Most of peninsular India exposes Precambrian crust which was formed, reworked and consolidated over several major orogenic events. Two of these events were associated with the assembly and breakup of the supercontinents Rodinia respectively Gondwana in Mesoproterozoic and Neoproterozoic to Early Phanerozoic times. The subcontinent's crustal evolution during these periods is well documented in two high-grade crustal provinces situated at the southern tip (Southern Granulite Terrain) and eastern coast (Eastern Ghats Belt) (Fig. 1).
Unravelling the Proterozoic history of India was long impeded by the very high grade of metamorphism and the strong deformation that most rocks had experienced, and the resulting unclear primary relations between them. In his fundamental treatise of the Precambrian of peninsular India, Fermor (1936) regarded the granulite terrains of east and south India as constituents of a coherent crustal province, the Charnockite Province, that swerves around an older non-charnockitic Province. This notion remained widely accepted with minor modifications (summarized in Gopalakrishnan 1998) until geochronological methods allowed reliable dating of magmatic and metamorphic rocks and to establish the geological significance of the attained ages. The obtained isotope data showed that the Eastern Ghats Belt and the Southern Granulite Terrain underwent pervasive metamorphism at appreciably different times (as reviewed in Jayananda & Peucat 1996 and Mezger & Cosca 1999).
In the case of the Eastern Ghats Belt, the considerable increase of knowledge during the last decade calls for a critical evaluation of new and old data and permits the formulation of an up to date synthesis of its crustal architecture and geological evolution. This offers new prospects for the discussion on the assembly of the supercontinents Rodinia and Gondwana.
The two high-grade crustal provinces skirt the Archaean protocontinental core of peninsular India (Fig. 1), which may be subdivided into the Dharwar, Bhandara (or Bastar), Singhbhum (or East India), and Aravalli (or Bundelkhand) cratons (Naqvi & Rogers 1987). As the latter is not in contact with the Eastern Ghats Belt, its geological evolution will not be described below. The unified cratonic forelands of the granulite terrains are termed Proto-India.
Dharwar Craton. Intense research during the past decades has greatly advanced our knowledge of the crustal make-up and evolution of this important cratonic terrain. Excellent syntheses of existing and new data were provided by Chadwick et al. (2000) and Jayananda et al. (2000). The craton (Figs 1 & 2a) comprises two distinct parts that are separated by a crustal scale N-S trending sinistral shear zone (Chitradurga-Kollegal shear zone). The western part is made up by a polyphase gneissic basement of tonalite-trondhjemite-granodiorite-associations (TTG) with narrow high-grade belts of greenstone-type volcanosedimentary sequences (Sargur Group). The Sargur Group was consolidated between 3.4 and 2.9 Ga before the deposition of a low-grade volcanosedimentary sequence (Dharwar Supergroup) in extended intracratonic basins at 2.9 - 2.6 Ga. The final intrusion of granite plutons occurred at c. 2.6 Ga. The eastern part of the craton consists of a series of N-S trending arc-type plutonic complexes, the Dharwar batholith, emplaced during a major episode of calc-alkaline magmatism at 2.55 - 2.51 Ga. Interdispersed are thin low to medium-grade greenstone-type schist belts, interpreted as intra-arc volcanosedimentary sequences. This major accretionary event led to the final stabilization of the entire craton at c. 2.5 Ga. Chadwick et al. (2000) have inferred an Andean margin setting in which the western cratonic domain forms the foreland to an accretionary arc system represented by the batholith and schist belts. Jayananda et al. (2000) opposed the subduction scenario and argued that the geochemical and isotopic characteristics of late Archaean magmatism in the Dharwar batholith and the associated structural and metamorphic features are best explained in terms of a plume model.
Bhandara Craton. Compared to the neighbouring Dharwar and Singhbhum cratons, the Bhandara craton (Figs 1 & 2a) has not been the target of integrated research in recent times. Its geological evolution thus remains poorly constrained. As in the adjacent cratons, its basement consists of polyphase TTG-gneisses and granitoids with minor infolded schist belts of dominantly intra-cratonic sedimentary sequences (Crookshank 1963; Narayanaswamy et al. 1963). The limited isotopic data indicate a mid-Archaean (c. 3.5 Ga) protolith age for trondhjemitic gneisses and wide spread felsic magmatism and metamorphism at 2.65 - 2.41 Ga (Sarkar et al. 1993, 1994; Choudhary et al. 1996; Singh & Chabria 1999).
Singhbhum Craton. The nucleus of the Singhbhum craton (Figs 1 & 2a) comprises a gneissic basement, voluminous granites, and a low-grade volcanosedimentary cover sequence (Naqvi & Rogers 1987; Mukhopadyay 2001). Deposition of the gneiss precursors occurred after c. 3.5 Ga and was followed by the intrusion of tonalite at c. 3.44 Ga (Sengupta et al. 1996; Mishra et al. 1999). Basaltic melts were emplaced within the sedimentary sequence at c. 3.3 Ga (Bakshi et al. 1987; Sharma et al. 1994) possibly during intense deformation and metamorphism, and prior to the multistage emplacement of the Bonai, Singhbhum, and Mayurbhanj granites, and the deposition of the low-grade volcanosedimentary Iron Ore Group between 3.4 and 3.1 Ga (Mishra et al. 1999). The age of unconformable deposition of a predominantly clastic cover sequence (Singhbhum, Dhanjori, Kolhan Groups) and the emplacement of mafic dyke swarms is not constrained (Mukhopadyay 2001).
Proterozoic intracratonic basins. The eastern margins of the Bhandara and Dharwar cratons are covered by sedimentary sequences of Proterozoic basins (Figs 1 & 2b). The largest and best investigated, the Cuddapah basin, unconformably overlies the granitoid basement of the eastern Dharwar craton (Nagaraj Rao et al. 1987; Raman & Murty 1997). This very low to low-grade metamorphic volcanosedimentary sequence with an estimated aggregate thickness of 9 to 13 km, has been subdivided into a lower Cuddapah Supergroup and an upper Kurnool Group. The Kurnool Group is restricted to the western basin and overlies the Cuddapah Supergroup with a marked unconformity. Both sequences consist of mature clastic deposits and intercalated calcareous sediments. Dolerite dykes and a few lamproites transect the entire Cuddapah Supergroup, while mafic flows and sills, ignimbrites, and ash-fall tuffs occur mostly in its lower portion (Chaudhuri et al. 2002). The Pb/Pb ages of c. 1.8 Ga, interpreted to date uranium-mineralization in stromatolitic dolomites of the basal sequence (Zachariah et al. 1999), provide a minimum age for carbonate sedimentation and comply with a Rb-Sr whole rock age of 1817 ± 24 Ma for a mafic sill from the intermediate part of the sequence (Bhaskar Rao et al. 1995). A K-Ar total fusion age of 1350 ± 52 Ma for micas of the Chelima lamproite (Chalapati Rao et al. 1996) post-dates sedimentation of the Cuddapah Supergroup. The age of sedimentation of the Kurnool Group is not constrained by palaeontological or radiometric data, but the presence of diamondiferous conglomerates at the base of the sequence suggests a minimum age of c. 1090 Ma, the emplacement age of kimberlite pipes in the eastern Dharwar craton (Chalapati Rao et al. 1999), which are regarded as the most likely source rocks (Saha 2001). From west to east, the Cuddapah basin deepens, the grade of metamorphism increases and strain becomes more intense. A marked increase in the intensity of deformation occurs at the Rudravaram line, which separates the weakly deformed (low amplitude open folds, tilted blocks) western sector from the Nallamallai fold belt that is characterized by isoclinal, overturned folds and thrusts (Meijerink et al. 1984). The Kurnool Group may have been deposited only after this deformation (Meijerink et al. 1984).
The basins of the Bhandara craton are less well studied. The major basins are the Chattisgarh, containing the Chattisgarh Soupergroup, and the Indravati containing the Indravati Group (Chaudhuri et al. 2002, and references therein). Both groups represent similar platformal sequences of unmetamorphosed conglomerates, sandstones, shales, locally stromatolitic limestones, cherts and dolomites with an aggregate thickness in excess of 600 m (Chaudhuri et al. 2002). No reliable age data exist to constrain the time of sedimentation, but basaltic dykes with Nd model ages of between 2.1 and 2.9 Ga and probable emplacement ages of c. 750 Ma (Fachmann 2001) transect the entire sequence in the Chattisgarh basin.
Traditionally, the term Eastern Ghats Belt (or Eastern Ghats Mobile Belt, Eastern Ghats Granulite Belt) denotes a contiguous terrain of granulite facies rocks at the northeastern coast of peninsular India. To the north and west, it borders on Archaean rocks of the Singhbhum, Dharwar and Bhandara cratons, and to the southeast it disappears underneath alluvial plains and the Bay of Bengal (Fig. 1). Based on field work by the Geological Survey of India and an earlier concept of Nanda & Pati (1989), Ramakrishnan et al. (1998) compiled a 1:1,000,000 geological map of the Eastern Ghats Belt and proposed its subdivision into four major lithological units, all aligned parallel to the contact with the cratons (Fig. 2a). These units, from west to east, are:
The contact with the Archaean cratons is assumed to be a discrete boundary shear zone, the Eastern Ghats Boundary Fault, in the north and northwest, but a wide "Transition Zone" along most of the western margin (Fig. 2a) has been inferred from the presence of orthopyroxene-bearing rocks within presumably cratonic gneisses (Ramakrishnan et al. 1998). Alternatively, Chetty (1995, renewed 2001) presented a terrane model based on extensive satellite imagery studies. Megalineaments (Fig. 2a), interpreted as shear zones or thrusts (Chetty 1995, 2001; Chetty & Murthy 1994, 1998b; Mahalik 1994; Ramakrishnan et al. 1998; Biswal & Jena 1999; Biswal 2000), outline the contact with the cratonic forelands and separate the belt into several structural domains, postulated to represent suspect terranes. Earlier presented but no more considered conceptions on the litho(-tectonic) divisions of the belt may be taken from the literature (Gopalakrishnan 1998; Ramakrishnan et al. 1998).
Models for the tectonothermal evolution generally include a first event of granulite facies metamorphism and pervasive deformation in the Late Archaean to Early Proterozoic, and a protracted similar event in the Mesoproterozoic followed by the intrusion of late granitoids and alkaline rocks at c. 0.9 Ga (Chetty & Murthy 1998b; Ramakrishnan et al. 1998; Sarkar & Paul 1998). Sarkar & Paul (1998) provide a comprehensive account of geochronological data published until 1994.
Recent research based on adequate isotope and structural data evidently shows that the internal crustal architecture and regional differentiation of the tectonothermal and magmatic events are much more intricate than previously expected. A re-definition of the 'Eastern Ghats Belt' thus seems unavoidable as the granulite terrain appears to consist of several crustal segments with distinct geological histories. The evaluation further shows that granulite facies grade is not a suitable criterion for the discrimination of these segments, as two of them consist of both, granulite-grade and lower grade rock assemblages. In consequence, we propose to abandon the term 'Eastern Ghats Belt' and the above given synonyms. Table 1 lists the classification of Ramakrishnan et al. (1998) and compares it to our new province-based scheme. We define a province as a crustal segment with a distinct geological history. A province may be further divided into several domains characterized by specific features (e. g. lithology, structure, metamorphic grade) which distinguish them from neighbouring domains. Figure 2b shows the extension of the newly defined provinces and their subdivision into domains.
Below we review and discuss recently published data and ideas to develop a model for the geological evolution of each province and of the entire composite orogenic belt exposed in the Eastern Ghats and adjacent areas. The provinces are presented in order of the age of crustal consolidation, starting with the oldest.
Jeypore Province
The newly defined Jeypore Province comprises the northern portion of the 'Western Charnockite Zone' of Ramakrishnan et al. (1998) and extends from southwest of Bhawanipatna (Fig. 3) for approximately 300 km towards the SW along the western margin of the Bhandara craton (Fig. 2b). In the absence of detailed mapping, the southern termination cannot be exactly located but field observation suggest that the province terminates north of Upper Sileru (Fig. 4).As the presence of metasedimentary rocks reported by Ramakrishnan et al. (1998) cannot be confirmed for this sector of the 'Western Charnockite Zone', it appears that the Jeypore Province exclusively consists of granulite facies meta-igneous rocks. The igneous character of the protoliths is established by locally abundant ultrabasic or basic cumulates and xenoliths as well as geochemical data (Subba Rao et al. 1998; Kovach et al. 2001; authors' unpublished data). Rocks of intermediate composition, enderbites and charnoenderbites, dominate. Basic granulites occur as disrupted bands, lenses and xenoliths within the enderbitic rocks as well as massive, regionally extensive bodies (Nanda & Pati 1989). Basic granulites, locally garnet-bearing, form a large complex located between the Machkund (Subba Rao et al. 1998) and Balimela dam sites, and charnockites concentrate around Jeypore (authors' unpublished observation). At places, NW-SE trending, unfoliated fine-grained mafic dykes occur, which show only a low-grade metamorphic overprint.
Subba Rao et al. (1998) concluded from a limited geochemical data set that the igneous protoliths of basic granulites and enderbites originated from poorly fractionated tholeiitic and intermediate magmas, respectively, that evolved through crustal assimilation and fractionation, suggesting emplacement in a rift environment. The dominance of intermediate rocks with calc-alkaline affinity (chemical analyses of Subba Rao et al. 1998, authors' unpublished data), however, points to arc-related magmatism in an Andean-type active continental margin setting.
Archaean Nd model ages of 3.9 - 3.0 Ga (TDM) for enderbites, in combination with strongly retarded Pb isotopic signatures for feldspars, which imply Pb homogenization in Archaean times, suggest a formation of the igneous protoliths before 3 Ga (Kovach et al. 2001; Rickers et al. 2001). Preliminary U-Pb data from complexly zoned zircon grains point to a high-grade metamorphic event at c. 2.8 Ga and indicate the absence of any significant later thermal overprint (Kovach et al. 2001). The age of metamorphism, however, is not yet constrained by independent isotopic data.
Prolific development of orthopyroxene-bearing leucosomes in enderbites and charnoenderbites through dehydration-melting involving biotite and hornblende indicate that peak conditions of granulite facies metamorphism probably exceeded the P-T estimate of c. 790°C and 5.6 kbar obtained from thermobarometry on garnetiferous basic granulites of the Machkund area (Subba Rao et al. 1998). The structurally controlled segregation of leucosomes provides evidence for the contemporaneity of granulite facies metamorphism and deformation.
The deformation history is, however, poorly known as no structural studies exist. Own observations from the sector between Jeypore and Balimela indicate that ductile deformation during granulite facies metamorphism was largely uniform within the massive meta-igneous rocks, giving rise to a weak continuous schistosity and/or a compositional layering. Locally, narrow mylonitic shear zones, containing mineral assemblages indicative of a medium-grade metamorphic overprint, transect the pervasive fabric at acute angles. These shear zones trend parallel to the margins of the Jeypore Province and dip with 40°-50° towards the southeast. A well-developed mineral lineation plunges steeply, and C/S fabrics and asymmetric clasts indicate transport of the hanging wall towards the NW.
Rengali Province
Detailed satellite imagery analyses supplemented by extensive ground truth control have established that the northern part of the granulite terrain of the Eastern Ghats, where the regional trend of the dominant foliations changes from SW - NE to WNW - ESE, is fragmented into several fault-bounded crustal domains characterized by distinct lithological, structural and geochronological features (Mahalik 1994; Nash et al. 1996; Crowe 2002; Crowe et al. 2001; Fachmann 2001).
Mahalik (1994) distinguished two structural domains dominated by medium to high-grade quartzofeldspathic gneisses, the Rengali and Angul assemblages, and another three domains consisting of biotite schists locally grading into amphibolites or migmatitic hornblende gneisses, massive quartzite units, and pelitic schists. He assigned the high-grade gneiss assemblages to the 'Eastern Ghats Belt' and the low to medium-grade volcanosedimentary sequences to the Singhbhum craton, although one of the low-grade domains separates the gneiss domains. By contrast, Nash et al. (1996) and Fachmann (2001) regarded the meta-volcanosedimentary pile as cover sequence over locally migmatitic and charnockitic gneisses forming cores to regional antiforms. Based on lithological similarities between the meta-volcanosedimentary sequence and the Iron Ore Group they assumed that the entire assemblage formed part of the Singhbhum craton. The Rengali Domain was confined to the area between the Barakot and the Kerajang (or North Orissa Boundary) fault zones (Fig. 3), while southward adjoining gneisses were allocated to the Angul Domain of the Eastern Ghats (Province). Fachmann (2001) also reported NE - SW trending mafic dykes from the Rengali Domain and neighbouring provinces (Singhbhum craton, Angul Domain, Chattisgarh basin). Recently, Crowe et al. (2001) re-interpreted the geological framework of the area highlighting that an amphibolite facies rock assemblage equivalent to the Rengali Domain occurs to the south of the Kerajang Fault zone in a SW-trending, fault-bounded, triangular domain that terminates at the Eastern Ghats Boundary Fault. Further, on the basis of isotope data, a coherent block of granulite facies quartzofeldspathic gneisses and mafic granulites, the Badarama Complex, was interpreted to be included within the Rengali Province (Fig. 3).
Identical 207Pb/206Pb zircon SHRIMP ages of 2802 ± 3 Ma and 2801 ± 10 Ma for hornblende-orthogneisses from the Badarama Complex and the eastern termination of the Rengali Province, respectively, are interpreted to date crystallization of the precursor granitoids and thus point to extensive magmatism at 2.8 Ga within the Rengali Province (Crowe 2002). 207Pb/206Pb zircon ion microprobe ages of 2811 ± 3 Ma (leucocratic granitic gneiss) and 2803±4 Ma (hornblende orthogneiss) for xenoliths embedded in biotite granite corroborate this event (Mishra et al. 2000). The interpretation of Mishra et al. (2000) that the late Archaean ages provide a minimum age for granulite facies metamorphism in the Eastern Ghats Province cannot be accepted as none of the analyzed xenoliths contain mineral assemblages indicative of granulite facies. In addition, the samples were taken near the contact with the Singhbhum craton and far away from the contact with the Eastern Ghats Province. A 207Pb/206Pb zircon ion microprobe age of c. 3.5 Ga from a biotite-bearing quartzofeldspathic gneiss xenolith (Mishra et al. 2000) strongly argues for derivation of the xenoliths from the Singhbhum craton where similar zircon ages are reported from paragneisses (Mishra et al. 1999). Rb-Sr whole rock ages of 2924 ± 25 Ma for the host biotite granite (Mishra et al. 2000), and 2745 ± 103 Ma and 2735 ± 44 Ma for two charnockite samples (Sarkar et al. 1998), have to be critically evaluated in the light of likely disturbances of the Rb-Sr isotope system during later high-grade metamorphism. However,these rocks still seem to reflect the 2.8 Ga magmatic event. Likewise, Nd isotopic signatures of late mafic dykes near Rengali (TNd(DM) = 2.1 - 2.7 Ga, Nd(T) = - 9.9 ± 1.1) have been interpreted in terms of varying contamination of the primary basaltic melts with Archaean crustal material (Fachmann 2001).
The age and P-T evolution of metamorphism in the Rengali Province are poorly constrained. The assemblage garnet-staurolite-kyanite in pelitic schists (Crowe 2002) requires conditions of 550-700°C and >5 kbar (see Spear & Cheney 1989), whereas the presence of arrested-type or massive charnockite (Nash et al. 1996; Crowe et al. 2001) and garnet+hornblende-bearing migmatites (Crowe 2002) indicates considerably higher peak temperatures. However, upper amphibolite to granulite facies assemblages are largely restricted to the SW-trending arm of the Rengali Province (Crowe et al. 2001). A minimum age for the amphibolite facies metamorphism is inferred from an 40Ar/39Ar plateau age for hornblende, indicating cooling below c. 500°C at 699 ± 8 Ma (Crowe et al. 2001), and a Sm-Nd mineral isochron age of 792 ± 85 Ma for the emplacement of a mafic dyke which experienced only a low-grade overprint (Fachmann 2001).
The pronounced WNW - ESE lithostructural trend results from syn-metamorphic, multiple tight to isoclinal folding of the meta-volcanosedimentary sequences with gently to the E or W plunging fold axes (Nash et al. 1996; Crowe et al. 2001; Fachmann 2001). Pervasive secondary foliations in gneisses and meta-volcanosedimentary rocks are parallel to the subvertical axial planes of the folds. At a later increment of the bulk deformation, strain was partitioned in and at ductile shear zones. This deformation was accompanied by retrogression to amphibolite and greenschist facies conditions. Steeply plunging mineral lineations on the generally steeply to the N or S dipping mylonitic foliation within major shear zones indicate N-W directed compression during amphibolite facies metamorphism (Fachmann 2001). As no shear sense indicators are reported, it remains undecided if compression was associated with thrusting (Fachmann 2001) or right-lateral transpression (Nash et al. 1996). 40Ar/39Ar plateau ages for white mica and biotite from pervasively retrogressed schists and amphibolites indicate reactivation of the Kerajang fault zone between 490 and 470 Ma. Truncation of Late Carboniferous to Early Triassic deposits of the Talchir basin (Fig. 3) proves brittle reactivation of the Kerajang fault (Mahalik 1994; Nash et al. 1996; Fachmann 2001).
Pooled apparent apatite fission track ages between 270 ± 13 Ma and 252 ± 9 Ma suggest that the Rengali Domain entered the partial annealing zone for apatite (120°C - 60°C) at the latest in the late Permian (Lisker & Fachmann 2001). Samples from major fault zones varying between 208 ± 8 Ma and 181 ± 13 Ma further indicate a localized thermal overprint and possible reactivation of the fault zones in the Jurassic (Lisker & Fachmann 2001).
Krishna Province
The Krishna Province comprises the granulites of the newly defined Ongole Domain (equivalent to the southern part of the 'Western Charnockite Zone' of Ramakrishna et al., 1998) and the low to medium-grade Nellore-Khammam schist belt (Fig. 4). Merging of the two crustal terrains, so far considered as unrelated geological units, appears justified by congruent geochronological data for magmatic and metamorphic events (Tab. 2), which point to a common Palaeoproterozoic evolution.
Nellore-Khammam schist belt
Along strike the Nellore-Khammam schist belt has been divided into the Nellore schist belt in the south and the Khammam schist belt in the north, both being separated by a faulted block of cratonic gneisses (Raman & Murty 1997; Okudaira et al. 2000; Fig. 4). The Nellore schist belt has been split up in a western upper unit and an eastern lower unit, which are suspected to be separated by a fault or thrust (Vasudevan & Rao 1983; Narayana Rao 1983). The Khammam schist belt has been correlated with the lower unit (Okudaira et al. 2000). Following the subdivision of crustal provinces in fault or shear zone-bounded domains by Nash et al. (1996), we propose to provisionally rename the upper unit the Udayagiri Domain and the lower unit the Vinjamuru Domain (Fig. 4) until the nature of the contact is clarified.
The low-grade Udayagiri Domain consists of a greenschist facies volcanosedimentary sequence of pelites, psammites and conglomerates with local intercalations of cherts, limestones and felsic volcanics. Basic meta-igneous rocks are extremely rare (Raman & Murthy 1997). By contrast, basic volcanic rocks are ubiquitous in the amphibolite facies Vinjamuru Domain which also comprises metasedimentary schists, gneisses and migmatites, and locally abundant felsic metavolcanics. Marbles and calc silicate gneisses occur together with intermediate and felsic metavolcanic rocks, banded barite-magnetite layers, and kyanite-bearing schists to the west of Vinjamuru (Raman & Murty 1997; Vasudevan et al. 2002). The southern Vinjamuru Domain hosts several economically important muscovite pegmatites (Babu 1998). Anorthositic rocks concentrate near Inukurti in the extreme south (Kanungo & Chetty 1978; Kanungo et al. 1986) and at Chimalpahad near the southwestern margin of the Godavari Rift (Fig. 4). Anorthosites at Inukurti are embedded in voluminous amphibolites which suggests that they form cumulate layers and domes within a layered intrusion. The Chimalpahad layered mafic complex synkinematically intruded a sequence of partly migmatitic metasedimentary gneisses and amphibolites (Raman & Murty 1997, Leelanandam & Narashima Reddy 1998). Coarse cumulate gabbros with rare interlayered ultramafic rocks, metamorphosed pillow basalts, and mafic (sheeted?) dykes constitute the Kandra Igneous Complex near the southwestern margin of the Vinjamuru Domain. Nagaraj Rao et al. (1991) interpreted this assemblage as remnant ophiolite sequence. An ophiolitic melange of chaotically intermingled blocks of ultramafic rocks, metamorphosed pillow basalts, and sedimentary rocks including finely laminated cherts, embedded in a serpentinized matrix is reported from an exposure near the contact with the Ongole Domain at Kanigiri (Leelanandam 1990). The ophiolitic nature of both occurrences is not independently confirmed.
On the basis of chemical discrimination diagrams, Mukhopadyay et al. (1994) proposed a back-arc setting for the amphibolites of the Vinjamuru Domain, whereas Hari Prasad et al. (2000) argued that the protoliths have formed in two different tectonic settings comparable to recent ocean island arcs and continental margin arcs.
The timing of felsic magmatism in the Vinjamuru domain has been constrained by 207Pb/206Pb single grain evaporation ages for magmatic zircons from metarhyolites (1868 ± 6 Ma and 1771 ± 8 Ma; Vasudevan et al. 2002). 207Pb/206Pb evaporation ages for some of the numerous zircon xenocrysts show that the felsic rocks, which are seen as products of intracrustal melting, incorporated crustal material as old as 2431 Ma. A Rb-Sr whole rock age of 1534 ± 14 Ma is interpreted to reflect incomplete homogenization of the Sr isotopic system probably during the metamorphic overprint (Vasudevan et al. 2002). Several K-Ar, Rb-Sr and Pb-U-Th isotope data for muscovite and samarskite from muscovite pegmatites cluster around 1.6 Ga, but these data were generated in the late 1940's to late 1960's (for a review see Babu 1998) so that their reliability is uncertain. A K-Ar whole rock age of 989 ± 23 Ma for a metabasalt and a K-Ar muscovite age of 806 Ma for a metapelitic schist indicate final cooling of the southern Vinjamuru Domain in Neoproterozoic times (Gosh et al. 1994).
Metamorphism in the Udayagiri Domain did not exceed greenschist facies conditions (Babu 1998). Geothermobarometric calculations for amphibolites and kyanite-bearing metapelites of the Vinjamuru Domain yielded maximum P-T conditions of 700 - 750°C and 7 - 10 kbar, with apparent pressures increasing northward (Babu 1998; Okudaira et al. 2000). Further details of the P-T evolution are not known.
Several structural studies in both domains have established that a pervasive schistosity, S1, formed during peak metamorphic conditions and was subsequently multiplely folded. S1 trends parallel to the arcuate margins of the province and dips steeply to the east. Associated tight to isoclinal and westward-facing folds, F1, with subhorizontal fold axes were repeatedly folded into upright open folds with either near-coaxial or near-perpendicular fold axes (Babu 1998; Raman & Murty 1997; Okudaira et al. 2000). Muscovite-bearing pegmatites were emplaced along the foliations and in fold hinges (Babu 1998) but it remains unclear if their emplacement post-dates deformation. The development of NW - SE trending left-lateral shear zones at greenschist facies conditions in rigid amphibolites near the Godavari Rift (Hari Prasad et al. 2000; Okudaira et al. 2000) may be associated with rift formation.
Ongole Domain
The granulite facies rocks of the Ongole Domain extend from Ongole in the south to the southwestern margin of the Godavari Rift and re-appear on the other side of the rift (Fig. 4). This is exemplified by the occurrence of characteristic metasedimentary lithologies (interlayered diatexitic garnet-spinel-corundum-sillimanite gneisses, garnet-sillimanite quartzite, calc silicate rocks) up to Upper Sileru.
The domain essentially consists of garnetiferous basic granulites and large volumes of multi-intrusive enderbitic granulites interlayered with locally porphyritic leptynites. Metasedimentary rocks mostly occur in its eastern part as rafts and partly resorbed layers with black wall contacts in the meta-igneous rocks. Diatexitic pelitic garnet-sillimanite-spinel granulites dominate and are interbanded with quartzitic to psammitic gneisses and rare calc silicate granulites suggesting a typical association of shallow marine sediments as protoliths. A layered mafic-ultramafic complex at Kondapalle (Fig. 4) has intruded the high-grade metasedimentary package at deep crustal levels and was dismembered and invaded by enderbite (Sengupta et al. 1999). The layered complex consists of gabbronorite grading into finely interbanded leucogabbronorite and anorthosite with clinopyroxenite and chromite-bearing enstatite cumulate layers (Leelanandam 1997). The entire granulite assemblage is transected by allanite and monazite-bearing pegmatites and microgranites.
The deposition age of the sediments, the oldest component in the assemblage, is not known. Nd model ages (TDM) reflecting the mean crustal residence age of the provenance areas range between 2.8 - 2.6 Ga and compare well with Nd model ages for granitoids of the eastern Dharwar craton (Rickers et al. 2001). From this, these authors concluded that the clastic material was presumably derived from the Archaean granitoids without incorporation of juvenile material. The intrusion of the enderbite-charnockite precursors has been dated by near-concordant U-Pb ages for magmatic zircons at c. 1.72 - 1.70 Ga (Kovach et al. 2001). These data together with Nd model ages (TDM) of 2.5 - 2.3 Ga (Rickers et al. 2001) suggest a derivation from mantle-derived magma through assimilation-fractional crystallization processes involving substantial input of Archaean crustal material. The short-lived magmatism was closely followed by granulite facies metamorphism. Concordant U-Pb monazite ages of 1672 ± 3 Ma, which date the syn-kinematic formation of coarse-grained leucosome networks in the enderbitic granulites (Kovach et al. 2001) and the emplacement of a syn-kinematic pegmatite (Mezger & Cosca 1999), approximate the age of granulite facies metamorphism. A reproducible 207Pb/206Pb allanite age from a late-kinematic pegmatitic vein (c. 1598 Ma; Mezger & Cosca 1999) and the results of EPMA monazite dating (age populations of 1.64 - 1.55 Ga and 1.45 - 1.39 Ga; Simmat 2002), argue for a prolonged or polyphase metamorphism. The Pb isotope compositions of feldspars from metasedimentary rocks cluster tightly at the 1.5 Ga isochron (Rickers et al. 2001), consistent with last isotopic homogenization of Pb during this prolonged metamorphism. A Rb-Sr whole rock age of 1507 ± 59 Ma for a mangeritic granulite may not date emplacement (Sarkar & Paul 1998) but rather reflect incomplete homogenization of the Sr isotopic system during metamorphism. A 40Ar/39Ar total degassing age for hornblende documents final cooling below c. 500°C at 1111 Ma (Mezger & Cosca 1999).
The P-T evolution of the dominant granulite facies metamorphism in the Ongole Domain has not been studied in detail (Sengupta et al. 1999; Dasgupta & Sengupta, this volume). Metapelitic country rocks from the immediate contact with the Kondapalle layered complex exhibit mineral assemblages indicative of temperatures in excess of 1000°C and 8 kbar that are attributed to deep-crustal ultra high-temperature contact metamorphism preceeding the regional granulite facies event (Sengupta et al. 1999). The UHT event is not documented by the EPMA monazite age data of these rocks (Simmat & Raith 1998, Simmat 2002). Detailed studies of the regional deformation, which accompanied granulite facies metamorphism, are lacking. In general, deformed mafic dykes, potassium feldspar-bearing leucosomes in diatexitic granulites, and quartz veinlets evidence multiple isoclinal folding of the rocks. Compositional layering and axial plane foliations trend mostly parallel to the margins of the domain and dip moderately to steeply to the E/SE or NW. Orthopyroxene grains define a SE/NW-ward plunging mineral lineation. Locally present cm-wide subvertical shear zones are filled with orthopyroxene-bearing leucosomes and transected at acute angles by narrow mylonite zones. Shear sense indicators uniformly suggest transport of the hanging wall towards the NW. Axes of late open folds plunge moderately to steeply to the NE. Moderately to the NE dipping brittle shear zones have been observed only near the Godavari Rift and thus may be associated with rift formation.
The emplacement of variegated alkaline to calc-alkaline plutons and layered mafic-ultramafic complexes ('Prakasam Alkaline Province', Prasada Rao et al. 1988, Leelanandam 1989) into granulites of the Ongole Domain, and into migmatitic gneisses and amphibolites of the Vinjamuru Domain blurred the boundary between both domains (Fig. 4). The mafic-ultramafic complexes consist of olivine-bearing clinopyroxenites and gabbroic to noritic and anorthositic rocks (Dasgupta et al. 1997; Rao et al. 1998b). In the alkaline plutons, nepheline syenite clearly dominates over hornblende syenite, syenite and, quartz syenite. Additionally, ferrosyenite, shonkinite and alkali granite occur as subordinate components. Whether small volumes of gabbro are genetically related is debated (Leelanandam 1998; Nagsai Sharma & Ratnakar 2000).
Sarkar & Paul (1998) listed Rb-Sr whole rock ages for nepheline syenites of the Purimetla (1369 ± 28 Ma), Uppalapadu (1348 ± 41 Ma), and Elchuru (1242 ± 33 Ma) alkaline plutons and suggested that these ages date emplacement. Overall, however, the chronology of magmatism and the effects of regional deformation and metamorphism on the plutons are poorly constrained. Dasgupta et al. (1997) reported that the Chimakurthy layered complex (Fig. 4) and subsequently intruded alkaline rocks were affected by progressive deformation. The metamorphic regime during this inferred regional deformation event has not been specified. By contrast, Nagasai Sharma & Ratnakar (2000) stated that the rocks of the Chanduluru composite pluton are devoid of deformation features except for weakly strained margins. The chemical characteristics of the calc-alkaline pluton point to emplacement in an Andean-type arc. Clearly, further studies are required.
At the contact with the Chimakurthy mafic-ultramafic complex, cordierite-bearing metapelitic country rocks show imprints of a mid-crustal ultra-high temperature contact metamorphism along a near-isobaric heating-cooling trajectory within the temperature interval between c. 750°C and 1000°C (Dasgupta et al. 1997). Two-pyroxene thermometry established maximum temperatures of c. 1100°C in the centre of the complex (Rao et al. 1998b).
Eastern Ghats Province
The Eastern Ghats Province consists of an intensely deformed and metamorphosed assemblage of metasedimentary, basic and enderbitic granulites interspersed with massif-type anorthosites and probably multi-intrusive granite-charnockite complexes. The dominant metasedimentary lithologies are khondalites and diatexitic quartzofeldspathic gneisses, both frequently interlayered with sillimanite and garnet-bearing quartzites and leucocratic garnetiferous gneisses, commonly termed leptynite. Conformable bands and metre-sized xenoliths of high-MgAl granulites, containing sapphirine and spinel-bearing assemblages, have been found in several localities (summarized in Sengupta et al. 1999; Dasgupta & Sengupta, this volume). Calcareous rocks occur as rare layers or lenses of calc silicate granulites except for the northwestern margin where larger occurrences exist (Bhattacharya et al. 1998; Ramakrishnan et al. 1998) (Fig. 3). Marbles are known only from a small area to the north of Visakhapatnam (Bhowmik et al. 1995). Geochemical analyses confirm the sedimentary origin of the khondalites (Nanda & Pati 1989; Sen & Bhattacharya 1997; Rao et al. 1998a), but the notion that they represent metamorphosed equivalents of deeply weathered soil profiles (Dash et al. 1987) met with strong opposition (Nanda & Pati 1991). Field and geochemical evidence implies a magmatic origin for most of the leptynites (Sen & Bhattacharya 1997; Shaw 1996; Yamamoto et al. 1998), i.e. their formation through segregation and intrusion of felsic melts generated by biotite-melting in the high-grade crustal lithologies (Braun et al. 1996). Especially in the Chilka Lake area, leptynites constitute sizeable intrusive bodies of garnetiferous leucogranite (Dobmeier & Raith 2000; Dobmeier & Simmat 2002). The meta-igneous rocks range from basic to felsic in composition, but fine grained enderbites (Murthy et al. 1998), locally migmatitic charnockites (Charan et al. 1998), and above all garnet and/or orthopyroxene-bearing porphyric granitoids (Mukhopadhyay & Bhattacharya 1997; Krause 1998; Kovach et al. 1998), which constitute voluminous batholitic bodies apparently associated with megalineaments (Nash et al. 1996; Krause 1998; Ramakrishnan et al. 1998), predominate. The porphyric granitoids were not affected by the early deformation. Their bulk chemical compositions and isotope data imply a derivation from crustal sources (Krause 1998). Basic rocks form centimetre to metre-sized layers and lenses. The high-grade gneiss association in the northern part of the province hosts four massif-type anorthosite complexes of varying size (Bolangir: Bhattacharya et al. 1998, Dobmeier 2002; Chilka Lake: Sarkar et al. 1981, Dobmeier & Simmat 2002; Turkel: Maji et al. 1997; Jugsaipatna: Nanda & Panda 1999) (Fig. 3). The larger anorthosite massifs are bordered by crustal-derived felsic igneous rocks varying from monzodiorite to granite. Strongly Fe-enriched dioritic rocks showing exceptionally high concentrations of REE and high field strength elements occur at the immediate contact with the anorthosite which they intrude in cross-cutting dykes and sheets (Bhattacharya et al. 1998; Raith et al. 1997; Maji et al., 1997). These rocks are interpreted as residual melts of the anorthosite (Bhattacharya et al. 1998) and, due to the high modal abundance of zircon have enabled dating of the age of anorthosite emplacement at Chilka Lake and Bolangir (Krause et al. 2001).
The regional distribution of khondalitic lithologies led to the subdivision of the Eastern Ghats Province in a Western and an Eastern Khondalite Zone, and an intermediate Charnockite-Migmatite Zone (Nanda & Pati 1989; Ramakrishnan et al. 1998). However, the results of Nd isotope mapping (Rickers et al. 2001) oppose such a subdivision and rather suggest that the province is divided along the Nagavalli-Vamsadhara and Mahanadi megalineaments (Chetty 1995; Fig. 2a) into a northeastern domain (Angul and Tikarpara Domains; Fig. 2b) with Nd model ages (TDM) of 2.2 - 2.8 Ga, a central domain with Nd model ages restricted to 1.8 - 2.2 Ga, and a southwestern domain with a more heterogeneous distribution of Nd model ages (paragneisses 2.1-2.5 Ga, orthogneisses 1.8-3.2 Ga). As the majority of analyzed granitoids exhibit pronounced S-type characteristics (Krause 1998), the data reflect regional-scale isotopic and hence compositional heterogeneity of the crustal source. Evidently, for the central domain the presence of Archaean components can be ruled out.
U-Pb ages of detrital zircons (Shaw et al. 1997; Jarick 1999) suggest that the deposition of sediments did not cease before c. 1.35 Ga. An early episode of felsic magmatism at c. 1.45 Ga in the Eastern Ghats Province seems to be indicated by Sm-Nd whole rock isochron ages for orthogneisses from Rajagada (Shaw et al. 1997), which has to be verified by U-Pb zircon dating. Near-coeval with this crustal magmatism, an alkaline complex was emplaced near Khariar at c. 1.5 Ga (Sarkar & Paul 1998; Aftalion et al. 2001). However, it remains unclear if the alkaline complex (Fig. 3) intruded the Bhandara craton or the Eastern Ghats Province (Madhavan & Khurram 1989; Aftalion et al. 2001).
Discordant U-Pb zircon data (Aftalion et al. 1988), a Pb/Pb isochron age (Paul et al. 1990), and a concordant 207Pb-206Pb SHRIMP age (Crowe 2002) for orthopyroxene-bearing meta-igneous rocks from the Angul and Phulbani Domains point to a second pulse of felsic plutonism at 1.2 - 1.15 Ga. Further field and isotopic work is needed to substantiate this magmatic event and its regional extent.
An early phase of ultra-high temperature metamorphism (c. 950°C and 8 kbar) preserved in sapphirine-bearing high-MgAl granulites (Lal et al. 1987; Kamineni & Rao 1988; Dasgupta 1995; Dasgupta & Sengupta, this volume) has been dated with the common Pb method at 1099 ± 56 Ma (Jarick 1999) and thus may be linked to this event. A further time constraint is provided by a near-concordant 207Pb-206Pb SHRIMP age of 1105 ± 9 Ma for zircons with euhedral zoning, interpreted to date partial melting of leucogranite (leptynite) in the Phulbani area (Crowe 2002).
The emplacement of the voluminous complexes of porphyric granitoids between 985 and 955 Ma has been firmly established by near-concordant U-Pb ages (Grew & Manton 1986; Paul et al. 1990; Kovach et al. 1998) and concordant 207Pb-206Pb SHRIMP ages (Crowe 2002) for magmatic zircons. An accompanying pervasive granulite facies metamorphism attained peak conditions of 750-800°C at 7-8 kbar, and was followed by a steeply decompressive retrograde P-T segment to final conditions of 650-700°C and 4-5 kbar (Mukhopadhyay & Bhattacharya 1997; also reviewed in Sengupta et al. 1999; Dasgupta & Sengupta this volume). U-Pb monazite ages of 973 - 954 Ma from metasedimentary and meta-igneous rocks (Aftalion et al. 1988; Paul et al. 1990; Mezger & Cosca 1999), Pb/Pb single zircon evaporation ages of 952 - 946 Ma for metamorphic zircons from high-MgAl granulites (Jarick 1999), and comprehensive EPMA monazite data clustering between 1.05 and 0.95 Ga (Simmat 2002) are interpreted to date this metamorphism, which was accompanied by pervasive regional deformation.
From regionally limited structural studies (Sarkar et al. 1981; Halden et al. 1982; Bhattacharya et al. 1994; Shaw 1996) it has been concluded that the entire Eastern Ghats Province experienced three discrete major episodes of regional deformation (summarized in Bhattacharya 1997; Sarkar & Paul 1998), the first of which may be coeval with ultra-high grade metamorphism (Jarick 1999; Crowe 2002). Yet, most of these structural analyses have only been based on the discrimination of fold axis populations using the plunge of the axes as discriminant. Further, it was taken for granted that the entire granulite terrain has been deformed homogeneously (cf. Bhattacharya 1997). However, the results of combined structural and geochronological studies at the Bolangir and Chilka Lake anorthosite complexes (Krause et al. 2001; Dobmeier 2002; Dobmeier & Simmat 2002) and in the Angul Domain (Halden et al. 1982) have shown that most structures are the imprints of progressive regional deformation that occurred at appreciably different time in different sectors of the Eastern Ghats Province:
In the Khariar Domain (Figs 2b & 3), the time of emplacement of the massif-type anorthosite at Bolangir (Bhattacharya et al. 1998) in a thrust shear-dominated deformational regime with N - S directed shortening (Dobmeier 2002) is constrained by an upper intercept age of 933 ± 32 Ma for slightly discordant U-Pb zircon data from ferrodiorite (Krause et al. 2001) and a near-concordant 207Pb/206Pb zircon SHRIMP age of 918 ± 20 Ma for garnetiferous leucogranite (Crowe 2002). Pervasive secondary foliations in the anorthosite complex and the country rocks trend W - E and dip moderately to steeply to the south, with lineations and fold axes plunging in southern directions. U-Pb titanite data for calc silicate rocks show that cooling below c. 650°C occurred at 570 to 530 Ma (Mezger & Cosca 1999) and thus point to a significant Pan-African thermal overprint of this domain.
In the Angul Domain (Figs 2b & 3) an Early Neoproterozoic age of granulite facies metamorphism and deformation, during which the N-S trending isoclinally folded steep compositional layering and associated schistosities were formed (Halden et al. 1982), is indicated by U-Pb and EPMA monazite data which cluster around 960 Ma (Aftalion et al. 1988; Mezger & Cosca 1999; Simmat 2002). U-Pb titanite data (Mezger & Cosca 1999) show that the domain was cooled below c. 650°C at c. 930 Ma. Halden et al. (1982) interpreted a Rb-Sr muscovite age of 854 ± 6 Ma to date the emplacement of pegmatites in axial planes of late folds. However, a coeval 40Ar/39Ar age (854 ± 4 Ma) for amphibole from the same area (Lisker & Fachmann 2001) rather suggests to interpret the Rb-Sr age as cooling age, unrelated to deformation. The geochronological data imply that the Angul Domain was tectonically and thermally decoupled from the larger part of the Eastern Ghats Province prior to the regional Pan-African tectonothermal event.
The sequence and geometry of structures in the Chilka Lake Domain strongly resemble those described in the Angul Domain, except for the orientation of the planar and linear elements that both trend predominantly SSE - NNE (Bhattacharya et al. 1994; Dobmeier & Raith 2000). However, the U-Pb zircon and EPMA monazite dates have established that there the regional deformation occurred in this domain much later. The emplacement of the Chilka Lake anorthosite complex (792 ± 2 Ma; Krause et al. 2001) and leucogranites (762 - 743 Ma) probably mark the initial phase of a prolonged Neoproterozoic orogenic event that culminated with a pervasive transpressive deformation and granulite facies metamorphism at 690 - 660 Ma (Dobmeier & Simmat 2002).
Similar ages have also been reported from the northern margin of the Khariar Domain near Bolangir (U-Pb zircon SHRIMP age of 764 ± 35 Ma, Crowe 2002; EPMA monazite data of 820-780 Ma, authors' unpublished data) and its eastern extension, the Kantilo-Ranipathar shear zone (207Pb/206Pb zircon SHRIMP age of 789 ± 7 Ma; Crowe 2002), and the Rampur Domain (Rb-Sr mineral isochron data between 833 ± 10 Ma and 781 ± 38 Ma, Sm-Nd mineral isochron data between 815 ± 9 Ma and 808 ± 64 Ma; Shaw et al. 1997). The latter data are thought to relate to thermal metamorphism accompanying the emplacement of a porphyritic granite. An 40Ar/39Ar amphibole age of 650 ± 4 Ma for the Koraput alkaline complex (Rb-Sr whole rock age 864 ± 25 Ma; cited in Sarkar & Paul 1998) indicates final cooling below c. 500°C near the contact with the Jeypore Province (Mezger & Cosca 1999).
Further 40Ar/39Ar stepwise heating (Mezger & Cosca 1999; Aftalion et al. 2001; Crowe et al. 2001; Lisker & Fachmann 2001) and laser spot analysis data (Lisker & Fachmann 2001), Sm-Nd and Rb-Sr mineral isochrons (Shaw et al. 1997), U-Pb mineral ages (Kovach et al. 1998; Mezger & Cosca 1999; Aftalion et al. 2001), and EPMA monazite data (Dobmeier & Simmat 2002; Simmat 2002) provide evidence for a final pulse of medium to high-grade thermal activity associated with deformation localized in shear zones at c. 515 Ma. Moreover, these data highlight a rather complex cooling history for the northern Eastern Ghats Province between 622 Ma and 456 Ma, which necessitates differential movements at major shear zones in late Neoproterozoic to Cambrian time.
Pooled apparent apatite fission track ages, scattering essentially between 300 and 250 Ma, document that the Angul and Tikarpara domains entered the partial annealing zone for apatite at the latest in late Permian times (Lisker & Fachmann 2001). Data for samples from major fault zones varying between 152 ± 12 Ma and 119 ± 6 Ma indicate a localized thermal overprint and reactivation of the fault zones during the Cretaceous, probably coeval with the onset of sea floor production in the Bay of Bengal at c. 117 Ma (Lisker & Fachmann 2001).
Reconnaissance studies have shown that all crustal provinces are bounded by sheared margins. Most prominent of all is the western margin of the Eastern Ghats Province, which coincides with a salient satellite megalineament that runs from the Godavari rift via Sileru and Koraput to close to Bhawanipatna along the contact with the Jeypore Province (Fig. 2). Further to the north, where the Eastern Ghats Province abuts on the Bhandara craton, the margin is less distinct in satellite imagery but better studied on the ground. The megalineament is either termed Sileru shear zone (Chetty 1995, 2001; Chetty & Murthy 1998b) or Elchuru - Kunavaram - Koraput shear zone (Chetty & Murthy 1998a), while its northern extension is known as Eastern Ghats Boundary Fault (Ramakrishnan et al. 1998) or Lakhna thrust (Biswal 2000). The width of the moderately to the E/SE inclined Eastern Ghats Boundary Shear Zone averages 3 kilometres, and is composed of granulites and cratonic rocks. All rocks are complexly folded but contain a late shear foliation that trends strictly parallel to the shear zone boundary and a mylonitic foliation in frequently encountered mylonitic rocks (Gupta et al. 2000). Feldspar, quartz, pyroxene, amphibole, epidote or biotite mark an associated down-dip lineation, and shear sense indicators uniformly imply transport of the hanging wall towards the W/NW. Chetty (1995) additionally reported subsequent dextral transcurrent displacement. The folded foliations contain mineral assemblages indicative of widely different metamorphic conditions, whereas the late foliation formed at similar P-T conditions everywhere. Gupta et al. (2000) documented the preservation of a narrow inverted thermal gradient within the adjacent cratonic domain, with temperature increasing from c. 350°C in the west to c. 700°C near the contact with the granulites of the Eastern Ghats Province. The thermal profile indicates aggregation of the terrains through westward thrusting of the hot granulite assemblage over a cooler cratonic foreland. The granulites show only minor retrogression. Significantly, the shear zone is the locus of emplacement for the Khariar and Koraput alkaline plutons (Leelanandam 1998). Considerably different emplacement ages for the alkaline complexes of c. 1500 Ma (Aftalion et al. 2001) and c. 864 Ma (Sarkar & Paul 1998) imply either repeated alkaline magmatism or disturbance of the Rb-Sr systematics during later tectonometamorphic events. In fact, EPMA monazite data suggest that the late foliation formed only in the early Phanerozoic, i.e. during the Pan-African event at c. 530 Ma (authors' unpublished data). Further to the north, the shear zone swings sharply to the east. Essentially, the structural evolution remains the same near the moderately S-dipping shear zone, but shear sense indicators suggest N-directed transport of the hanging wall (Dobmeier 2002). U-Pb zircon SHRIMP data (Crowe 2002) and EPMA monazite data (authors' unpublished data) from the northern Bolangir anorthosite complex indicate repeated activity along this segment of the shear zone in Neoproterozic and early Phanerozoic times.
The propagation of the Eastern Ghats Boundary Shear Zone east of the Bolangir anorthosite is unclear (Nash et al. 1996; Ramakrishnan et al. 1998; Crowe et al. 2001). The contrasting evolution of the Angul Domain suggests that the shear zone continues into the Kantilo shear zone and the western segment of the Ranipathar shear zone (Mahanadi shear zone of Chetty 1995; Fig. 3). The western segment of the Kantilo shear zone is blurred by the emplacement of the Sonapur granite at 982 ± 6 Ma (Crowe 2002). Farther to the east mylonites formed at 789 ± 7 Ma in a porphyritic granite gneiss which was emplaced at 980 ± 7 Ma (Crowe 2002). Yet, the absence of a regional Pan-African imprint in the Tikarpara Domain has to be substantiated by more radiometric data.
The contact zone of the Jeypore Province with the Bhandara craton shows marked similarities with the Eastern Ghats Boundary Shear Zone. It is characterized by pervasive retrogression of the granulites to amphibolites and hydrated gneisses, and the formation of a narrow spaced planar fabric in all rocks (Nanda & Pati 1989). Mylonites, however, are largely confined to the cratonic rocks, which exhibit a steadily increasing metamorphic overprint reaching lower amphibolite facies conditions towards the contact zone. Planar fabrics in the high strain zone trend strictly parallel to the contact and dip moderately to steeply towards E/ESE. An associated down-dip mineral stretching lineation (amphibole, epidote, feldspar, quartz) and unambiguous shear sense indicators in the mylonites indicate west-directed transport of the hanging wall during retrogression (authors' unpublished field observation). The age of this deformation is not constrained. It has to be noted that this zone of retrogression/prograde metamorphism is much narrower than shown by Ramakrishnan et al. (1998).
To the north of the Godavari Rift, problems in locating the margin of the Krishna Province result from poor exposure and an apparent transitional character. The sheared Kunavaram alkaline complex may have been emplaced at or near this margin at 1265 ± 58 Ma (Rb-Sr whole rock age; Sarkar & Paul 1998) and thus broadly synchronous with the alkaline complexes of the Prakasam Alkaline Province. Alternatively, the Rb-Sr age may be afflicted by disturbance of the Sr isotopic system during later deformation and medium-grade metamorphism. To the south of the rift, the Krishna Province essentially abuts against the Cuddapah basin. The contact is a several kilometres wide tectonic melange, produced by thrust imbrication and intense folding, and is intruded by several felsic plutons (Meijerink et al. 1984). Further, the available data indicate that westward thrusting of the Krishna Province caused the strong deformation of the easternmost Cuddapah basin with creation of the Nallamalai fold belt. However, age and evolution of the deformation remain unconstrained.
Repeated intense reactivation of the margins of the Rengali Province, the Barakot - Akul and Kerajang faults (Mahalik 1994; Nash et al. 1996) at semiductile and brittle conditions in Phanerozoic times (Crowe et al. 2001; Fachmann 2001; Lisker & Fachmann 2001) makes it extremely difficult to assess their Proterozoic significance. From results of their satellite imagery studies, Nash et al. (1996) and Crowe et al. (2001) inferred considerable right-lateral displacement along the Kerajang Fault and the ductile precursor shear zone in Neoproterozoic or early Phanerozoic times, respectively. However, no combined tectonic and geochronological studies exist to confirm this assumption.
There is no evidence for a continuation of the Eastern Ghats Belt into the Southern Granulite Terrain (see the discussion of proposed continuations in Gopalakrishnan 1998). Notably, the early Mesoproterozoic period of crust formation and tectonism in the Eastern Ghats Belt coincides with an episode of tectonometamorphic inactivity in southern India. The limited geochronological data base implies that crust formation in the Southern Granulite Terrain was essentially completed in early Palaeoproterozoic time (reviewed in Braun & Kriegsman, this volume). To date, however, the internal configuration of the Southern Granulite Terrain is not yet sufficiently known to arrive at a conclusive model of its relations to the neighbouring crustal terrains.
pre-Rodinia evolution
Our review and synthesis Tab. 2) has shown that the Eastern Ghats and adjacent areas expose a deep section through a composite orogenic belt that comprises four granulite facies crustal provinces with widely different geological evolutions.
Isotope data establish that the Jeypore Province has evolved in the Archaean (Rickers et al. 2001; Kovach et al. 2001). It now constitutes a narrow high-grade igneous belt along the eastern fringe of the Bandhara craton. In the absence of adequate geochemical and isotopic data the palaeotectonic setting cannot be assessed, although the lithological composition and the spatial arrangement of the province in relation to the Bhandara Craton suggests formation in an active continental margin setting. Alternatively, the highly sheared margins allow to understand the province as an exotic terrane that has been amalgamated with Proto-India previous to or together with the Eastern Ghats Province.
At least parts of the Rengali Province also formed in the Archaean. Crowe (2002) argues that the Rengali Province earlier constituted a part of the northern Bandhara craton and was juxtaposed with the Eastern Ghats Province by right-lateral transport along a transcurrent shear zone possibly in late Mesoproterozoic times. This interpretation offers the interesting possibility to directly interrelate the Rengali and Jeypore provinces, as the latter probably experienced igneous activity and/or high-grade metamorphism at approximately the same time when voluminous granitoids intruded the Rengali Province. However, additional data are required from the poorly studied Jeypore Province before further conclusions can be drawn.
Strikingly similar isotope data for magmatism and metamorphism documented in the granulites of the Ongole Domain and lower grade rocks of the adjacent Nellore-Khammam Schist Belt Tab. 2) strongly argue for a common geological history of these units which together form the Krishna Province. The extrusion of conformable felsic lavas at 1.87-1.77 Ga (Vasudevan et al. 2002) proves that the volcanosedimentary sequence of the Vinjamuru Domain evolved in the late Palaeoproterozoic and thus evidently rules out any correlation with the Archaean greenstone belts of the eastern Dharwar craton as suggested by earlier workers (e. g. Raman & Murthy 1997; Kumar et al. 1999, and references therein). The igneous activity was closely followed by a tectonothermal event between c. 1.65 and 1.55 Ga. A second event, probably associated with alkaline magmatism (Sarkar & Paul 1998), may have occurred between c. 1.45 and 1.3 Ga (Mezger & Cosca 1999; Simmat 2002). Coeval kimberlites and lamproites of the eastern Dharwar craton and the Cuddapah basin (Chalapati Rao et al. 1996) are most likely further products of this alkaline magmatism. The extent of the mantle-derived igneous province points to extensive magmatic underplating, but the causative geodynamic processes remain unclear.
The limited knowledge of the nature of its boundaries and lack of time constraints on deformation in the eastern Cuddapah basin prohibit far-reaching interpretations of the geodynamic evolution of the Krishna Province and the adjacent craton. However, the metamorphic zonation within the Krishna Province and the Cuddapah basin is suggestive of E-W-oriented convergence, causing an inverse metamorphic stacking of granulites (Ongole Domain), amphibolite facies rocks (Vinjamuru Domain), greenschist facies rocks (Udayagiri Domain), and very low-grade rocks (Nallamalai fold belt), in the late Paleoproterozoic. The available deep seismic sounding data support this interpretation (Kaila et al. 1987; Nayak et al. 1998; Fig. 5).
Noticeably, the lithological successions of the Krishna Province resemble the Cuddapah Supergroup, with similar crystallization ages for igneous rocks. It thus appears likely that the Krishna Province constitutes metamorphic equivalents of the Cuddapah Supergroup. Nd model ages for paragneisses of the Ongole Domain suggest a derivation of the sedimentary material from the Dharwar craton (Rickers et al. 2001), which is also considered as provenance of the Cuddapah Supergroup sediments (Crawford & Compston 1973; Meijerink et al. 1984).
The west-vergent stacking presumably resulted from the collision of the Napier Complex with the eastern Dharwar Craton in late Palaeoproterozoic times. This perception is supported by recent isotope data that indicate a tectonothermal event at c. 1.6 Ga in the Napier Complex (Grew et al. 2001; Owada et al. 2002). It is further consistent with latest models for the pre-Pan-African configuration of Antarctica (Fitzsimons 2000; Boger et al. 2001).
The alkaline complexes exposed along the western margin of the Eastern Ghats Province and at domain boundaries within the Krishna Province may denote incipient continental rifting (Leelanandam 1998) in the early Mesoproterozoic. Further, rifting may have initiated the deposition of the sedimentary sequences of the Eastern Ghats Province. However, this geodynamic scenario remains highly speculative since the cause and the mandatory eastern cratonic counterpart are not known.
Rodinia assembly
The presented isotope data clearly dismiss the repeatedly expressed view that the Eastern Ghats Province underwent a first granulite facies metamorphism and deformation in the late Archaean (e. g. Chetty 1995; Chetty & Murthy 1998b; Yoshida et al. 1992; Ramakrishnan et al. 1998; Sarkar & Paul 1998). Clearly, most of the crust of the Eastern Ghats Province evolved in Proterozoic times during continent-continent collision processes related to the assembly of Rodinia (Rickers et al. 2001). Two models have been proposed for the crustal evolution: either the Eastern Ghats Province may have grown from SW to NE in several phases or it may consist of three distinct crustal blocks separated by the composite Vamsadhara-Nagavalli shear zone and the Ranipathar (or Mahanadi) shear zone (Fig. 2) (Rickers et al. 2001), both of which were the site of voluminous felsic plutonism during Early Neoproterozoic times (Fig. 3). The second model is supported by the sharp contacts between the blocks and their considerably different Nd isotope compositions. Incomplete reworking of the crust in the southwestern block, a typical feature of orogenic fronts, contrasts with the homogeneous crustal material of the central block, indicative of a position far away from the orogenic front. The central block, which corresponds to the Phulbani Domain (Fig. 2b), may represent an indenter-like protrusion of internal portions of the orogen. If so, the bounding shear zones should exhibit opposing shear sense during the same time interval, i.e. left-lateral at the Vamsadhara-Nagavalli composite shear zone but right-lateral along the Ranipathar shear zone.
Ultra high-grade regional metamorphism presumably documents amalgamation of the Eastern Ghats Province with Proto-India during the global Grenvillian orogeny at c. 1.1 Ga. Imprints of this event became largely erased during subsequent extensive granitoid emplacement and coeval pervasive deformation and high-grade metamorphism between 985 - 950 Ma. It appears that crustal consolidation of most parts of the Eastern Ghats Province was achieved during this episode. Yet, the Chilka Lake Domain became pervasively reworked at high-grade metamorphic conditions between 800 Ma (anorthosite emplacement) and 690-660 Ma (transpressional deformation). This event also led to reactivation of major shear zones at least in the northern part of the province.
Imprints of Gondwana assembly
Isotope data establish for most of the Eastern Ghats Belt that the rocks experienced a medium grade thermal overprint in early Phanerozoic times, i.e. during the global Pan-African orogeny. The combination of EPMA-monazite age data (authors' unpublished data) in combination with structural and petrological data from the Eastern Ghats Boundary shear zone near Junagarh (Gupta et al. 2000; Fig. 3) imply that final thrusting of the Eastern Ghats Province onto the Bhandara craton occurred during this event, and that considerable parts of the Eastern Ghats Province left the high temperature regime only at that time. The near-coeval activity of several other major shear zones leaves no doubt that the Pan-African event modified the internal organisation of the Eastern Ghats Belt. It has to be pointed out, that the above proposed indentation of the Phulbani Domain has not necessarily occurred during the pervasive late Grenvillian tectonothermal event. Unfortunately, no combined structural and geochronological data are available for the Vamsadhara-Nagavalli lineament.
Recognition of suture zones in or at the Eastern Ghats Belt
Apparent ophiolitic rock assemblages have been reported only from the Vinjamuru Domain (Leelanandam 1990; Nagaraj Rao et al. 1991), but the occurrences lie well away from the domain margins, and the ophiolitic nature has not been firmly established. However, the highly sheared margin of the Eastern Ghats Province separates crustal blocks of markedly different ages and distinct geochemical and isotopical signatures, and ocean floor remnants could therefore be expected. The isotopical and geochronological data suggest that the Rengali and Jeypore provinces constitute parts of the Bhandara craton, and that the supracrustal rocks of the Krishna Province have been accumulated on the eastern extension of the Dharwar craton in Palaeoproterozoic times. By contrast, the geological evolution of the Eastern Ghats Province differs decidedly from the adjacent cratonic foreland. Cooling of the Ongole Domain granulites below c. 500°C at c. 1.1 Ga (Mezger & Cosca 1999) is in strong contrast to the protracted granulite facies metamorphism in the adjoining Eastern Ghats Province through the Grenvillian and Pan-African eras. This difference necessitates important differential movements along the Sileru shear zone during and following granulite facies metamorphism (Simmat & Raith 1998; Simmat 2003). Accordingly, the boundary shear zone of the Eastern Ghats Province may constitute a suture zone, even if inherent characteristics of suture zones, such as ophiolite bodies or metasedimentary rocks of indisputable oceanic affinity, are absent.
The high-grade crustal terranes in the Rodinia and Gondwana context
The eastern margin of Proto-India was the locus of repeated accretionary activity during the Proterozoic era. A first orogenic cycle started with the deposition of volcanosedimentary sequences on the eastern Dharwar craton at c. 1.9 - 1.7 Ga. These platform sequences were segmented and transported towards the west during subsequent nappe tectonics. There is no evidence for incorporation of cratonic crust in the thrust belt. The weakly deformed and metamorphosed autochthonous western Cuddapah basin forms the western foreland of the W-vergent thrust-nappe stack of the Krishna Province. The pile consists of three major tectonic units with markedly different degree of metamorphism and intensity of deformation. The greenschist facies Udayagiri Domain at the base of the pile is overlain by the amphibolite facies Vinjamuru Domain and the granulite facies Ongole Domain. Collision took place between 1.7 and 1.55 Ga. Although plate re-organization during subsequent assembly and breakup of Rodinia and Gondwana makes it difficult to identify the eastern (present coordinates) continental counterpart and the possible extent of the orogen, latest isotope data suggest that the orogeny resulted from collision of the eastern Dharwar Craton with the Napier Complex of East Antarctica. The emplacement of basic-ultrabasic and alkaline intrusive complexes ("Prakasam Alkaline Province") indicates post-collisional relaxation of the thickened lithosphere between 1.4 - 1.2 Ga. Uplift and cooling below c. 500 °C was completed before the onset of the Grenvillian tectonothermal event.
The Eastern Ghats Province exposes the root zone of a collisional orogen predominantly consisting of intensely deformed high-grade pelitic to quartzo-feldspathic continental-derived metasedimentary sequences that were detached from their basement and intruded by voluminous syn-orogenic granitoids. Demonstrably, the basement of the supracrustals, which were probably deposited in early Mesoproterozoic times, has not been incorporated into the orogen. A derivation of the sediments from Proto-India can be ruled out and an adequate provenance has not yet been identified. The welding together of this province with Proto-India probably occurred at 1.1 Ga and was associated with ultrahigh-grade metamorphism. Pervasive deformation at high-grade metamorphic conditions and extensive crustal magmatism between 985 - 950 Ma mark the climax of the orogenic activity and result from intracontinental collisional processes. The similar evolutionary history of the Rayner Province of East Antarctica (Black et al. 1987; Harley & Hensen 1990; James et al. 1991; Boger et al. 2000; Kelly et al. 2000) suggests that both provinces are segments of the same Grenvillian orogenic belt that supposedly formed during the collision of Proto-India and the Central Antarctic craton (see discussions in Mezger & Cosca 1999; Fitzsimons 2000; Rickers et al. 2001). The plate tectonic significance of an episode of strong crustal reworking in the Chilka Lake Domain dated between 800 and 660 Ma (Dobmeier & Simmat 2002) remains unclear. This age group is unknown from adjacent terrains in the Eastern Ghats province, but provides a possible link with the Southern Granulite Terrain and Sri Lanka (see Braun & Kriegsman 2003) and the Rayner Complex (Black et al. 1987), where intrusive activity in this period has been reported. The post-Grenvillian architecture of the orogen was considerably modified during the Pan-African orogeny. The deep crustal section was internally re-organized, uplifted and thrust over its western cratonic forelands, possibly together with the Jeypore Province, in late Neoproterozoic to early Phanerozoic times. This implies a long-lasting stability of the thickened Grenvillian crust. In contrast to the situation in the Lützow-Holm-Bay and Prydz Bay complexes of East Antarctica (Fitzsimons 2000; Boger et al. 2001), the Southern Granulite Terrain of southernmost India, and Sri Lanka (see below), the Pan-African compressional tectonics was confined to shear zones and reflects localized intracontinental deformation. This gives yet another similarity with the Rayner Province (Clarke 1988).
The palaeotectonic model for East Antarctica of Boger et al. (2001) lends credibility to the consideration of Torsvik et al. (2001) that NW-India was not united with peninsular India by c. 750 Ma. In consequence, the short-lived magmatic event along the east coast of peninsular India at 800-780 Ma cannot be correlated with subduction processes at the Mozambique suture.
References
