Equus nalaikhaensis Kuznetsova & Zhegallo, 1996

Eisenmann, Véra & Kuznetsova, Tatiana, 2004, Early Pleistocene equids (Mammalia, Perissodactyla) of Nalaikha, Mongolia, and the emergence of modern Equus Linnaeus, 1758, Geodiversitas 26 (3), pp. 535-561 : 535-561

Equus nalaikhaensis Kuznetsova & Zhegallo, 1996

SKULL

An exceptionally well preserved skull (PIN 3747- 500) belongs to a very old male ( Fig. 5 ). The basilar length is 545 mm. Among extant wild species, only E. grevyi reaches this size. Similar or larger sizes may be observed in Pliocene equids like Allohippus stenonis of Saint-Vallier or Plesippus shoshonensis Gidley, 1930 of Hagerman Quarry (Idaho, USA). But species belonging to primitive forms such as Allohippus and Plesippus Matthew, 1924 differ from extant species by basicranial proportions ( Eisenmann & Baylac 2000). In extant species the distance between the Basion and the Hormion ( Fig. 6 ) is long relative to the overall palatal length. By that character, the skull of Nalaikha must be considered as modern and referred to Equus .

During the middle Pleistocene, the caballine equids were frequently very large. The skulls of extant caballines wild and domestic ( E. przewalskii Poliakov, 1881 , E. caballus Linnaeus, 1758 or E. ferus Boddaert, 1785 ) and fossil caballines may be distingushed from the skulls of asses (wild and domestic) by the so-called Franck’s Index. In asses, the distance between the staphylion and the hormion ( Fig. 6 ) is the longest, while in caballines, the longest is the distance between hormion and basion. The separation is not good between caballines and other extant species. Fortunately, the Palatal Index (Eisenmann in press) gives much better discriminations. Both indices indicate that E. nalaikhaensis is not a caballine.

Further comparisons were done with non caballine extant Equus : Grevy’s zebras ( E. grevyi ), plains zebras ( E. burchelli Gray, 1824 and E. quagga Gmelin, 1788 ), mountain zebras ( E. zebra ), asses ( E. africanus Heuglin & Fitzinger, 1866 and E. asinus Linnaeus, 1758 ), hemiones ( E. hemionus and E. kiang Moorcroft, 1841 ). The overall

350 400 450 500 550 600 650

FIG. 4. — Relation between basilar lengths and distal articular breadths of metatarsals (MTIII) in different species of equids. Abbreviations: A., Allohippus ; E., Equus .

resemblance of the skull of E. nalaikhaensis with that of E. grevyi is striking. E. nalaikhaensis fits inside the range of variation of E. grevyi by almost all its dimensions ( Fig. 7 ; Annexe: Table 2). In particular, like in E. grevyi and most other zebras, and contrary to asses and hemiones, the skull is narrow ( Fig. 8 ), even narrower than the average of them. But unlike most zebras and asses, and like in hemiones, the face is high. Moreover, relative to the basilar length, the muzzle is short. Similar shortness is exceptional in extant Equus , even in hemiones. Summing up, E. nalaikhaensis combines skull characters of extant zebras, asses and hemiones in an altogether original way.

Comparison with three fossil skulls are of special interest. They belong to E. graziosii , E. apolloniensis and E. coliemensis .

1. The type skull of E. graziosii is preserved at the Palaeontological Museum of Montevarchi and was described by Azzaroli (1966, 1979) who kindly provided a cast for the former Laboratoire de Paléontologie (Muséum national d’Histoire naturelle, Paris). This skull lacks the posterior part but it is perfectly preserved otherwise. E. graziosii (middle-upper Pleistocene) is much smaller than E. nalaikhaensis to which it resembles by its face and muzzle proportions; the frontal is however as wide as in E. grevyi ( Fig. 7 ; Annexe: Table 2).

2. The type skull (APL 148) of E. apolloniensis of latest Villafranchian age (early-middle Pleistocene) excavated in Macedonia, was described by Koufos et al. (1997). It is almost entire but laterally compressed which makes comparisons difficult. In particular it is difficult to estimate its frontal width. However, like E. nalaikhaensis and E. graziosii , it definitely belongs to Equus by its basicranial proportions, as already demonstrated by Koufos et al. (1997). Like in E. nalaikhaensis , the muzzle is very short, but the face looks wider at the frontals (APL 147) and short, and the narial opening is also relatively shorter.

3. The northeastern Siberian skull (IA 1741, Geological Institute, Iakutsk) from Kolyma, probably late-early or early-middle Pleistocene, is the type of E. coliemensis and was described by Lazarev (1980). By the basicranial proportions, E. coliemensis is certainly also an Equus . It is about the same size as E. nalaikhaensis (basilar length: 538 mm) but different in proportions: the skull is wide at the frontals ( Fig. 8 ), the supra-occipital crest is very narrow, and the muzzle longer and wider. Among extant species, the greatest similarities are with the much smaller E. hemionus khur Lesson, 1827 – the Indian hemione ( Fig. 19 ; Annexe: Table 2). In particular, the shape and size of the supra-occipital crest are strikingly similar.

UPPER CHEEK TEETH

Unfortunately the skull of E. nalaikhaensis belongs to a very old animal and its teeth are worn to the roots. Other cheek teeth are extremely variable in size and in protocone length. Figure 9 is a scatter diagram of the protocone length versus the average occlusal diameter (length + width)/2. It shows that the teeth from Nalaikha are larger than in E. grevyi . Besides, the wide range of their protocone length implies the presence of more than one taxon as already noted by Kuznetsova (1996) and Kuznetsova & Zhegallo (1996). The sorting into separate taxa is however very difficult. Morphologically, one upper series ( Fig. 10B ) resembles extraordinarily a Grevy’s zebra ( Fig. 10A ). One rather small premolar PIN 3747-606 ( Fig. 10C ) resembles modern hemiones by its deep post-protoconal groove. Two associated M2 and M3 ( Fig. 10D ) have long, flat, and wide protocones, such as found in some caballine horses like the type of E. lambei Hay, 1917 from the late Pleistocene of Yukon, the middle Pleistocene E. orientalis ( Lazarev, 1980: pl. VII-3), and the lectotype ZIN 3966 of the lower-middle Pleistocene E. nordostensis Russanov, 1968 ( Tcherski 1893: pl. VI-3; Lazarev 1980: pl. VII-2). We cannot presently decide which belonged to E. nalaikhaensis .

The upper cheek teeth of E. graziosii ( Azzaroli 1979: pl. 2) and E. apolloniensis ( Koufos et al. 1997: pl. I) resemble hemiones and some asses by their deep postprotoconal grooves and lack of plis caballins. The teeth of E. coliemensis ( Lazarev 1980: pl. VII-1) are altogether different. The postprotoconal groove is deep but the plis caballins are very well developed. On P3 and P4 they have the same wide base as some teeth of Akhalkalaki and Süssenborn ( Musil 1969: pls XXXIX-1, 3, 6 and others). This pattern was not observed in the Nalaikha sample.

LOWER CHEEK TEETH

They are also variable in size and morphology. Some have a very deep vestibular groove ( Fig. 10F ) like at Süssenborn ( Musil 1969: pls XXXVIII-3, 4). Several have a very shallow lingual groove and a very long bilobated metaconid ( Fig. 10G, H ). Long and bilobated metaconids are frequent in hemiones, and are also found in E. granatensis Alberdi & Ruiz Bustos, 1985 of the lower Pleistocene of Venta Micena ( Eisenmann 1999: pl. I-7). Some molars have a shallow vestibular groove ( Fig. 10J ) like in modern asses. Some may be caballoid. As for the upper cheek teeth, we do not know which belonged to E. nalaikhaensis .

The lower teeth of E. apolloniensis APL 570 ( Koufos et al. 1997: pl. III-2) are absolutely ass-like: rounded double knots, shallow ectoflexids on molars. There are no lower cheek teeth associated with the type skulls of E. graziosii or E. coliemensis . Azzaroli (1979: pl. 7) refers to his species a lower mandible that looks rather caballine. Lazarev (1980: pl. VIII-1) refers to E. coliemensis a hemione-like worn series lacking the m2.

THIRD METACARPALS

We refer to E. nalaikhaensis eight relatively small metacarpals (Annexe: Table 3). Compared to E. hemionus onager Boddaert, 1785 (reference line in Fig. 11 ) E. nalaikhaensis has a wider diaphysis. Among fossil metacarpals, the less different are those of E. hipparionoides of Akhalkalaki and E. altidens of Süssenborn.

THIRD METATARSALS

We refer to E. nalaikhaensis eight relatively small metatarsals (Annexe: Table 4) and a few small fragments. Compared to E. hemionus onager ( Fig. 12 ), E. nalaikhaensis has a wider and deeper diaphysis. Among fossil metatarsals, the less different are again those of E. hipparionoides of Akhalakalaki and E. altidens of Süssenborn.

FIRST PHALANGES

Most of the first phalanges of Nalaikha have about the same lengths. To sort out the phalanges of E. nalaikhaensis , we have used the relation existing between metapodial distal articular breadths and first phalange minimal diaphysis breadths ( Fig. 13 ). We have plotted individual measurements for extant species (where limb bones are associated), and average values for fossil species. We refer to E. nalaikhaensis two anterior and six posterior relatively small first phalanges (Annexe: Table 5).

Their proportions are illustrated by ratio diagrams ( Fig. 14 ). One of the anterior phalanges resembles those of some African wild asses while the other is more similar to one specimen of Onager. Among the first posterior phalanges, one is especially interesting to consider because it is associated with the type skull (PIN 3747-500). The ratio diagram shows its resemblance to one Grevy’s zebra and one African wild ass. Another posterior phalanx is rather similar to an Onager.

SECOND AND THIRD PHALANGES

Second phalanges were sorted into two size groups according to the proximal breadth. We refer to E. nalaikhaensis eight specimens (Annexe: Table 6). One posterior phalanx is associated with the type skull (PIN 3747-500). Three other have about the same smallest breadth. The remaining four are considered as anterior.

Most third phalanges may be referred to E. nalaikhaensis . The largest (Annexe: Table 7) could belong to Equus sp. A. Anterior were distinguished from posterior on the basis of their greater breadths.

OTHER LIMB BONES

We refer to E. nalaikhaensis a proximal fragment of radius (diaphysis breadth: 36 mm, proximal

maximal width: 73.1 mm, proximal articular width: about 71 mm, antero-posterior articular depth: 37.1 mm) and a distal fragment possibly of the same individual (distal maximal width: about 80 mm, distal articular width: 66.5 mm, distal anteroposterior depth: 40.3 mm).

LIMB SEGMENTS PROPORTIONS

The small amount of data does not warrant definitive conclusions. It seems, however, that E. nalaikhaensis was a cursorial species with relatively narrow third phalanges. It differs from all extant zebras (less cursorial) and asses (very long posterior first phalanges and very narrow third phalanges). It differs also from extant hemiones and from E. apolloniensis by the longer MTIII relative to MCIII and first phalanges.

BODY MASS

There is no satisfactory way to estimate the body mass of all and any kind of equid by using its cheek teeth ( Alberdi et al. 1995) or its limb bones dimensions. That is because different species do not plot on the same regression lines. Applying the equations proposed by Eisenmann & Sondaar (1998), the body mass of an equid “in general” can be calculated by using one of the following formulas:

Ln body mass = -4.525 + 1.434 (Ln of the product of MC10 by MC13).

Ln body mass = -4.585 + 1.443 (Ln of the product of MT10 by MT13).

(Ln is the natural logarithm; MC10 and MT10 are the supra-articular distal widths of the third metacarpal and metatarsal, MC13 and MT13 the distal minimal depth of their medial condyles; measurements in mm; body mass in kg).

These formulas are acceptable in a broad way, giving correct estimations for donkeys, domestic horses, and plains zebras, but they overestimate the weights of Przewalski horses and underestimate the weights of mountain and Grevy’s zebras. In the case of E. nalaikhaensis , the body weights calculated using metacarpals and metatarsals are respectively 381 and 371 kg.

ECOLOGICAL PROBLEMS

A long chain of assumptions is used to reconstruct environments from fossil equids. Previous studies have shown the relation that may be established between muzzle proportions and ways of feeding ( Solounias & Moelleken 1993; Dompierre & Churcher 1996; Eisenmann 1999). Roughly speaking, square and broad shapes belong to grazers, long and narrow shapes to selective browsers. There are, moreover, numerous intermediary forms that may indicate mixed feedings. Recent grazers like hemiones and Przewalski horses usually feed on abrasive plants that grow in open dry landscapes. Their teeth have high crowns and simple enamel pattern. The proportions of their limb segments are very cursorial.

Palynological data on Nalaikha ( Zhegallo et al. 1982) give evidence of mixed forest, bushes, forest-prairie, and prairie-steppe vegetation pointing to a mild, humid climate. The long and narrow muzzle of E. nalaikhaensis suggests it was a browser. But, if we were right in the referral of the limb bones, it was a rather cursorial species, which is somehow surprising.

Figures

FIG. 5. — Type skull of Equus nalaikhaensis Kuznetsova & Zhegallo, 1996 (PIN 3747-500); A, profile; B, dorsal view; C, ventral view. Scale bar: 10 cm.

FIG. 6. — Basicranial measurements in Equus; 1, basilar length; 2, overall palatal length; 2-5, palatal length; 3, vomerine length; 4, postvomerine length; 5, muzzle length.

FIG. 7. — Ratio diagrams of skulls of Equus nalaikhaensis Kuznetsova & Zhegallo, 1996, extant Grevy’s zebras, and late Pleistocene E. graziosii Azzaroli, 1966 of Italy. Measurements 16, 23, 3, etc. are defined in Annexe (Table 2).

FIG. 8. — Scatter diagram of frontal width versus basilar length in extant Equus and in E. nalaikhaensis Kuznetsova & Zhegallo, 1996 and E. coliemensis Lazarev, 1980.

FIG. 19. — Ratio diagrams of Equus skulls. Measurements 16, 23, 3, etc. are defined in Annexe (Table 8).

FIG. 9. — Scatter diagram of upper cheek teeth occlusal diameters (occlusal length + occlusal width)/2, observed inside the Nalaikha sample compared to the range observed in extant Equus grevyi Oustalet, 1882.

FIG. 10. — Upper and lower cheek teeth of equids of Nalaikha; A, upper right cheek teeth series of Equus grevyi Oustalet, 1882 (23-10-20-16, British Museum); B, upper right associated P2-M3 (PIN 3747-522, 600, 358, 402, 362, 395); C, upper left P3 or P4 (PIN 3747-606); D, associated upper left M2 and M3 (PIN 3747-457, 458); E, lower right p3 or p4 (PIN 3747-491); F, lower left m1 or m2 (PIN 3747-345); G, lower left p3 or p4 (PIN 3747-394); H, lower left m1 or m2? (PIN 3747-300); I, lower p3 or p4? (PIN 3746-546); J, lower m1 or m2 (PIN 3747-604). Scale bar: 2 cm.

FIG. 11. — Ratio diagrams of third metacarpals of Equus. Measurements 1, 3, 4, etc. are defined in Annexe (Table 3).

FIG. 12. — Ratio diagrams of third metatarsals of Equus. Measurements 1, 3, 4, etc. are defined in Annexe (Table 4).

FIG. 13. — Scatter diagram of minimal breadths of first phalanges versus metapodial distal articular breadths. Abbreviations: A., Allohippus; E. suess., Equus suessenbornensis; E. hipp., E. hipparionoides.

FIG. 14. — Ratio diagrams of first phalanges. Abbreviations: BL, Zoologisches Museum der Humboldt Universität, Berlin; HA, Zoologisches Institut und zoologisches Museum, Hamburg; MU, Zoologische Sammlung des Bayerischen Staates, Munich. Measurements 1, 3, 4, etc. are defined in Annexe (Table 5).