Order
Artiodactyla
Suborder
Ruminantia
Infraorder Tragulina
Lophiomerycidae Janis, 1987
Genus
and Species Indeterminate
Locality Z 108: Z 760, fragment of a left P3; Z 761, fragment of a very worn upper right molar; Z 2277, fragment of a right p3; Z 2278, left mandible with m1-m2; Z 2299, right astragalus.
Known from only one Zinda Pir site, Locality Z 108 in the lower part of the Chitarwata Formation. Late Oligocene and approximately 25 Ma (Lindsay et al. this issue).
The mandible, Z 2278 (Figure 2.1-2.4) is the only
identifiable specimen. It has the broken posterior root of dp4, intact crown of
m1, and anterior half of a newly erupted m2. Identification of the position of
the teeth is based on the shape of the jaw margin posterior to the m2.
The lower molars are brachyodont, with smooth enamel, distinctly selenodont protoconid and hypoconid, and a cuspidate metaconid. On the m2 the unworn protoconid is much taller than the metaconid, but the slightly worn hypoconid of the m1 appears to have originally been about the same height as the entoconid. The trigonid is distinctly narrower than the talonid (trigonid: talonid = 0.89), giving the molars a slightly trapezoidal aspect in occlusal view. The anterior face of the metaconid is rounded and lacks any hint of a premetacristid. The antero-lingual face of the protoconid is also rounded, with the preprotocristid sharply differentiated. After first descending at a steep angle, the preprotocristid turns abruptly lingually and extends around the anterior of the tooth as a low crest (“paralophid”), reaching the lingual margin and having a barely discernable twinned cuspule at its anterior. The large anterior fossette has a broad lingual opening. The entoconid is slightly anterior to the hypoconid, and its posterior face is rounded, without a postentocristid. The antero-lingual face of the entoconid is flat, without an entoconid groove (“Zhailimeryx-fold”). The hypoconid has a broad and shallow vertical groove on its posterior face. The posthypocristid extends to the lingual margin of the tooth, leaving a narrow opening to the posterior fosset, but without forming a distal tubercle (variously interpreted in closely related forms as an entostylid or hypoconulid).
There is a distinct “M” shaped structure (“Dorcatherium‑fold” in the broad sense) on the posterior of the trigonid of both the m1 and m2 (Figure 2.4). The metaconid leg of the “M” (“Dorcatherium‑fold” in the sense of Janis 1987) is formed from a deep cleft with the resulting lingual fold reflected labially to form a thickened pillar-like structure. The cleft starts at the unworn tip of the metaconid and closes before reaching the base of the cusp. The labial leg on the protoconid (the “Tragulus-fold”) is formed from an equally deep groove with a reflected and thickened labial margin forming a ridge connecting to the prehypocristid. The longer postprotocristid and shorter postmetacristid (the two cross connections of the “M”) are subhorizontal, coming together at a very shallow angle and joining ventrally with a vertical crest that connects with the preentocristid. The postprotocristid, postmetacristid, and vertical crest together form a “Y” shaped structure on the posterior of the trigonid, with the vertical stem being as long as the diverging arms. At the base of this structure where it contacts the preentocristid there is a very small vertical fold in the entoflexid and another on the labial side making a connection to the “Tragulus‑fold”‑prehypocristid complex. Finally, there is a strong anterior basal cingulum that arises at the base of the preprotocristid and then passes around the front of the tooth to form a lingual tubercle, a very strong and high posterior basal cingulum that extends labially, and a low bulbous “ectostylid” that is probably a vestige of a labial cingulum.
The upper molar (Z 761) is very worn and possibly is an m1. It preserves a faint lingual cingulum, extending around the protocone. The two premolars are fragments of the anterior of a P3 (Z 760) and posterior of a p3 (Z 2277). The three specimens preserve little by way of distinctive morphology and are assigned to this taxon on the basis of size.
This taxon differs from both extant and extinct tragulids (including Dorcabune Pilgrim 1910) in the development of the preprotocristid and absence of the premetacristid, as well as in the length of the common stem of the postmeta- and postprotocristid complex and the closing of the “Dorcatherium‑fold” before it reaches the base of the metaconid.
Archaeotragulus krabiensis Métais, Chaimanee, Jaeger, and Ducrocq 2001, was described as a tragulid, but shares some similarities with the Zinda Pir mandible not otherwise seen in tragulids. These principally involve the formation of the trigonid, in that the preprotocristid of Archaeotragulus krabiensis is developed as a substantial antero-lingually directed loph that, in combination with the absence of a premetacristid, leaves the trigonid open lingually. In addition, the postprotocristid and postmetacristid join at a shallow angle to form a short vertical stem that passes down the rear of the trigonid to connect to the preentocristid. There are, however, differences that in our opinion preclude assigning the Zinda Pir specimen to Archaeotragulus. By comparison to the Zinda Pir mandible, the lingual cusps of Archaeotragulus krabiensis are more bulbous, and the unworn trigonid cusps are more equal in height; the preprotocristid is more lingually directed, and the anterior fosset is less open; the “Dorcatherium-fold” extends to the base of the metaconid and the postprotocristid and postmetacristid join at a more acute, steeper angle with a shorter vertical stem; and there is no sign of a “Zhailimeryx-fold” on the entoconid nor a distal swelling on the posthypoconid.
The absence of the postentocristid and premetacristid (producing a “figure-8” morphology), together with strong anterior and posterior cingula and a wide, lingually open anterior fossette are characters of taxa placed in the Lophiomerycidae by Janis (1987) and others (Guo et al. 2000; Métais et al. 2001). To this character list we would add 1) the “Y” shaped configuration of the postprotocristid and postmetacristid, with a long common stem connecting to the preentocristid; 2) the relative heights of the protoconid and metaconid and the relative lengths of the postprotocristid and postmetacristid; and 3) the distinctive form of the “Dorcatherium‑fold,” which is confined to the tip of the metaconid cusp and does not extend to its base. The presence of these characters in the lower molars of the Zinda Pir form indicate it is a lophiomerycid.
Of currently recognized lophiomerycids, Zhailimeryx Guo, Dawson and Beard 2000 and Krabimeryx Métais, Chaimanee, Jaeger, and Ducrocq 2001 are similar to the Zinda Pir specimen in possession of the typical lophiomerycid features, but differ in many details. Most notably, differences include the presence of a well defined “Zhailimeryx-fold” on the anterior face of the entoconid, the apparent absence of a well defined “Tragulus-fold” connected to the prehypocristid, and the weakness or absence of a bulbous ectostylid. In addition, in neither of these two taxa does the preprotocristid extend as far lingually as in the Zinda Pir tooth, although in Zhailimeryx there is an anterior cusp described by Guo et al. (2000) as a “rudimentary paraconid” that might correspond to the “twinned cuspule” observed on Z 2278.
Both Lophiomeryx Pomel 1853 and Iberomeryx Gabunia 1964 (including Cryptomeryx Schlosser, 1886) share the characteristic lophimerycid trigonid features, and both lack the “Zhailimeryx-fold” on the entoconid, a similarity shared with Z 2278. Nevertheless, Iberomeryx and especially Lophiomeryx differ substantially from the Zinda Pir specimen. European and perhaps Asian species of Lophiomeryx are more hypsodont and lack any vestiges of a “Tragulus-fold” or of basal tubercles on the labial side of the tooth, while having a more open posterior fossette, a distinct cuspule terminating the posthypocristid, and an incomplete postentocristid that extends from the unworn tip of the entoconid partway down the cusp, leaving the entoconid base rounded. In addition, in Lophiomeryx the antero-lingually directed prehypocristid joins the preentocristid‑postmetacristid complex, rather than being more medially terminated.
Species of Iberomeryx have bulbous basal tubercles on the labial side of the tooth (interpreted as ectosylids, but possibly a labial cingulum remnant), a weak “Tragulus-fold” variably developed on anterior molars, and, also variably, the prehypocristid connected directly to the trigonid rather than to the preentocristid-postmetacristid complex (Nanda and Sahni 1990; Métais et al. 2001). These are similarities shared with the Zinda Pir mandible and suggest Iberomeryx is the mostly closely related of known taxa. The Zinda Pir specimen, however, differs from the better-known species of Iberomeryx such as I. minus and I. parvus in the strength of the “Tragulus-fold,” which on both its m1 and m2 is very well formed with a thickened labial margin contacting the anterior end of the prehypocristid. The Zinda Pir mandible also has the terminal end of the prehypocristid more lingually directed, and there is no trace of a cuspule on the lingual termination of the posthypocristid.
The enigmatic Iberomeryx savagei Nanda and Sahni 1990 is known from a few specimens of Late Oligocene age from the Kargil Formation of northern India. While too small to be the same species, it appears to also have had a well-formed “Tragulus-fold” contacting the prehypocristid and a lingually extended preprotocristid.
The Tragulina are regarded by most current workers as a paraphyletic assemblage, and the lophiomerycids may also be paraphyletic, with some genera having closer relationships to the Tragulidae (Métais et al. 2001). As currently recognized, the family contains Middle Eocene to Late Oligocene species.
Métais et al. (2001) suggested the Kargil species represented a new genus, and it is likely the Zinda Pir specimens are a closely related species. Forster-Cooper (1915) based Gelocus indicus from Dera Bugti on a single upper molar. His specimen, which is taken to be a third molar, has a strong lingual cingulum and very small metaconule. It is of appropriate size for an upper molar matching Z 2278 and might be the same species.
The “M” structure (or “Tragulus-fold”) is thought to be a uniquely derived character of tragulids (Métais et al. 2001) or tragulids and leptomerycids (Geraads et al. 1987). The clear presence of the structure in combination with a lingually extended preprotocristid and absence of a premetacristid suggests it may be a primitive feature shared by both early tragulids and lophiomerycids.
