1 Introduction

Compared with the footprints of Mesozoic dinosaurs, Cenozoic vertebrate traces are still less studied and often do not attract the same attention of ichnological research. However, in recent years, their potential for paleobiogeography, paleoecology, and biostratigraphy was recognized, and a growing number of papers that focus especially on mammal tracks from the Tertiary and Pleistocene (Vialov 1965, 1966; Demathieu et al. 1984; Kordos 1985; McCrea et al. 2004; Hunt and Lucas 2007; Lucas 2007; McDonald et al. 2007) have been published. From the Tertiary, most records come from the Neogene, whereas Paleogene localities are less abundant (McDonald et al. 2007). Important surfaces with footprints of brontotheres, tapirs, rhinoceroses, ancient horses and other ungulates as well as bird tracks from the Eocene–Oligocene are known from North America, South America, Europe and China (Lockley and Hunt 1995; Lockley et al.1999; Lockley and Meyer 1999; Xing et al.2013). Some trackbearing units of South America and China were originally mistaken for Mesozoic deposits, but their stratigraphic age is now identified as Cenozoic (Lockley et al. 1999; Melchor et al. 2002, 2013; DeValais and Melchor 2008). Thus far, the Cenozoic footprint record from India is scarce. From the Siwaliks of India, footprints of Miocene artiodactyls have been described by West et al. (1983). Here we describe the first mammal (perissodactyl) footprints from Lower Oligocene Flysch deposits of the Barail Group in Manipur, India. The ichnofossils supplement the skeletal record of perissodactyls from the Paleogene of South Asia that comprises numerous tapiromorph forms (Antoine et al 2003; Holroyd et al. 2006; Missiaen and Gingerich 2012).

2 Geological setting

The Indo–Myanmar Ranges (IMR) is thought to be the northern elongation of the Indonesian island arc. It is an arcuate shaped tectonic belt having a convexity towards west. Brunschweiler (1966) described the region in terms of three major lithostratigraphic units: the Naga Metamorphic Complex, the Naga Hills Flysch and the Upper Chindwin Molásse. The Naga Hills is the northernmost segment of the IMR and trending nearly NE–SW. The segment mainly consists of the Naga-Patkoi Hills of Nagaland and the northern part of the Manipur Hills. Acharyya et al. (1986) described the geological and tectonic setting of the Naga Hills in terms of two distinct longitudinal belts; the Central Naga Hills Palaeogene Flysch sediments and the Naga-Chin Hills Ophiolite Belt. The Central Naga Hill Palaeogene Flysch sediments, which can simply be referred to as the Flysch Belt are principally made up of sediments of Disang and Barail groups, constituting the western part of the Naga Hills segment. Most of the lithostratigraphic units in the region occur in the form of an imbricate thrust system where older lithostratigraphic units lie above the younger ones (Soibam 2000; Soibam and Pradipchandra 2006). The principal lithostratigraphic units of the Flysch Belt in the state of Manipur are composed of turbidite sediments of the Disang and Barail groups and molásse sediments of the Surma and Tipam groups, respectively (table 1). The Disang Group (Upper Cretaceous to Upper Eocene) consists of a monotonous sequence of dark grey to black splintery shales, and occasional rhythmites of shales and siltstones/fine-grained sandstones, forming the principal lithostratigraphic units of the eastern half of the state of Manipur (23°50′−25°55′ N latitudes and 093°00′−094°45′E longitudes). The Barail Group, mainly Oligocene in age, with a possible range to the Upper Eocene, consists of \(\sim \)3000 m thick arenaceous beds intercalated with shales and carbonaceous shales, silty shales and siltstones intercalated with thin shale beds and sandstones that rest on the Disang Group. The occurrence of Barail sediments in the state mainly starts from the western foothill regions of the Imphal Valley (figure 1). The basal part of the Barail Group has a gradational (with local tectonic) contact with the Disang Group (Soibam 2000). This gradational contact is related with a gradual change from dominantly argillaceous deep marine to a mainly arenaceous shallow marine depositional environment (Guleria et al. 2005). The Barail Group has been divided into three formations namely the Laisong, Jenam, and Renji (Evans 1932). The Laisong Formation, about 1200 m thick, is characterized by alternations of shale, silty shale, siltstone and fine to medium grained sandstones, giving rise to a typical turbidite sequence (figure 2). This formation contains abundant and varied sedimentary structures such as ripple marks, undulating bedding, cross-stratification, hummocky cross-stratifications, herringbone structures, etc., indicating wave and current influence, and deposition above storm wave base. The studied section reflects a transition zone and could represent a large fluctuation in the sedimentation rate. Therefore, sedimentary facies deposited above storm wave-base comprises large sedimentary structures, otherwise they are characteristic of sediments formed below the storm wave-base level. The footprints described herein come from a silty shale near the base of the Laisong Formation at the Gelmon locality (23\(\circ44^{\prime }19.9^{\prime \prime }\mathrm {N}-092\circ40^{\prime }16.5^{\prime \prime }\)E, elevation of 905 m). The Laisong Formation section at Gelmon locality may be considered to be of Lower Oligocene age taking reference to earlier studies carried out in the state in sections that belong to Upper Disang (Middle to Upper Eocene) and Lower Barail (Upper Eocene to Oligocene) transition sections. For example, Kachhara et al. (2000) treated the Biozone B at Thoubal (Kabrang Hill, 24\(\circ39^{\prime }67^{\prime \prime }\mathrm {N}-094\circ00^{\prime }57^{\prime \prime }\)E; Kaina Hill, 24°41′20″N−094°01′31″E) as the lowermost lithological contact of the Barail Group (Laisong Formation) with the Upper Disang Formation based on bivalves such as Cardium, Chlamys, Lucina, Ostrea, Trigonia, Pinna, Spondylus, Tellina, Pecten and Turritella. However, a proper age to this formation may only be assigned after proper dating of the sediments, which has not been done so far.

Table 1 Stratigraphic succession of Manipur (modified after Soibam 2000).
Full size table
Figure 1
figure 1

Geological map of Manipur (northeast India) indicating study (track) site.

Full size image
Figure 2
figure 2

Lithostratigraphic section showing the position of the perissodactyl track horizon.

Full size image

3 Tetrapod footprints

3.1 Material

Two successive footprints of a trackway catalogued as MSM-1 and displayed in the Manipur State Museum (figure 3). A cast of the specimen has been made using modelling clay and is kept at the Geological Museum, United College, Manipur.

Figure 3
figure 3

Pes tracks of a perissodactyl mammal (MSM-1) from the base of the Laisong Formation (Barail Group, Lower Oligocene) of Manipur (northeast India) preserved as concave epireliefs. (A) Photograph of footprints (i) left and (ii) right from partial trackway. (B) Cast of (A) where ia and iia are respectively the corresponding casts of tracks i and ii in (A). (C) Sketch of the partial trackway. (D) Perissodactyl pes imprints from the Paleogene of Xinjiang Province, China for comparison (after Lockley et al. 1999). (E) Close-up of pes i in (A); notice the leaf impression on the top left corner of (E). (F) Cast of (E).

Full size image

3.2 Locality and horizon

Silty shale beds with lingoid ripples and scour marks, about 30 mm thick, at Gelmon locality (Manipur, India). Track-bearing layer close to the base of the Lower Laisong Formation of the Barail Group.

3.3 Description

Two tridactyl imprints, probably those of the pes, representing a single step (right–left) of a trackway (pace length = 18.6 cm) (figure 3A). They have been preserved as moderately impressed concave epireliefs. Their overall shape is mesaxonic with digit trace III being longest and projecting more anteriorly compared with lateral digits. The first imprint (figure 3A-ii) is 6.3 cm in length and 8.2 cm in width (length/width \(=\) 0.8), the better preserved second imprint (figure 3A-i), 7 cm in length and 7.7 cm in width (length/width \(=\) 0.9). The digit traces are moderately spread. Their divarication was measured in the well-preserved second imprint with 41° (II–III), 18° (III–IV) and 58° (II–IV). They terminate in broad rounded, hoof-like impressions. In particular, the digit trace III of the second imprint shows three distinct circular to oval impressions of the digital pads and the distal hoof. The digit traces are embedded in a broad oval, faint impression of the sole that shows a posteriorly concave embayment forming a bilobate posterior margin. The embayment is positioned in line with digit trace III. There is no indication of the presence of manus imprints.

4 Discussion

The tridactyl, mesaxonic shape and the hoof impressions indicate a perissodactyl mammal as the trackmaker. Footprints of perissodactyls are known, for example, from Upper Eocene deposits of Trans-Pecos, Texas, USA and have been assigned to tapiroid, rhinocerotoid and brontotheroid group trackmakers (Sarjeant and Langston 1994). Important surfaces with perissodactyl footprints come from the Eocene Chuckanut Formation and the Pudget Group of Washington (USA) (Mustoe 2002; Mustoe and Hopkins 2013). These tracks have been attributed to Hyracotherium-like equids and/or tapiroids, known also by a rich skeletal record from the North American west. Famous localities with perissodactyl and other ungulate footprints are in the Duchesne, Uinta, Green River and Debeque formations of Utah and Colorado (USA) as well as in Paleogene (Eocene–Oligocene) deposits of France, Spain and Peru (Demathieu et al. 1984; Lockley and Hunt 1995; Hamblin et al. 1998, 1999; Lockley and Meyer 1999; see Lockley et al. 1999 for overview). The locality in the Apt Basin of southeastern France which is in Lower Oligocene calcareous lake deposits, shows trackways with typical pes and manus imprints of a rhinocerotoid and was named Rhonzotherichnus (Demathieu et al. 1984). However, the imprints are much larger compared with those from Manipur and the middle digit trace is more broad with a sharp distal edge of the hoof. One of the best localities with Cenozoic tracks, that also contains perissodactyl footprints, is in the lower Miocene of Hungary (Kordos 1985; Szarvas 2007). The well-preserved tracks of rhinocerotoids are also larger than those of the Manipur specimen and the middle digit has a broad-oval shape. From Asia, perissodactyl footprints are known from Eocene and Miocene deposits of Iran (Ataabadi 2007) and especially from the Xinjiang Province of China (Lockley et al. 1999). The tridactyl footprints from Iran show a more extended heel portion with a posteriorly convex margin and are over 20 cm in size. By their shape and size, the imprints described herein resemble the footprints from China. Both are mesaxonic tridactyl with a bilobate posterior margin. The latter are 6 cm in length and 8 cm in width, which matches the measurements given for the Barail specimen above. On the contrary, perissodactyl footprints described from North America and Europe mostly show a posteriorly convex and continuous margin (Lockley et al. 1999; Mustoe 2002). Small tridactyl imprints of perissodactyls have been described from the Oligocene of Spain by Murelaga et al. (2000) and Astibia et al. (2007), but all have an elongate ‘heel’ and a convex posterior margin, whereas the imprints from the Barail Group are relatively short, broad and concave posteriorly. The lack of a manus imprint in the Barail specimen may be due to overstep by the pes which is a common feature in mammal tracks (Mustoe 2002). Costeur et al. (2009) figure footprints of larger rhinocerotoids from the Oligocene of southeastern France, but these have a mediolaterally extended hoof impression on the third digit which is characteristic of rhinoceros tracks. By their overall shape, the perissodactyl footprints from the Barail Group can rather be assigned to small equoids or tapiroids, even if basal rhinocerotoids cannot be excluded. Ichnotaxonomically, perissodactyl footprints have been assigned to different ichnogenera such as Apoxypus, Rhinoceripeda, Plagiolophus, Plagiolophustipus, Lophiopus or Palaeotheripus. In most cases, a differentiation is difficult to make and the ichnotaxonomy will certainly need revision. Also, further prospection at the Gelmon footprint locality and nearby areas might yield larger samples that permit a more detailed determination. Therefore, the footprints from the Barail Group are not assigned here to a distinct ichnotaxon.

5 Conclusions

The footprints from the Laisong Formation of the Barail Group are the first record of Lower Oligocene perissodactyl tracks from India, strongly resembling those previously described from Xinjiang Province of China. Remarkable is the paleoenvironment which is considered as marginal marine. Singh et al. (2012) indicated a tropical to sub-tropical climate for the depositional basin of Laisong sediments. The depositional setting/environment of the Upper Disang–Lower Barail transition sediments was considered as moist tropical with high humidity by Guleria et al. (2005) based on morphologies of Paleogene monocot and dicot plant leaves from Manipur such as Dicotylophyllum kainai sp. nov., Phoenicites indica sp. nov., Amesoneuron manipurensis sp. nov., Dicotylophyllum asymmetricum sp. nov., etc. These taxa differ from any known fossil dicotyledonous floras from various parts of India (Guleria and Mehrotra 1999; Mehrotra and Mandaokar 2000. It is likely that the source material of plant fossils came from the Myanmar landmass rather than from the Indian subcontinent. The similarities between the Laisong tracks and those from Xinjiang Province of China could be correlated to a continuous landmass connection between China and Myanmar during Lower Oligocene time, favouring a likely migration and dispersal of different vertebrates. However, this is only a speculative observation and based on a small sample of footprints, this cannot be proved presently. Interestingly, skeletal remains of middle Eocene perissodactyls (rhinocerotoids) from Myanmar show a relationship with coeval forms from North America and Asia (Holroyd et al. 2006). Vertebrate body fossils from the Indian sub-continent indicate a faunal exchange with other parts of Asia as well as with Europe (Antoine et al. 2003; Rose et al. 2006; Missiaen and Gingerich 2012). Further material and study is needed to explain this conclusively. It has to be noted that the deposition of plant fossils was prior to the subduction of the Indian plate beneath the Myanmar micro-plate (cf. Soibam and Khuman 2008). Therefore, the exchange of terrestrial faunas between the Indian landmass and the Myanmar continental margin was very unlikely at that time. With respect to the palaeoenvironment, Rajkumar et al. (2008, 2012) reported a marginal marine depositional setting for the Laisong sediments and the trackbearing unit based on various invertebrate-ichnofossils. On the other hand, perissodactyl footprints from North America, Europe and China all come from fluviolacustrine deposits. The new discovery from Manipur indicates the wide distribution of the trackmakers that may have been early equoids, tapiroids or rhinocerotoids frequenting the subtropical to tropical Lower Oligocene coasts of the western Myanmar continental margin.