 31 and SPV 39 ). In most cases (32/36) the posteroloph ends freely, leaving open the posterosinus ; except on four teeth on which it closes the latter, curving toward the posterior wall of the metacone.

11.9 14.2 10.9 12.6 M2. Ŋe anterior lingual cingulum is most often (22/31) well developed and clearly defines a protosinus; on four specimens (4/31), it is reduced to a low ridge without individualized protosinus; on the contrary, it is highly developed on five other teeth (5/31) where it extends over the inner side of the tooth, skirting the protocone. On SPV 88 , there is a labial connection between the anteroloph and the protolophule. Ŋe latter is attached on the anterior arm of the protocone. Beyond its junction with the endoloph, the posterior arm of the protocone extends mostly towards the paracone without reaching it (23/31) ; a complete double link of the anterior cusps is performed on 6 specimens (6/31). No trace of this double linking is visible on SPV 66 and SPV 80 .

Ŋe sinus, almost transverse on SPV 75 , does not extend deeply behind the protocone (24/31) ; it can be seen clearly around the protocone on 6 specimens. On SPV 63 there is a ridge on the posterior flank of the protocone, while on SPV 79 a complete but low protocone – hypocone connection is present .

Ŋe mesoloph is usually of medium length (18/30), long in other cases (12/30), and it can reach the labial edge (6/30) but it is interrupted on two specimens. A link metaloph – mesoloph is seen on two teeth. A second mesoloph is present on SPV 83. A complete (6/29) or incomplete labial cingulum, is present on the majority of teeth (28/29); on SPV 71, it is a small cusp independent of the mesoloph. On about half of the teeth (15/29) this cingulum is in line (9/15), either directly or after an interruption, with a more or less developed ridge descending obliquely from the top of the paracone. Ŋe metaloph, proverse (31/31), is inserted on the anterior arm of the hypocone more or less in front of this cusp. On SPV 79, behind the metaloph, a ridge runs towards the posteroloph without reaching it. Ŋe posterior cingulum joins the metaloph, closing the posterosinus (9/30).

M3. Ŋe lingual anteroloph is present in seven of eight specimens, only weakly indicated for three of them on which the protosinus is absent. Four other teeth show a strong anteroloph defining a sharp protosinus; the anterolingual cingulum extends to the hypocone. Ŋe protolophule is connected to the anterolophule (4/6) or anterocone. Ŋe sinus is absent (3/8), small (3/8) or deep (2/8), the neoendoloph being missing or incomplete. Ŋe mesoloph is either absent (1/8), of medium length (2/8), or long (2/8), reaching the labial edge on three other specimens. Ŋe endoloph is complete

(5/8) or absent (3/8). Ŋe axioloph is complete (2/8), incomplete (3/8) or absent (3/8). A complete labial cingulum closes the mesosinus only on half the teeth (4/8). Ŋe metacone forms a swollen cusp on five specimens (5/7); it is merged in the labial ridge on the others (2/7). In continuation of this ridge, the posteroloph closes the posterosinus on most specimens (6/7).

Lower teeth

m1. Ŋe anteroconid position is axial or slightly offset lingually.Ŋe lingual anterolophid is absent on SPV 110. On the other teeth, it joins (12/17) or not (4/17) the metaconid.Ŋe labial anterolophid is short (5/15), or progressively runs down the base of the protoconid (10/15). Ŋe anterolophulid is complete (18/18), obliquely to the axis of the ectolophid. Ŋe metalophulid is absent (6/16) or reduced to a short spur (10/16). Ŋe posterior arm of the protoconid is always connected to the metaconid (18/18); the mesolophid is short (2/16), mostly halflong (11/16) or long (3/16), without reaching the lingual side. A second mesolophid is present on four teeth. Ŋe sinusid is most often slightly oriented backward, and transverse on two teeth (2/18); it is closed by a discrete and often incomplete cingulum. A poorly developed ectomesolophid is present on four specimens (4/17). Ŋe hypolophulid is attached in front of the anterior arm of the hypoconid (17/18); it is interrupted before this junction on SPV 112, and directly connected at the hypoconid on SPV 113. Ŋe posterosinus is closed by the prolongation of the posterolophid toward the entoconid. Ŋe posterior arm of the hypoconid is absent.

m2. Ŋe tooth SPV 149 has a short labial anterolophid; it descends to the base of the protoconid on all other specimens (32/33). Ŋe anterolophulid is complete (32/32) and almost as high as the anteroconid. Ŋe metalophulid is complete, obliquely forwards, and is always inserted on the anterior arm of protoconid. A crest behind the metalophid (metalophulid IIṙ) connects to the posterior arm of the protoconid on SPV 130. It generally ends free. It is short on SPV 124 or generally of medium length (28/33); on some specimens (4/33) it curves toward the metaconid. Ŋe mesoconid is rarely distinct. Ŋe mesolophid may be absent (8/33), short (6/33), halflong (16/33) or long (3/33), and in this case, it can be very low and slender (SPV 130). About one third of the teeth (10/33) have an ectomesolophid. Ŋe sinusid, tranverse (15/33) or weakly retroverted (18/33) is closed on most specimens (28/31) by a low and often incomplete cingulum. On the lingual side, the distal flank of the metaconid forms a ridge down to the base of the entoconid, but without joining it. Ŋe hypolophid is generally (30/33) inserted onto the anterior arm of the hypoconid and oblique forward; it is transverse, inserted just before the hypoconid on three specimens (3/33). Lingually, the posterolophid joins the top of the entoconid, closing strongly the posterosinusid. Ŋe posterior arm of the hypoconid is absent.

+

–

+

–

–

m3. Ŋe labial portion of the anterolophid always reaches the base of the protoconid (10/10), while its lingual part is connected to the metaconid. Ŋe metalophulid is complete and connected to the anterolophulid (7/10) or to the anteroconid (3/10). Behind the metalophid, two teeth (2/11) show a ridge that does not reach the posterior arm of the protoconid. Ŋe latter is long, freely ending (10/11), and joins the lingual edge only on SPV 155. Ŋe mesolophid is absent (11/11). Ŋere is no mesoconid, and two teeth (2/11) have an ectomesolophid. On all teeth a crest, variable in thickness, runs down from the metaconid and closes the mesosinusid. Ŋe sinusid is most often closed by a cingulum more or less complete (9/10). Ŋe hypolophulid oblique forward fits the anterior arm of the hypoconid, more or less in front of this cuspid. Ŋe posterolophid is connected lingually to the top of the entoconid, strongly closing the posterosinus.Ŋe posterior arm of the hypoconid is absent.

Incisors. Numerous lower incisors display an ornamentation of longitudinal ribs associated with ribs becoming oblique labially. Ŋis pattern is similar to that of two fragments of incisors from BoujacA (Gard, Alès basin; MP 27) assigned to Pseudocricetodon incertus ( Comte 1985: 34, fig. 11g, h, and 47, 48).

REMARK AND COMPARISONS

Ŋe species Pseudocricetodon incertus was reported from BoujacA (MP 27; Alès Basin, France ) on the basis of a relatively small sample ( Comte 1985). Freudenthal (1994: 19) notes strong contradictions in the size distributions at MIR4C and BoujacA, suggesting that the material contains more than one species in this latter locality. Ŋe comparison of BoujacA material with the more abundant one now known from Saint-Privat-des-Vieux, in the same basin, allows adopting this view. A few teeth (one M1, one M3, one m1 and two m2) must be separated out of BoujacA material initially assigned to P. incertus . Ŋey demonstrate the existence of a larger form that could be affine to Pseudocricetodon landroveri (Daams, Freudenthal, Lacomba & Alvarez, 1989) , a species initially considered as a small species of the genus Heterocricetodon Schaub, 1925 . Ŋerefore, a few teeth from BoujacA belong to Pseudocricetodon aff. landroveri ( Daams et al. 1989: 43-48) : m1 (BJA763): 1.86 × 1.23 mm, m2 (BJA780): 1.64 × 1.51 mm, m2 (BJA781): 1.72 × 1.43 mm, (BJA726): 1.81× 1.63 mm, M3 (BJA750): 1.29 × 1.31 mm. Ŋe dimensions of these teeth fall within the range of variation of the type species from Pareja (Guadalajara, Spain; Daams et al. 1989), except for the M2 (L = 1.81 mm) which is longer than in Pareja (L. max = 1.71 mm), and rather in the upper part for m2. Ŋe single m1 BJA763 is moderately worn. Its lingual anterolophid is strong and fully closes the very deep antesinusid. Ŋe labial anterolophid gradually extends to the base of the protoconid. Ŋe anterior arm of the protoconid (anterolophulid) is oblique and joins the anteroconid. Ŋe posterior arm of the protoconid is connected to the metaconid. Ŋere is no metalophid. Ŋe mesoconid has a very broad base, but there is no trace of mesolophid or ectomesolophid. On the lingual side, the trace of a crest remains behind the metaconid. Ŋe sinusid is wide open with no trace of labial cingulum. Ŋe posterolophid is widely connected to the entoconid.

Ŋe rest of the BoujacA material remains allocat- ed to P. incertus . Ŋis material has been measured again after exclusion of the largest teeth ( Table 4 ). Normality of the samples was controlled by the Shapiro-Wilk test (with values of W: 0.82 <W <0.96 always exceeding the limit value). Tooth size of this small population was compared to that of P. incertus from Ehrenstein7 ( Germany), but also to that of P. cornelii Freudenthal, 1994 of Mirambueno1 ( Spain). Although precise comparison of dimensions of P. cornelii of MIR1 with the population from BoujacA suffers from the reduced sample in the latter locality, the dental proportions remain comparable ( Table 5 ). Ŋe differences concern the M2, m1, m2 and m3, longer in BoujacA; these are significant, with t-values that are less than 6. Ŋe m1, m2 and M2 are also wider, but the differences are weakly significant. With the same caution, the comparison between Mirambueno1 materials with that of BoujacA indicates that the M2 and m2, best represented molars in the latter locality, are significantly longer and wider. Ŋe dimensions of the m1 and the length of m3 of the species from BoujacA would also be stronger. In general terms, the differences are most pronounced between P. cornelii of MIR1 and the species of BoujacA, which indicate a slightly greater size for the latter.

Ŋe frequencies of morphological features were compared using the chi-square. Morphologically, the m1 from BoujacA differ very significantly from those of P. cornelii from Mirambueno1 by the absence of the characters “high anterolophid” and “absence of ectomesolophid” ( Table 6 ; 7 ). In BoujacA, the posterior arm of the protoconid is connected to the metaconid on m2 and can reach the lingual side of m3, while these characters are absent in P. cornelii . Ŋe development of the mesolophid of m2 is again comparable in these two localities.

COMPARISONS AND DISCUSSION ABOUT PSEUDOCRICETODON POPULATIONS

Ŋe size of the teeth of P. incertus from Mirambueno4C (MIR4C) and Saint-Privat-des-Vieux were compared using the t-test ( Table 5 ). Ŋe difference between the dimensions of the M3 is only weakly significant (with a confidence coefficient of 95%). Ŋe m2 length difference is also weakly significant (t = 2.7). Ŋe difference between the M1 lengths of the two localities disappears when excluding the large specimen SPV 23 (2.11 × 1.39 mm). In addition, the distribution of the lengths of the M1 population of Saint-Privat-des-Vieux, by mean of Shapiro test, is normal (Shapiro-Wilk with W = 0984, above the limit values for risk 5% [W = 0935] and 1% [W = 0912]). Otherwise, the dimensions of the tooth SPV23 fall outside those of the M1 of Pseudocricetodon landroveri . Ŋerefore we keep the attribution of this tooth to P.incertus . Differences, including the length of M1, could thus result from the small sample from Mirambueno4C. Ŋe dimensions of most molars of P. incertus from Mirambueno4C and Saint-Privat-des-Vieux appear very similar with a possible trend of size increase of some of them in the Alès basin.

According to Freudenthal (1994: 20), P. cornelii from the younger locality MIR1 at Mirambueno (MP 27), with identical dimensions, differs from P.incertus from MIR4C in its morphology ( Tables 7 , 8 ). Frequencies of the distinctive features retained by Freudenthal (1994) as well as a few others are compared using the chi-square, for the populations of Saint-Privat-des-Vieux and MIR4C ( P. incertus ) on one hand, Saint-Privat-des-Vieux and MIR1 ( P. cornelii ) on the other. For M1, the characters that distinguish P.incertus (MIR4C) from P. cornelii (MIR1)are not significant. For all other morphological characters selected by Freudenthal, the population of Saint-Privat-de-Vieux does not show any difference with that of P.incertus of MIR4C, which however represents a small sample. However if one refers to the differential diagnosis (and M1 aside), significant differences appear allowing to separate the population of BoujacA with P. cornelii from MIR1: lingual anterolophid never high, ectomesolophid of m1 more frequently absent, mesolophid more frequent and better developed on m2, posterior arm of the protoconid reaching the lingual edge of m3. Finally, in terms of dimensions and of morphology, the populations of Pseudocricetodon incertus from Saint-Privat-des-Vieux (MP 26) in southern France and MIR4C (MP 26) in Spain are similar. Ŋey both differ from the population described as P. cornelii (Freudenthal 1994) , from Mirambueno1 (MP 27).

Ŋe species incertus was described by Dienemann (1987: 45-56, fig. 23) under the genus name Eucricetodon Ŋaler, 1969 from different localities in Germany, the species being particularly well represented in Ehrenstein 7 (MP 27). Ŋe comparison with the forms of BoujacA and Mirambueno1 shows that the m1 from Ehrenstein 7 exhibits features of P. incertus (lingual anterolophid always low, almost complete absence of ectomesolophid) comparable to those from BoujacA, but also from Mirambueno4C and Saint-Privat-des-Vieux. Ŋe m2 of Ehrenstein 7 are different in having the posterior arm of the protoconid not connected to the metaconid (Dienemann 1987: fig. 23). Ŋe absence of mesolophid shows a comparable frequency in Ehrenstein 7 and BoujacA. Ŋis confirms the attribution of the population of BoujacA to P.incertus .

Ŋus,in two subcontemporaneous localities close to the MP 27 standard level, the frequency of character “mesolophid absent” (7/10 at BoujacA and 29/32 at Erhenstein 7) in P. incertus is of same order as that in P. cornelii of Mirambueno1 (33/47). Ŋe decrease in the frequency of the mesolophid on m2, which is 2/17 at Mirambueno4C and 8/33 at Saint-Privatdes-Vieux (MP 26), compared to its frequency at BoujacA and Ehrenstein7 (MP 27), may indicate a simplification of the tooth pattern over time.However, it is important to be cautious because the number of specimens is small in Mirambueno4C and BoujacA. Ŋe lack of the character “posterior arm of the protoconid not connected to metaconid” in Ehrenstein 7, also suggests geographical variations affecting these subcontemporaneous populations.

Freudenthal (1994: 29) considered premature to establish a phylogeny of the three species P. incertus , P. cornelii and P. landroveri . However, he suggested that P.incertus could have given rise to P.cornelii with increasing complexity of teeth, whereas P. landroveri may not be the ancestor of the latter because of its size and its occurrence together with P. cornelii in Vivel del Rio ( Spain, MP 28). Ŋe occurrence of P. aff. landroveri in BoujacA (MP 27) introduces a new step in this lineage, while that of P. incertus in this level invalidates the hypothesis of a direct filiation P. incertus - P. cornelii .

However, the differences noted between P. incertus from Saint-Privat-des-Vieux or Mirambueno4C (MP 26) and P.cornelii from Mirambueno1 (MP 27) are less pronounced between P. incertus of BoujacA and Ehrenstein7 (MP 27) on the one hand and P.cornelii on the other hand. Ŋus P. cornelii could be a descendant of P. incertus or a geographical variant of this species, a hypothesis suggested by its differences with the contemporary populations from BoujacA and Ehrenstein7. However, the particular characteristics of m1of P. cornelii (“high anterolophid” and “absence of ectomesolophid”) are absent in all populations assigned to P.incertus whereas they occur in P. landroveri . Pseudocricetodon cornelii and P. landroveri , which coexist in the MP 26 and MP 27 levels, could have evolved in a parallel way from a common stock.Since m1 characters – like presence of high anterolophid and absence of ectomesolophid – occur in P.montalbanensis , this species may represent the stem species.

Figures

FIG. 10. — Pseudocricetodon incertus (Schlosser,1884) from Saint-Privat-des-Vieux (Gard;Lower Chattian,MP 26):A, left m1, SPV109: A1, occlusal view, A2, lingual view; B, left m2, SPV131: B1, occlusal view, B2, lingual view; C, left m3, SPV155: C1, occlusal view, C2, lingual view; D, left M1, SPV26: D1, occlusal view, D2, lingual view; E, left M2, SPV61: E1, occlusal view, E2, lingual view; F, left M3, SPV93: F1, occlusal view, F2, lingual view. Scale bar: 1 mm.