DISCUSSION

The India-Asia collision event is the best example of continent-continent collision and has attracted wide attention in recent years with respect to the biota present before and during the collision (Jaeger et al. 1989, Sahni and Bajpai 1991, Rage 1996, Prasad and Sahni 1999, Briggs 2003, Sahni 2006, Whatley and Bajpai 2006). In this context, the Late Cretaceous - Early Eocene terrestrial vertebrate assemblages have an important bearing on our understanding of the degree of isolation of the Indian subcontinent and the resulting endemism and possible dispersal routes, if any, during this time. Investigations of the Upper Cretaceous Deccan intertrappean biota have long shown the presence of a mixture of Laurasiatic (Sahni 1984, Jaeger et al. 1989, Sahni and Bajpai 1991, Prasad and Rage 1991, Prasad and Sahni 1999) and Gondwanic (Krause et al. 1997, Prasad and Sahni 1999) affinities. In recent years, however, a considerable degree of endemism has been demonstrated for the Maastrichtian continental biota of India based on freshwater ostracods, with about 100 new species having been recorded (e.g., Whatley and Bajpai 2000, 2005, 2006). Biotic endemism during the northward journey of India was also demonstrated by the extant Nasikabatrachus sahyadrensis (family Nasikabatrachidae), the sister group of the family Sooglossidae, the extant frog family of the Seychelles (Bossuyt and Milinkovitch, 2001, Dutta et al. 2004, Karanth 2006).

The continental vertebrate fauna of the Vastan Lignite Mine, which comprises the southeastern extension of the Ghazij Formation of Baluchistan (Pakistan), is highly significant from a paleobiogeographic point of view because it comes from a point in time during which the Indian plate was docking with Asia. During the Early Eocene, and possibly prior to the firm suturing between India and Asia, changes in ecological niches may have led to the origin and endemic evolution of certain groups of mammals such as quettacyonid condylarths (Gingerich et al. 1997), cetaceans (Bajpai and Gingerich 1998), cambaythere perissodactyls (Bajpai et al. 2005b, 2006a), and possibly anthracobunid tethytheres (Ginsburg et al. 1999). Krause and Maas (1990), in a seminal paper, hypothesized that certain mammalian groups such as artiodactyls, perissodactyls, and primates, which appeared at the Paleocene-Oligocene boundary (ca. 55 Ma) in the northern hemisphere, dispersed out of India when it made subaerial contact with Asia. More recently, this 'Out-of-India' dispersal hypothesis has been proposed on the basis of molecular data collected from many of the modern faunal and floral groups, such as ranid frogs (Bossuyt and Milinkovitch 2001), acrodont lizards (Macey et al. 2000), ratite birds (Cooper et al. 2001), and Crypterioniaceae plants (Conti et al. 2002). These studies suggest that the origin of these and many other animal and plant lineages occurred in the Indian subcontinent or other Gondwanan landmasses, and that they were rafted northwards on the drifting Indian plate, eventually spreading to the northern landmasses (Asia, Europe and North America) as a consequence of the contact between India and Asia. At present, there is limited fossil evidence for the 'Out-of-India' dispersal hypothesis: which includes ranoid frogs (Prasad and Rage 2004), freshwater ostracods (Whatley and Bajpai 2006) and Lagerstroemia (Lythraceae, Liu et al. 2007) from the Deccan intertrappean beds of India. However, the 'Out-of-India' hypothesis has been contested by Clyde et al. (2003) who, based on their study of the Early Eocene Ghazij mammal fauna of Pakistan, suggested that this fauna dispersed 'In-to-India' rather than out of it. This idea is based on their interpretation that the mammal faunas of the lower and middle parts of the Ghazij Formation are relatively endemic, while in its upper part the fauna becomes dominantly holarctic.

The new dental material described here from the Lower Eocene Vastan Lignite deposits resembles the dental morphology of Tikiguania and Tinosaurus in having long jaws with an anteroposteriorly elongated, elliptical symphyseal face and three pleurodont teeth. Before the present discovery, Tinosaurus was reported only from the holarctic landmasses (North America, Europe, Kazakhstan, and China). Only two fragmentary teeth of Tinosaurus sp. were documented from the early Middle Eocene Kuldana Formation, Kohat, Pakistan (Rage 1987). The specimens described here are the oldest well-preserved jaws of agamid lizards from the Cenozoic rocks of South Asia. The fact that Tinosaurus occurs in the fossil record as early as in the Paleocene might suggest that Tinosaurus had originated somewhere in the Northern Hemisphere and dispersed in to India during the earliest Eocene or even Late Paleocene. Although the timing of the initiation of the India-Asia collision is controversial, most estimates vary between 65 Ma to approximately 50 Ma. Thus, the Vastan agamids, if they were immigrants from the north, document a dispersal event that occurred either after the establishment of well-developed dispersal corridors following a firm suturing between India and Asia, or via a series of intermittently emergent crustal blocks in the Neotethys sea (including island arcs), that provided a continuous terrestrial connection required for faunal interchanges before the two plates collided.

However, the oldest agamid lizard, Tikiguania, comes from the Late Triassic of India (Datta and Ray 2006). The close similarity in the development of central and lateral cuspules on the posterior teeth of Tikiguania, Vastanagma susani gen.et sp.nov. and T. indicus sp. nov., the presence of a broad, flat or convex, platform-like subdental ridge in Tikiguania and Vastangama, and the antiquity of Tikiguania suggest that the agamid lizards from Vastan may have been derived from a Tikiguania-like animal. The resemblance between T. indicus sp. nov. and Tikiguania in possessing caniniform teeth anteriorly and tricuspid pleuroacrodont teeth posteriorly, may offer an alternative explanation that the incipient development of labial cuspules in Tikiguania and T. indicus sp. nov. is the plesiomorphic state and that the well separated labial cuspules with deep vertical grooves and with reduced height difference between these cuspules and central cusp represents the derived state seen in North American and Eurasian species of Tinosaurus. This interpretation leads to the conclusion that Tinosaurus descended from the Late Triassic Tikiguania of India and dispersed out of India into Eurasia and North America when the initial land connection was established between India and Asia, or possibly even earlier by trans-Tethyan dispersals.

The assignment of some of the Vastan specimens to Tinosaurus is primarily based on the presence of tricuspid acrodont and pleuroacrodont teeth, and a narrow and cylindrical subdental ridge between the alveolar margin and the Meckelian fossa. However, as mentioned before, the holarctic species of Tinosaurus differ from the Vastan species T. indicus sp. nov. in the development of central and labial cuspules. In contrast to the North American and Eurasian species, T. indicus sp. nov. has incipiently developed labial cuspules lacking marked vertical grooves separating them from the central cusp, and greater height difference between the central and labial cuspules. Because this character is variably present in different species of Tinosaurus and extant agamids, it cannot be used as an apomorphy. Moreover, Tinosaurus is a poorly defined taxon and the monophyly of its component species is not established with apomorphies. For the convenience of description, however, the Vastan specimens (T. indicus) are tentatively assigned to Tinosaurus until better preserved material is available.

Estes (1983) proposed that the stem group squamates had a Pangaean distribution and that the split between southern iguanians and northern scleroglossans took place as a result of the separation of Laurasia and Gondwana during the Jurassic and subsequent land connections permitted north to south migration of these groups. However, some later discoveries of agamid lizards from the Late Cretaceous of Asia (Alifanov 1989, 1993, Borsuk-Bialynicka and Alifanov 1991, Gao and Hou 1995, Gao and Nessov 1998) have questioned the Gondwanan ancestry for Iguania, favoring a Laurasian center of origin for the iguanians. In their detailed study of the phylogenetics of modern acrodont lizards by molecular data, Macey et al. (2000) identified a number of modern agamid clades and suggested a Gondwanan origin for them. According to these authors, all of these agamid clades were introduced into Asia by the accretion of the Gondwanan plates at different times. The molecular studies were subsequently supported by new fossil discoveries (Bharatagama, Evans et al. 2002) from the Middle to Upper Jurassic Kota Formation and from the Upper Triassic Tiki Formation (Tikiguania, Datta and Ray 2006). The new fossil data from Vastan does not unequivocally support the 'Out-of–India' hypothesis, but we cannot rule out the possibility that some lineages of acrodont lizards, such as the new agamid lizard material described here, were introduced into Laurasia from India during one of the several trans-Tethyan dispersal phases (Macey et al. 2000, Gheerbrant and Rage 2006). Because the fossil record from the Gondwanan landmasses is scanty and incomparable to that from Laurasia, a clearer paleobiogeographic picture will only emerge when the fossil record from these landmasses becomes better represented.