Pataecops Radinsky, 1966

Averianov, Alexander O. & Godinot, Marc, 2005, Ceratomorphs (Mammalia, Perissodactyla) from the early Eocene Andarak 2 locality in Kyrgyzstan, Geodiversitas 27 (2), pp. 221-237 : 228-233

Genus Pataecops Radinsky, 1966

Pataecops minutissimus ( Reshetov, 1979) n. comb. ( Fig. 3 )

Rhodopagus minutissimus [nomen nudum] Reshetov, Shevyreva, Trofimov & Chkhikvadze, 1978: 151.

Pataecops microdon [nomen nudum] Reshetov, Shevyreva, Trofimov & Chkhikvadze, 1978: 151.

Rhodopagus minutissimus Reshetov, 1979: 29 , fig. 5-1. — Gabunia & Kukhaleishvili 1991: 111.

Pataecops microdon Reshetov, 1979: 31 , fig. 5-3.

“ Pataecops ” microdon – Dashzeveg 1991: 36.

REFERRED MATERIAL. — ZIN 35286, right D3?; ZIN 34030, left D4; ZIN 35287, left maxillary fragment with P4 (broken) and M1; ZIN 34029, right maxillary fragment with M1-2; ZIN 35288, left maxillary fragment with P3-M3 and alveolus of P2; ZIN 35289, right maxillary fragment with M3 and partially preserved M1-2; ZIN 35290, left maxillary fragment with D4 M1-2; ZIN 35291, right maxillary fragment with M2-3; ZIN 35293, left maxillary fragment with M1; ZIN 34031, right M1; ZIN 35292, left M2; ZIN 35296, right M2; ZIN 35295, left M3; ZIN 35294 and 35297, right M3; ZIN 34028, right dentary fragment with d3-4 and alveoli of d2 and m1, and fragments of m 2 in the alveolus; ZIN 35298, dentary symphysis fragment with alveoli of i3, c1 and p2.

DESCRIPTION

The infraorbital foramen (seen only in ZIN 35288, Fig. 3A ) is relatively small and placed above P2. The anterior zygomatic process of the maxilla begins above M1 and terminates at the posterior end of M3. The zygomatic arch is placed relatively low on the maxilla, close to the tooth row. The bony palate terminates approximately between M2 and M3.

No P1 is present in our material. This tooth is most probably lacking in Pataecops parvus ( Radinsky, 1965) and Rhodopagus pygmaeus Radinsky, 1965 : a two-rooted alveolus in front of P 2 in these species (e.g., Radinsky 1965: 212, fig. 8) may actually belong to a D1, which persisted late enough to leave unresorbed alveoli present with the definitive upper cheek dentition, and which was never replaced by a P1. This situation would be symmetrical to that on the lower jaw, where d1 was present but not replaced by p1 (see below).

P2 was two-rooted, as can be judged from its alveolus (ZIN 35288), with a large rounded and labiolingually elongated posterior root and a much smaller anterior root.

The P3-4 are non-molariform teeth, subtriangular (P3) to subrectangular (P4) in crown outline. The parastyle is smaller than the paracone. The ectoloph is straight (P3) or slightly convex labially (P4). On P3 the metaloph is a short ridge well separated from the protocone by a narrow valley. In P4 this feature is obliterated by wear. The anterior and posterior cingula are well developed. The lingual and labial cingula are completely lacking on P3-4.

The D3? and D4 are fully molariform. They differ from M1 only by their smaller size and straighter ectoloph.

The M1 and the M2 differ from one another in size and proportions: M2 is larger and more elongated antero-posteriorly, M1 is more squarish. On M2 the ectoloph is slightly more deflected lingually. The parastyle is well developed, but lower than the paracone and columnar. Posteriorly to the paracone, the ectoloph is bent, concave or convex labially. The protoloph and the metaloph are oblique crests. The protoloph joins the ectoloph between the parastyle and the paracone. The metaloph runs to the metacone (“premetaconule crista” of Hooker 1989). There is a relatively strong anterior cingulum and a weaker posterior cingulum. The labial cingulum is absent. The lingual cingulum is better developed on smaller teeth; on larger teeth it is interrupted at the protocone and hypocone bases. On all available M2s the postmetacrista is bent labially and not confluent with the metaloph.

The M3s vary significantly in size ( Table 3) and crown shape. The crown has a generally triangular outline. The parastyle is relatively large, but much lower than the paracone and columnar. The ectoloph is much deflected lingually and nearly parallel to the protoloph. The metaloph is half the length of the protoloph. The postmetacrista is usually shorter than the metaloph and confluent with the latter. There is only an anterior cingulum, which is well developed. Sometimes at the base of the protocone there is a remnant of lingual cingulum.

The specimens ZIN 35288, 35290, and 35293 give some information about the roots of the upper cheek teeth. On ZIN 35288 the roots of P3-4 are extremely long and gradually tapering towards their free ends. The same can be seen on ZIN 35290, a juvenile specimen, for D4 and M1. However, the molars of ZIN 35288 and ZIN 35293, much older individuals, have bulbous (swollen) roots and sometimes curved at their free ends. Swollen roots is probably an age correlated character, which is apparently more variable than was thought by Radinsky (1965), who used it for distinguishing Rhodopagus pygmaeus from Pataecops parvus . At least in one specimen of P. parvus ( Dashzeveg 1991: fig. 16) the upper molars appear to have lingual roots inflated distally, like in R. pygmaeus .

The dentary symphysis is completely fused, relatively short (11.5 mm, ZIN 35298) and terminates at the level of p2 (first cheek tooth). The alveolus for c1 is relatively large. The length of the lower diastema is 6.9 mm. There is a small mental foramen between c1 and p2.

There was no d1, judging from absence of its alveolus in ZIN 34028. Apparently, it was not replaced by p1, because this tooth is lacking in P. parvus and Rhodopagus .

The d2 was two-rooted, distinctly smaller than the d3.

The d3 and the d4 have broadly similar morphologies, differing mostly in size and proportions. The trigonid and talonid basins are widely open lingually. The labial part of the paralophid and the crista obliqua are almost parallel to the labial side of the crown. The protolophid and the hypolophid are oblique crests. d3 has a relatively longer and more open trigonid with a distinct paraconid. On d4 the antero-lingual portion of the paralophid goes parallel to the protolophid.

The m2 is preserved only on ZIN 34028 ( Fig. 3M, N ), where it is slightly incomplete and not fully erupted. The protolophid and hypolophid are obliquely oriented to the longitudinal axis of the tooth. The hypolophid is higher than the protolophid. The cristid obliqua is relatively unreduced, high, relatively long, and arcuate, meeting the protolophid rather lingually. The posterior cingulid is rather short, limited to the lingual portion of the crown.

Measurements

For upper dentition: see Table 3. ZIN 34028: d3: L = 4.5, WTR = 2.2, WTL = 2.5; d4: L = 5.4, WTR = 3.0, WTL = 3.1.

DISCUSSION

As was pointed out by Gabunia & Kukhaleishvili (1991: 111), the holotype of R. minutissimus (PIN 3486-1) sharply differs from the other specimens of Rhodopagus and other lophialetids except Pataecops by a much more considerable lingual displacement of the elongated ectoloph on M1 and M2. They also noted some differences in the ectoloph structure between the molars of R. minutissimus and those of Pataecops . Indeed, the ectoloph in the R. minutissimus holotype is concave labially and the metaloph seems to be slightly less reduced than in Pataecops parvus from the middle Eocene of Mongolia ( Radinsky 1965; Dashzeveg 1991). Our specimens show, however, that these characters are quite variable. Among five M1-2 at our disposal, three have a concave ectoloph and two a convex one. Moreover, the degree of metaloph reduction on M1-2 is greatly variable. Crucial for our taxonomic conclusion is the association of an M2 with a labially concave ectoloph and a relatively long metaloph and an M3 of typical Pataecops pattern (ZIN 35288; Fig. 3B ) (the holotype and only known specimen of R. minutissimus lacks M3). This fact lumps together specific characters of R. minutissimus and Pataecops microdon , which was known previously from four isolated upper molars from the same locality, Andarak 2. Thus we conclude that there is only one species of small archaic rhinoceratoids (“rhodopagids”) in the Andarak 2 locality. We consider Pataecops microdon Reshetov, 1979 (p. 31) as a junior synonym of Rhodopagus minutissimus Reshetov, 1979 (p. 29) by page priority, and we refer this species to the genus Pataecops as P. minutissimus ( Reshetov, 1979) n. comb. The names Pataecops microdon and Rhodopagus minutissimus were used in an earlier publication, in an abstract by Reshetov et al. (1978) which provided a prelimi- nary faunal list for the Andarak locality. However, in that publication these names clearly appeared as nomina nuda as they were not accompanied by any description, nor by a holotype designation.

Dashzeveg (1991: 36, tabl. 8) considered “ Pataecops ” microdon as belonging neither to Pataecops nor to Rhodopagus , because of its small size and the shape of its crown, with a lingually shifted and short ectoloph parallel to the protoloph. However, the holotype of P. microdon (PIN 3486-2) represents the extreme case of M3 variation in Andarak’s sample. Among four M3 at our disposal, three (ZIN 35288, 35294, 35295; Fig. 3B, H, L ) approximate the condition of PIN 3486- 2 in having a reduced metaloph parallel to the ectoloph as well as other small differences in crown shape, and one specimen (ZIN 3 5 2 9 7) approximates the condition of Mongolian P. parvus with a relatively longer metaloph not parallel to the ectoloph. P. minutissimus n. comb. is further similar with P. parvus in having a small infraorbital foramen placed above P2 (in R. pygmaeus this foramen is larger and placed above P1; Radinsky 1965: 208). All this allows us the placement of Andarak’s species in the genus Pataecops .

The Andarak species clearly differs from the Mongolian middle Eocene P. parvus by its small- er size, the lack of a labial cingulum on P3-4, more variable M1-2 some of which have a concave ectoloph and more variable M3 which in some cases are more derived than those of the Mongolian species in having a more reduced metaloph parallel to the ectoloph. The upper molars in all known specimens of P. parvus from the Kholbolzdhi Formation at the Valley of Lakes, Mongolia ( Radinsky 1965; Dashzeveg 1991) have a convex ectoloph.

The genera Rhodopagus and Pataecops were originally placed by Radinsky (1965) in the family Lophialetidae with a question mark. He noted an evolution of these genera toward a functionally bilophodont dentition (by the lingual depression of the metacone), while the typical lophialetids show the development of a rhinocerotoid-like molar occlusion. However, he also noted later (Radinsky 1969) some similarities of these genera with rhinocerotoids in the wear facet pattern. Reshetov (1975) united Rhodopagus and Pataecops into a new subfamily Rhodopaginae provisionally placed within the Lophialetidae . Subsequently, Lucas & Schoch (1981) transferred both genera to the Hyracodontidae Cope, 1879 and cited eight shared characters which they considered synapomorphies for this family. Prothero et al. (1986) placed “rhodopagids” in a trichotomy with Hyracodontidae and Rhinocerotidae Gray, 1821 . They considered Rhodopagus and Pataecops as primitive sister taxa of either hyracodontids, or possibly the sister-group of the rhinocerotids plus hyracodontids. Attribution of “rhodopagids” to Hyracodontidae was upheld by Hopson (1989) and Dashzeveg (1991), but was not supported by the cladistic analysis of Hooker (1989), who suggested that Rhodopaginae may be the sister group of Deperetellidae . Prothero et al. (1986: 361) cited as an additional “hyracodont feature” of Rhodopagus the reduced c 1 in R. minimus (AMNH 26112; this character may be seen also in R. zdanskyi Lucas & Schoch, 1981 : fig. 4f). In ZIN 35298, a symphysis fragment of P. minutissimus n. comb., the alveolus for c1 is nearly two times larger than the alveolus for the preceding incisor, so the lower canine in this species was not reduced (unknown for P. parvus ).

Gabunia & Kukhaleishvili (1991) reviewed the characters listed by Lucas & Schoch (1981) as synapomorphies between Rhodopagus and Hyracodontidae , based on the discovery of the oldest Rhodopagus species , R. radinskyi from the early Eocene Chakpaktas Svita in Zaisan Basin, eastern Kazakhstan. They came to the conclusion that Rhodopagus is closer to the tapiroids ( Lophialetidae ) than to the rhinocerotoids ( Hyracodontidae ). Dashzeveg & Hooker (1997) found that this conclusion was based on plesiomorphic or wider distributed characters. They found also that “ Rhodopagus ” radinskyi should be placed “in a new rhodopagid genus” ( Dashzeveg & Hooker 1997: 128). According to these authors, Veragromovia desmatotheroides Gabunia, 1961 , known by one isolated M3 from the middle Eocene Obayla Svita of the Zaisan Basin, Kazakhstan and originally described as a helaletid, should be transferred to the Rhodopagidae . The lower crowned Rhodopagidae do not show special relationships with other rhinocerotoid families, but two characters advocate for their inclusion into the Rhinocerotoidea ( Dashzeveg & Hooker 1997: 128): 1) a distinct break between the metaloph and the protocone, which is characteristic of nonmolariform teeth of hyrachyids, hyracodontids, amynodontids, and rhinocerotids, and well marked in Pataecops and “ Rhodopagus ” radinskyi , although less obvious in Rhodopagus ; and 2) the lack of any sign of the major narial incision which typifies all tapiroids except Heptodon . The latter character among rhodopagids is known only for “ Rhodopagus ” radinskyi , a more or less complete skull of which is known ( Gabunia & Kukhaleishvili 1991: fig. 1a, b). A relatively anterior position of the infraorbital foramen in R. pygmaeus , P. parvus , and P. minutissimus n. comb. may suggest the lack or the incipient stage of a narial incision, although these features are not unequivocally related ( Antoine 2002). However, this character may not be so important because in the primitive lophialetid Eoletes gracilis the narial incision is also poorly developed ( Lucas et al. 1997: fig. 6). Only the presence of a distinct break between the metaloph and the protocone on the upper premolars, found also in Andarak’s P. minutissimus n. comb. (ZIN 35288, see above), remains a good character indicating rhinocerotoid affinities of the Rhodopagidae . We subsequently classify this family as Rhinocerotoidea incertae sedis.

Interestingly, the upper cheek teeth of Pataecops look more rhinocerotoid-like (more advanced) than those of Rhodopagus : on the premolars the metaloph is more distinctly separated from the protocone, P4 may bear a short crista ( Dashzeveg 1991: fig. 16); the parastyles on M1-3 are more columnar, M3 is essentially more rhinocerotoidlike in its triangular shape, lingually displaced metacone, and greater reduction of the postmetacrista. Moreover, Rhodopagus has simpler upper premolars with a less evident break between metaloph and protocone (approximating the lophoid loop of lophialetids), relatively larger scale-like parastyles (another “tapiroid” feature) and more trapezoid M3 with less reduced metacone and postmetacrista (which is typical for “tapiroids”, but also found in some early hyracodontids, e.g., Triplopus Cope, 1880 ). However, Rhodopagus and Pataecops still share some derived traits ( Hooker 1989; Holbrook 1999): loss of p1, lingually displaced metacone on upper molars, loss of m3 hypoconulid (for P. parvus see Dashzeveg 1991: 36).

Pataecops minutissimus n. comb. is the smallest known ceratomorph ever found: length of M1-3 is 17.1 (ZIN 35288) and 17.5 (ZIN 35289). This length is 19.2 in the previously known smallest ceratomorph Fouchia elegans Emry, 1989 from the middle Eocene of North America ( Emry 1989).

Family AMYNODONTIDAE Scott & Osborn, 1883 Genus Sharamynodon Kretzoi, 1942

Figures

FIG. 3. — Pataecops minutissimus (Reshetov, 1979) n. comb. in labial (A, N) and occlusal (B-M) views; Andarak 2, Kyrgyzstan, Alay beds, Irdinmanhan; A, B, ZIN 35288, left maxillary fragment with P3-M3 and alveolus of P2; C, ZIN 35290, left maxillary fragment with D4 M1-2; D, ZIN 35287, left maxillary fragment with M1 and root of P4; E, ZIN 34030, left D4; F, ZIN 35286, right D3(?); G, ZIN 35292, left M2; H, ZIN 35294, right M3; I, ZIN 35296, right M2; J, ZIN 35293, left maxillary fragment with M1; K, ZIN 34031, right M1; L, ZIN 35295, left M3; M, N, ZIN 34028, right dentary fragment with d3-4, alveoli of d2, m1, and with partial m2 crown in crypt. Abbreviation: if, infraorbital foramen. Scale bar: 1 cm.