1 Introduction

It is generally accepted that the Indian subcontinent was stabilized through accretion and assembly of several Archaean nuclei, namely, the Dharwar, Aravalli–Bundelkhand, Singhbhum, and Bastar (Bhandara) cratons by c. 2.6–2.5 Ga (Radhakrishna and Naqvi 1986; Rogers 1986; Ramakrishnan and Vaidyanadhan 2008; Sharma 2009; Saha and Mazumder 2012; Jayananda et al. 2013). Scattered across the Indian Archaean nuclei are several unmetamorphosed and only locally or weakly deformed volcano-sedimentary successions of Proterozoic age (figure 1). These erosional remnants are referred collectively in Indian literature as the ‘Purana basins’ (Holland 1907). Despite generally similar sedimentological features and lithofacies, as well as a similar time span of deposition between 2.0 and 0.7 Ga, the origin and evolution of these basins are not unanimously accepted (e.g., Basu and Bickford 2015; Joy et al. 2015; Saha et al. 2016).

Fig. 1
figure 1

The regional geology of India, with the WIB (western Indian block), NIB (northern Indian block), and SIB (southern Indian block), ADFB (Aravalli–Delhi fold belt) and CITZ (Central Indian Tectonic Zone) marked after Basu and Bickford (2015).

Full size image

Recently, Basu and Bickford (2015) divided the Indian continent into three main cratonic domains (southern Indian, northern Indian, and western Indian blocks – SIB, NIB and WIB, respectively; figure 1) to more readily compare the evolution of the Purana basins within the context of supercontinent development and the tectonic evolution of the Indian Proterozoic mobile belts, including the Eastern Ghats mobile belt (EGMB), the Central Indian Tectonic Zone (CITZ), and the Aravalli–Delhi fold belt (ADFB). However, it is now well understood that the evolution of a cratonic block is constrained by the tectonic history of the constituent terranes, which differ in intensity of deformation and metamorphism. Therefore, geological and geodynamic modelling requires robust age information to assess how and when these radically different terranes had stitched into the present day cratonic configuration. Despite growing evidence of the assembly of the first supercontinent Kenorland in the Late Neoarchaean (e.g., Bradley 2011; Meert 2012; Nance and Murphy 2013; Nance et al. 2013), comparisons and correlations have only been made between the Neoarchaean development of the Slave province in North America, a major component of Kenorland, and the Dharwar craton of southern India (Bleeker 2003). Recently, a correlation of different stages of evolution of the Purana basins with breakup of Kenorland at ca. 2.0 Ga, reassembly of Kenorland fragments to form the Columbia supercontinent at around 1.8 Ga, fragmentation of Columbia at 1.3 Ga, and finally formation of Rodinia at 900 Ma has been made (Basu and Bickford 2015; Saha et al. 2016 and references therein).

The tectono-stratigraphic development of the Nellore–Khammam schist belt (NKSB) in relation to global Proterozoic events, including possible links with the assembly of Columbia and its final dispersal, has been suggested by Saha et al. (2015). However, correlations between the Purana basins and the various Proterozoic mobile belts of southern India such as the Nellore–Khammam schist belt, Eastern Ghats mobile belt (EGMB) and the Karimnagar granulite belt (KGB) are unclear. Furthermore, their final incorporation into the southern Indian cratonic block (SIB) during supercontinent formation is controversially discussed (Basu and Bickford 2015).

The southern part of the SIB, comprising the western and eastern Dharwar cratons, two major Purana basins (i.e., Cuddapah basin in the south and P–G valley basin in the north), and the Nellore–Khammam schist belt, is an ideal area to test various models of supercontinent assembly, because the constituent terranes vary widely in their geographic disposition and in their stratigraphic and tectonothermal evolution. Furthermore, the SIB is bounded to the east by the Ongole domain of the EGMB, the tectonic journey with time of which is still debated. It is noteworthy that all the above-mentioned components of the SIB are bounded by discontinuities (thrust or normal faults) along their longitudinal trends. The available information on deformation features and metamorphic history indicate that they traversed differently through space and time to amalgamate at the present position. The components of the SIB could therefore be referred to as terranes.

Here we report new LA-ICP-MS U–Pb zircon ages from rocks of the Khammam schist belt and gneisses and granites within the Karimnagar granulite belt. We put forward a comprehensive picture of evolution of the SIB in the global context.

2 General geology

The geology of the study area comprises of the rocks of the Dharwar craton, the Proterozoic basins classified as part of the Purana basins of Holland (1907), the Nellore–Khammam schist belt, the Ongole and Eastern Ghats domains of the Eastern Ghats Mobile Belt (figure 2).

Fig. 2
figure 2

Geological map of the study area (modified after Chaudhuri et al. 2012) showing the location of the samples collected. The unit which was mapped as unclassified by Chaudhuri et al. (2012), was later classified as granitites and granitoids (Chaudhuri and Deb, personal communication).

Full size image

2.1 Dharwar craton and Purana basins

Based on the nature and relative abundance of greenstones, gneisses and granites the Dharwar craton has been divided into the west Dharwar craton (WDC) and the east Dharwar caton (EDC), separated by the Chitradurga Shear Zone, adjoining the western side of the Closepet Granite (Rogers 1986; Ramakrishnan and Vaidyanadhan 2008). The WDC comprises (a) older gneisses (\(\sim \)3.36–3.2 Ga) with TTG affinity, which have been grouped as Peninsular Gneiss-I (Balasubrhamanyan 2006) or as Peninsular Gneiss sensu stricto (Chadwick et al. 2000), (b) two generations of greenstone belts (older Sargur Group and younger Dharwar Supergroup) and late calc-alkaline to potassic plutons (Chadwick et al. 2000; Jayananda et al. 2008, 2013). In contrast, the EDC is a juvenile Neoarchaean province characterized by linear belts and rafts of 2.7–2.55 Ga supracrustal units (greenstones) with coeval TTG gneisses and migmatites, and voluminous 2.56–2.50 Ga calc-alkaline to potassic plutonic bodies, including the prominent Closepet Granite (Chadwick et al. 2000; Jayananda et al. 2008, 2013).

All the Purana basins formed in close spatial association with the stable cratons, and at present cover collectively about 20% of the Archaean basement (Kale and Phansalkar 1991). Outcrop in and around the NW–SE trending P–G valley basin is marked by symmetrical disposition of Archaean to Mesozoic rocks. The Proterozoic Godavari Supergroup (Chaudhuri and Chanda 1991), an unmetamorphosed and locally weakly deformed sedimentary rock succession, is exposed in two linear belts along the southwestern and northeastern margins of the P–G valley, separated by a linear belt of younger Gondwana rocks (220–65 Ma; Robinson 1971). The Proterozoic belts are flanked on their outer margins by granulites and gneisses (Rajesham et al. 1993; Vansutre et al. 2013); the Karimnagar granulite belt (KGB) and the gneisses of the Dharwar craton in the SW; and the Bhopalpatnam granulite belt (BGB) and gneisses of the Bastar craton in the NE. Most of the contacts among these lithologic units are faults that follow the NW–SE trend of the P–G valley (figure 2). The P–G valley basin developed along the join between the Dharwar and Bastar cratons (Naqvi and Rogers 1987) and it is referred to as a rift basin based on geophysical and stratigraphic analyses (Qureshy et al. 1968; Naqvi et al. 1974; Chaudhuri et al. 2002). However, there are alternative tectonic models that view the evolution of the Purana basins as foreland and sag basins (e.g., Acharyya 2003; Basu and Bickford 2015).

The other Proterozoic basin of the SIB is the oval shaped Cuddapah basin, the largest among all Purana basins, which hosts 1.9 Ga and younger unconformity-bound sedimentary rock successions (Saha and Tripathy 2012; Saha and Patranabis-Deb 2014; Saha et al. 2016). The eastern part of the basin evolved into the Mesoproterozoic Nallamalai fold-thrust belt, whereas the undeformed sedimentary rock successions in the west rest unconformably on the EDC gneisses.

2.2 Nellore–Khammam schist belt

The Nellore–Khammam schist belt (NKSB) is about 600 km long and 30–130 km wide (Hari Prasad et al. 2000). Its major southern extension (Nellore schist belt) separates the intracratonic Cuddapah basin to the west and the Ongole domain of the Eastern Ghats belt. In the northern extension, close to the P–G valley basin, the belt is referred to as the Khammam schist belt (Ramam and Murty 1997; Hari Prasad et al. 2000; Okudaira et al. 2001; Saha et al. 2015), where it is sandwiched between the EGMB and the Proterozoic sedimentary succession of P–G valley. The NKSB consists of several geologically and geochemically distinct volcano-sedimentary successions named the Vinjamuru Group, Kandra ophiolite complex (KOC), Kanigiri ophiolitic melange (KOM) and Udaigiri Group (Saha et al. 2015; Sain et al. 2017).

The Udaigiri Group is composed of predominantly greenschist facies metasedimentary rocks, whereas the Vinjamuru Group is dominantly an amphibolite facies volcano-sedimentary assemblage (Moeen 1998; Dobmeier and Raith 2003). There are two ophiolite complexes reported from the schist belt, namely the Kandra ophiolite complex (Vijaya Kumar et al. 2010) in the southern part and the Kanigiri ophiolite melange (Dharma Rao et al. 2011a) in the central part of the NKSB. The NKSB rocks show multiple deformation and metamorphism, as well as emplacement of granites and alkaline plutons (Saha et al. 2015 and references therein). Dobmeier and Raith (2003) grouped the granulite facies rocks of the Ongole domain to the NKSB and classified them as the Late Paleoproterozoic Krishna province, distinct from the Mesoproterozoic Eastern Ghats province. In contrast, Saha et al. (2015) proposed a Late Neoarchaean to Mesoproterozoic development of the NKSB independent of the Eastern Ghats belt. Although Ramakrishnan (2003) emphasised differences in tectono-stratigraphic features between the EDC greenstone belts and the Nellore–Khammam schist belt, this schist belt is often considered as the easternmost greenstone belt of the eastern Dharwar craton (Vadlamani 2010).

2.3 Karimnagar and Bhopalpatnam granulite belts

As mentioned earlier, the Dharwar and Bastar cratons are juxtaposed along a NW–SE trending join that was the keel of the P–G valley basin bordered by the Karimnagar and Bhopalpatnam granulites belts (KGB and BGB) to the southwest and northeast, respectively (figure 2). These granulites occur as several enclaves and discrete narrow bands mixed with granites, granite gneisses and amphibolites (Rajesham et al. 1993; Santosh et al. 2004; Vansutre et al. 2013). We use the term KGB to include all these units in a tectono-stratigraphic sense.

Both granulite belts have characteristic length: width ratios of at least 5:1. The Karimnagar granulite belt (KGB) is interpreted as an Archaean supracrustal-granite association, metamorphosed to granulite grade, representing the suture zone between the Dharwar and Bastar cratons (Rajesham et al. 1993). The KGB is characterised by at least two phases of deformation resulting in near isoclinal refolding with axial planar cleavages and the tectonothermal events in KGB match with those in the eastern Dharwar craton (Rajesham et al. 1993). The 300-km long Bhopalpatanam granulite belt (BGP) lies on the western edge of the Bastar craton (figure 2) and is composed of two pyroxene granulites, ultramafics, quartzite, calc-silicate rocks, Mg–Al metapelites (Vansutre and Hari 2010). The peak metamorphic condition of the high Mg–Al granulite within the KGB is estimated as 7.5–8 kb at 800–840\({^{\circ }}\)C (Prakash et al. 2017).

2.4 Eastern Ghats mobile belt

The Eastern Ghats mobile belt (EGMB), follows the east coast of India for over 1000 km adjoining the Dharwar, Bastar and Singhbhum cratons to the east. The EGMB is considered as an exotic distal Grenvillian terrane, juxtaposed to the Indian continent (Chaudhuri et al. 2012). The belt is a composite of accreted linear continental and oceanic fragments of moderate to high metamorphic grade, including granulite facies rocks, which have been intruded by suites of granites, anorthosites and nepheline syenites. A terrane boundary shear zone marks the juxtaposition of the EGMB with the Indian cratons, which is of strike slip character in the north and thrust character in the west and southwest (Ratre et al. 2010). The EGMB is divided into four provinces: the Jaypore province, the Rengali province, the Eastern Ghats province and the Krishna province, each with distinct tectonothermal histories (Dobmeier and Raith 2003). The Jaypore and Rengali Province accreted to proto-India prior to the 1000 Ma assembly of Rodinia (Mukhopadhyay and Basak 2009). The Eastern Ghats province records high grade metamorphism and orogenesis at 980–930 Ma during amalgamation of Rodinia (Korhonen et al. 2011). The Krishna Province records collisional orogenesis and high temperature metamorphism at 1600 Ma (Dobmeier and Raith 2003; Upadhyay et al. 2009; Henderson et al. 2013). The Krishna province is divided into the Ongole domain in the east and NSB in the west, which has been refuted by Saha et al. (2015), as discussed above in section 2.2, considering only the Ongole domain to be part of the EGMB.

3 Previous geochronology work

Geochronological data from different terranes of the SIB have enhanced our knowledge about many tectonothermal events of southern India. However, a more detailed data synthesis is presented below to better understand the evolution of the SIB.

3.1 Purana basins and eastern Dharwar craton

The EDC is composed of number of greenstone belts of Neoarchaean age (2700–2500 Ma) (Jayananda et al. 2013 and references therein). The Closepet Granite (2513±5 Ma; Friend and Nutman 1991; Jayananda et al. 2013) and equivalents form a widespread Neoarchaean plutonic phase in the Dharwar craton. It has been suggested that emplacement of the Closepet Granite represents the amalgamation of the WDC and EDC at \(\sim \)2.5 Ga (Jayananda et al. 2013).

Previous geological mapping combined with recent geochronological studies conducted on the Purana basin rocks suggest that deposition in these basins started as early as 1.9 Ga (Cuddapah basin) and basin evolution possibly ended at <1.0 Ga (Rasmussen et al. 2002; Ray et al. 2002; Patranabis-Deb et al. 2007; Malone et al. 2008; Das et al. 2009; Conrad et al. 2011; Sheppard et al. 2017). The Sullavai sedimentation continued until about 720 Ma or beyond in the P–G valley (Joy et al. 2015). However, the exact ages of the youngest sedimentary sequences in each of the basins are only poorly constrained.

The post-cratonization events in the EDC are composed of mafic dykes, kimberlites and lamproites. Protracted period of dyke activity (2.4–1.1 Ga) is reported from the EDC (table 1). Age of the kimberlites cluster around 1100 Ma and the lamproites in the Nallamalai fold belt at around 1400 Ma (Kumar et al. 2007; Chalapathi Rao et al. 2013; Chalapathi Rao and Srivastava 2016 and references therein).

Table 1 Summary of geochronology reported for dykes in the EDC.
Full size table

3.2 Nellore–Khammam schist belt

3.2.1 Vinjamuru Group

Sm–Nd isotope analysis from the Vinjamuru gabbro of the NKSB has yielded 2654±100 Ma whole rock age and a 1911±88 Ma isochron age (Vadlamani 2010). The data has been interpreted by Vadlamani (2010) to imply that the 2.7 Ga old NKSB was intruded by MORB-type gabbro at 1911 Ma in response to a major extensional event along the east Dharwar craton margin. The geochemical and geochronological evidences led (Vadlamani 2010) to consider the volcanic events at \(\sim \)1.9 Ga in the NKSB and Cuddapah basin (mafic–ultramafic sills, and gabbroic intrusives) as representing a single large anorogenic igneous event in the eastern Dharwar craton and NKSB. The andesites and rhyolites of the Vinjamuru Group have been dated through zircon Pb evaporation methods to be \(\sim \)1868 and 1771–1791 Ma, respectively (Vadlamani et al. 2012) and the authors propose that this volcanism occurred during a major convergent orogenic event along the southeastern margin of the eastern Dharwar craton.

The Vinukonda Granite intrusive to the Vinjamuru Group is dated as \(\sim \)1590 Ma (U–Pb zircon TIMS; Dobmeier et al. 2006), and represents the minimum age of the Vinjamuru Group.

Ghosh et al. (1994) report fission track and K–Ar dates showing pegmatite events at 1600, \(\sim \)1000 and 600±100 Ma from the Nellore mica belt (in the southern part of NKSB). The \(\sim \)500 Ma event is also represented by the phengite Rb–Sr ages from mylonites of the Vinukonda granite (Dobmeier et al. 2006).

Dharma Rao et al. (2011b) report an Sm–Nd model age of \(\sim \)1170 Ma from the Chimalpahad anorthosite complex in the NKSB and interpret it as an accreted arc fragment within the NKSB.

Yoshida et al. (1996) reported 1126 Ma from metapelite from the Khammam area based on an Pb–Pb mineral isochron. Okudaira et al. (2001), based on an Sm–Nd mineral isochron from amphibolites of the Khammam schist belt, proposed 824±53 Ma as the age of metamorphism of the Khammam schist belt and interpreted to be the result of the tectonic accretion of the EGMB to the Dharwar–Bastar craton. Okudaira et al. (2001) also report another thermal event from the same rocks at 481±16 Ma using a Rb–Sr mineral isochron.

3.2.2 Kandra ophiolite complex (KOC) and Kanigiri ophiolitic melange (KOM)

The Kandra ophiolite complex is dated as around 1850–1900 Ma (SHRIMP U–Pb zircon ages by Vijaya Kumar et al. (2010) and Sm–Nd isochron age by Vadlamani (2010)). The Kanigiri complex is dated at around 1330 Ma (LA-ICP-MS U–Pb zircon ages by Dharma Rao et al. 2011a). The timing of the post orogenic granite emplacement in the KOM is defined by the A-type Kanigiri granite intruding the KOM, dated at 1284 Ma (Sain et al. 2017). There is a 500 My age difference between these two ophiolite complexes and their history is proposed as an example of accretion along a craton margin over a prolonged period of convergence (Dharma Rao et al. 2011a). Saha (2011) proposed multiple cycles of ophiolite emplacement based on the divergent nature of thrusting in the area.

3.2.3 Udaigiri Group and Prakasam alkaline complex

The only age available from the Udaigiri Group is very poorly constrained at 1929±130 Ma based on a single grain xenotime analysis by Das et al. (2015).

The Prakasam alkaline plutons occurring along the boundary between the Vinjamuru and Ongole domains are dated by various methods at between 1242 and 1369 Ma (Upadhyay 2008 and references therein), interpreted to represent the rifting during the breakup of Columbia (Upadhyay 2008).

Table 2 Details of the samples collected.
Full size table
Fig. 3
figure 3

Examples of back scatter electron Z images captured using a CAMECA SX100 Electron Probe Micro Analyser for samples from the Karimnagar granulite belt. Ablation spots are 35 \(\upmu \)m in diameter and can be seen as circular holes.

Full size image

3.3 Karimnagar and Bhopalpatnam granulite belts

Santosh et al. (2004) report zircon (EPMA) ages from the KGB with the cores recording ages of up to 3.1 Ga and rims with ages of 2.6 Ga. In contrast, the cores of zircons recovered from the BGB demonstrate core ages of 1.9 Ga and the rims of 1.6–1.7 Ga. The monazites from the BGB also record the 1600±3 Ma age, interpreted as the most important tectonothermal event in the BGB. As there are no Mesoproterozoic tectonothermal events recorded from the KGB, the KGB and BGB are interpreted to have experienced different P–T conditions at different times (Santosh et al. 2004). The authors also note that the 1600 Ma age is not recorded from the KGB nor from the Bastar craton. Santosh et al. (2004) further note the absence of Grenvillian (ca. 1000 Ma) and Pan-African (ca. 520–550 Ma) ages from the KGB and suggest that the KGB was not directly involved in these younger collisional events that led to supercontinent amalgamation. The peak metamorphic age of the high Mg–Al granulite in the KGB the area has been recently reported by Prakash et al. (2017) as 2604±25 Ma (SHRIMP U–Pb zircon).

Granites intruding the supracrustal belt in the KGB are dated as 2490±115 Ma (Rb–Sr whole rock; Crawford 1969). This event is probably linked to the rim ages reported by Santosh et al. (2004) in the zircons.

3.4 Eastern Ghats mobile melt

The Jaypore and Rengali provinces accreted to proto-India prior to the 1000 Ma assembly of Rodinia (Mukhopadhyay and Basak 2009; Dasgupta et al. 2017). The EG province records high grade metamorphism and orogenesis at 980–930 Ma (SHRIMP U–Pb monazite dating) during amalgamation of Rodinia (Korhonen et al. 2011). The Ongole domain records collisional orogenesis and high temperature metamorphism at 1600 Ma (Dobmeier and Raith 2003; Upadhyay et al. 2009; Henderson et al. 2013).

Table 3 Results of U-Pb analysis of the zircons from sample EHV597 from the KGB.
Full size table

The crystallization of charnockites and enderbites in the centre of the Ongole domain is dated at 1720–1700 Ma (Kovach et al. 2001). The deposition of the sedimentary protolith in the Ongole domain is reported to be during 1.72–1.68 Ga and based on the U–Pb and Lu–Hf data has been proposed not to be sourced from the Dharwar craton, but from the Napier Complex in Antarctica (Henderson et al. 2013). A small proportion of pre-Neoarchaean detrital zircons from the Ongole domain is also reported by Henderson et al. (2013). The granulite facies metamorphism in the Ongole domain is dated as 1.68–1.6 Ga (Henderson et al. 2013). It has been proposed that the Ongole domain accreted to proto-India during 1.68–1.6 Ga, as part of a linear accretionary orogenic belt (Henderson et al. 2013).

North of the Godavari Rift the granulite metamorphism and magnetism are mainly in the interval 1.1–1.0 Ga and define the connection of EGMB with the Reyner–Napier complex of the Eastern Antarctica during the assembly of Rodinia (Rickers et al. 2001; Dobmeier and Raith 2003; Henderson et al. 2013; Dasgupta et al. 2017; Meert et al. 2017).

Saha et al. (2015) postulate long term episodic growth of the eastern craton margin of India in the Proterozoic, culminating in the final docking of the EGMB at \(\sim \)500 Ma (also Biswal et al. 2007; Vijaya Kumar and Leelanandam 2008).

4 Methodology

4.1 Sampling

Samples (\(\sim \)5 kg) of granite and gneiss were collected from the southeastern portion of the P–G valley basin close to the boundary with the Eastern Ghats domain, which tectono-stratigraphically belongs to the Karimnagar granulite belt and the Khammam schist belt, i.e., the northern part of NKSB (table 2 and figure 2).

4.2 Analytical method

The samples were concentrated at the De Beers sample treatment centre, Bangalore, India. The treatment process consisted of crushing, screening, standard dense media separation (DMS), low intensity magnetic separation and heavy liquid separation through a lithium heteropolytungstate solution (LST). The resultant heavy mineral concentrate was partitioned to two size fractions, viz., (+0.3–0.5 mm) and (+0.5–1.0 mm). The zircons were identified visually under binocular microscopes and analysed at the De Beers Exploration Indicator Mineral Laboratory (DBE IML), Johannesburg. The handpicked zircons were mounted in 25 mm diameter epoxy mounts and polished to fully expose the midsection of the grains. Epoxy mount surfaces were cleaned using 96% ethanol and were put in a vacuum chamber to ensure sufficient outgassing of the epoxy.

The U–Pb isotope analyses were done at the DBE IML, using a New Wave 193 Excimer laser ablation system equipped with a Large Format ablation cell and interfaced with a Thermo Fisher X-Series 2 quadrupole ICP-MS. The ICP-MS instrument was optimised for maximum sensitivity, stability and low background using the NIST-610 glass standard. Ablations were carried out in a helium atmosphere, with a laser spot size of 35 \(\upmu \)m. Ablation sites were placed at the centre of the zircon grains and away from cracks. Due to the small grain size, only one spot analysis was carried out per grain and it was not possible to carry out analyses of thin zircon rims. Unknowns and standards were ablated for 60 sec, followed by generous washouts of 5 min. The masses \(^{206}\)Pb, \(^{207}\)Pb, \(^{208}\)Pb, \(^{235}\)U, \(^{238}\)U and \(^{232}\)Th were measured for 50 ms each, yielding a total sweep time of 300 ms. Data reduction was done using the GLITTER Laser Ablation data reduction software (Van Achterbergh et al. 2001). The GJ-1 zircon standard (slightly discordant, ID-TIMS \(^{207}\)Pb/\(^{206}\)Pb age 608.5±0.4 Ma, Jackson et al. 2004) was used as the calibration standard. Concordia diagrams and age calculations were performed using IsoPlot v. 3.70 (Ludwig 2001). BSE images of the analysed zircon grains (figures 3 and 5) were captured using a CAMECA SX100 Electron Probe Micro Analyser.

5 Results

5.1 Samples from Karimnagar granulite belt

EHV597 (Augen gneiss): A total of 54 zircons were recovered from the sample; these are euhedral to subhedral and are 150–250 \(\upmu \)m in diameter (figure 3). Zircons are mostly unzoned with a few grains showing evidence of weak zonation (figure 3). Fourteen of the analyses are highly (>10%) discordant (table 3). These highly discordant data are rejected and are not used for the interpretation (table 3). The obtained U–Pb zircon age data form two concordant age groups. The dominant age group has a weighted average age of 1525±25 Ma (MSWD 25). The average Pb and U concentrations (table 3) are 148 (81–355 ppm) and 598 ppm (346–1398 ppm), respectively. The Th/U ratio (table 3) average is 0.8 (0.2–1.5). The second group (5 zircons) ranges in age between 873±13 and 1056±20 Ma and has a weighted average age of 953±79 Ma with MSWD of 18 (figure 4). The average Pb and U concentrations (table 3) are 20 (10–25 ppm) and 136 ppm (64–169 ppm), respectively. The Th/U ratio (table 3) average is 0.14 (0.1–0.3).

EHV592 (Banded gneiss): Zircons recovered are euhedral in shape and are 150–250 \(\upmu \)m in diameter and are mostly unzoned (figure 3). A total of 42 zircons have been analysed. Unfortunately 19 grains produced highly discordant (>10% discordance) results (table 4), which are rejected and not used in the interpretation. Analysis of this sample presents a single age group (figure 4) with an intercept age of 1621±36 Ma (MSWD 0.35). The average Pb and U concentrations (table 4) are 76 (47–138 ppm) and 299 ppm (184–509 ppm), respectively. The Th/U ratio (table 4) average is 1.1 (0.5–1.6).

EHV596 (Foliated granite): Zircon grains recovered from this sample are euhedral to subhedral, 150–250 \(\upmu \)m in diameter and in some grains there is evidence of zoning (figure 3). A total of 69 zircon grains have been recovered and analysed (table 5), of which four analysis have suspicious U, Pb and Th concentrations, which have been removed (table 5), further removing of another 5 apparently inherited zircons, result in an intercept age of 2428±16 Ma (figure 4). This is interpreted as the intrusion age of the granite. The average Pb and U concentrations (table 5) are 49 (15–197 ppm) and 113 ppm (35–454 ppm), respectively. The Th/U ratio (table 5) average is 1.0 (0.5–2.4).

EHV593 (Schistose granite): Zircons are similar to those recovered from EHV596. A total of 30 grains have been analysed, of which 9 have high discordance and were removed (table 6). The U-Pb isotope system appears to be disturbed and did not result in an interpreted concordia age for this sample. Weighted average of the accepted single system ages (figure 4) result in 2480±31 Ma (MSWD 6.0). This is interpreted as the age of the granite with low confidence. The average Pb and U concentrations are 124 (45–522 ppm) and 288 ppm (106–1253 ppm), respectively. The Th/U ratio (table 6) average is 0.9 (0.2–2.0).

Fig. 4
figure 4

U–Pb concordia diagram of zircons from the Karimnagar granulites and the granites. EHV593 is the weighted average of accepted single system ages.

Full size image

5.2 Samples from Khammam schist belt

EHV594 (Augen gneiss): Zircons recovered are euhedral to subhedral, are 150–250 \(\upmu \)m in diameter and demonstrate some evidence of zoning (figure 5). A total of 43 zircon grains were recovered and analysed out of which 12 analyses are highly discordant (>10%; table 7) and are not used in the interpretation.

There are four different age groups evident from the data (figure 6 and table 7). Age group 1 is represented by two grains only with very concordant analyses. A weighted average age of 3212±36 Ma (MSWD 0.102) is obtained for the two grains. The average Pb and U concentrations (table 7) are 154 (151–158) and 258 ppm (244–271 ppm), respectively. The Th/U ratio (table 7) average is 0.5 (0.4–0.5). Age group 2 has seven concordant data points and is represented with weighted average age of 2587±130 Ma (MSWD 23). The average Pb and U concentrations (table 7) for this group are 74 (44–103 ppm) and 164 ppm (81–217 ppm), respectively. The Th/U ratio (table 7) average is 0.9 (0.5–1.4). Age group 3 is the most prominent and is represented by 21 analyses of low discordance (<10%). This age group is represented with weighted average age of 1807±37 Ma (figure 6). The average Pb and U concentrations (table 7) for this group are 131 (32–405 ppm) and 454 ppm (115–1472 ppm), respectively with the Th/U ratio (table 7) average of 0.6 (0.2–1.9). The 4th group is represented by a single concordant zircon grain with an age of 522±25 Ma (figure 6 and table 7). The Pb and U concentrations for this zircon is 37 and 484 ppm, respectively with the Th/U ratio of 0.3 (table 7).

Table 4 Results of U-Pb analysis of the zircons from sample EHV592 from the KGB.
Full size table
Table 5 Results of U–Pb analysis of the zircons from granite sample EHV596 from the KGB.
Full size table
Table 6 Results of U-Pb analysis of the zircons from granite sample EHV593 from the KGB.
Full size table

EHV591 (Quartzo-feldspathic gneiss): Zircons recovered are euhedral to subhedral and are 150–250 \(\upmu \)m in diameter (figure 5). Zircons demonstrate stronger zonation than the samples from the KGB (figure 5). A total of 57 zircons were recovered and analysed, with eight analyses of high discordance (table 8 and figure 6). There are three different concordant age groups evident from the data (figure 6 and table 7).

Age group 1 is represented by three grains with <10% discordance and with a weighted average age of 3043±230 Ma. The average Pb and U concentrations (table 8) for this group are 50 (7–129 ppm) and 127 ppm (77–226 ppm), respectively with the Th/U ratio (table 8) average of 0.4 (0.4–0.5). Age group 2 is most prominent with 45 data points, five of which are highly discordant (>10%) and are not used in the interpretation (table 8 and figure 6). This age group ranges between 2005±28 and 2541±29 Ma and with a weighted average age of 2396±24 Ma. The average Pb and U concentrations (table 8) for this group are 104 (6–441 ppm) and 279 (19–1034 ppm), respectively with the Th/U ratio (table 8) average of 1.2 (0.1–2.7). The third age group is represented by a single zircon grain giving an age of 472±9 Ma (figure 6 and table 8). The Pb and U concentrations for this zircon is 200 and 82 ppm, respectively with the Th/U ratio of 1.8 (table 8).

Table 7 Results of U–Pb analysis of the zircons from sample EHV594 from the Khammam belt of NKSB.
Full size table

6 Discussion

Unravelling of overprinting multiple tectonothermal events by the in-situ U–Pb geochronology in single samples have been reported earlier (Wang et al. 2007; Zhang et al. 2008; Ma et al. 2012), based on resetting of the U–Pb system in pre-existing zircons due to tectonothermal effects or multiple recrystallization of new zircons during metamorphism (Carson et al. 2002). The zircons grown under metamorphic conditions have lower Th/U, typically lower than 0.1, whereas the magmatic zircons have higher Th/U ratios 0.18–0.47 (Rubatto 2002). However, Wang et al. (2011) demonstrate a huge variation in the Th/U in the granitic rocks (0.1–3.79) and in the intermediate rocks (0.02–6.82). Zircon growth with high Th/U ratios (>1) has also been reported during UHT metamorphism (Carson et al. 2002). The Th/U ratio for the gneissic samples from the KGB and the KSB show high values (0.14–1.8) on average (tables 3–8). We interpret the concordant ages obtained from the KGB and KSB to be representing relict magmatic zircons (representing either original igneous protolith or detrital), which has the U–Pb system reset by the tectonothermal events.

Fig. 5
figure 5

Examples of back scatter electron Z images captured using a CAMECA SX100 Electron Probe Micro Analyser for samples from the Khammam schist belt. Ablation spots are 35 \(\upmu \)m in diameter and can be seen as circular holes.

Full size image

The NKSB, EGMB, KGB are important pieces in the reconstruction of the Proterozoic history of the Indian continent and its part in the global tectonic framework and supercontinent cycle (Meert 2012; Nance et al. 2013; Meert et al. 2017). The results presented in this contribution demonstrate that the samples of KGB and northern NKSB are unique in terms of their zircon U–Pb geochronology suggesting a distinct, tectonothermal history relative to each other (figure 7). The different rock samples were collected from localities currently in close proximity (within \(\sim \)50 km of one another) from the eastern part of KGB and the northern Khammam schist belt of the NKSB.

The granite samples from the KGB are \(\sim \)2400 Ma old (intercept age of EHV596 is 2428±16 Ma and the schistose granite sample (EHV593) show a weighted average age of 2480±31 Ma). The interpreted age of 2400–2500 Ma for the granite within the KGB is in good agreement with the previously reported Rb–Sr whole rock age result of 2490±115 Ma (Crawford 1969). This Late Archaean event is also reported as a metamorphic imprint in the zircons of granulites from KGB by Santosh et al. (2004). However, Rajesham et al. (1993) consider the 2.5 Ga event to represent peak metamorphism of the KGB rocks to granulite grade. Recently, Prakash et al. (2017) date this peak metamorphism earlier than the granite event at 2604±25 Ma.

The two samples of gneisses (EHV597 and EHV592) from the KGB were collected from the same area and combined constitute a prominent 1560–1600 Ma age group. The 1000 Ma old event recorded in EHV597 is not evident from the concordant ages in EHV592 (figure 4). Combining the results from all four samples, the KGB is interpreted to record metamorphic resetting at \(\sim \)1600 Ma with a later tectonothermal event at \(\sim \)1000 Ma. The granite intrusions at 2400–2500 Ma most likely form part of the regional Late Archaean granite magmatism within the Dharwar craton (Jayananda et al. 2013). The 3100 and 2600 Ma ages reported by Santosh et al. (2004) for the western KGB could not be reproduced in our samples. This suggests that the 1600 Ma event was intense enough to reset the U–Pb system in the zircons completely. However, we report a 1000 Ma Grenvillian tectonothermal event in the KGB, which was hitherto unreported.

Santosh et al. (2004), based on the absence of the Mesoproterozoic tectonothermal event in the KGB and the prominence of the 1600 Ma event in the BGB, proposed that the KGB and BGB evolved under different P–T conditions at different times. We report an intense 1600 Ma event from the southeastern part of the KGB and propose that the two granulite belts are probably affected by the same tectonic event. The 1600 Ma event marks the fragmentation of the craton and the development of the P–G valley (Chaudhuri et al. 2012). The \(\sim \)1600 Ma event was also identified in detrital zircon populations from the Sullavai Group of the P–G rift (Joy et al. 2015). We note that Basu and Bickford (2015) consider the glauconite ages reported from the Somanpalli Group and Pandikunta Limestone to be unreliable (cf., Conrad et al. 2011). They consider the deposition of Somanpalli rocks to have started after the eastern Dharwar–Bastar amalgamation at \(\sim \)1600 Ma, with deformation being due to the continued crustal shortening that followed craton amalgamation.

The present report of a 1000 Ma tectonothermal event from the gneissic rocks of the KGB is contrary to Santosh et al. (2004). The presence of this Grenvillian age in the KGB, which is very prominent in the EGMB, highlights that both the KGB and EGMB were part of the global Grenvillian orogenic belt during the Rodinia supercontinent formation (figure 7). Unfortunately, the age of reassembly of the Dharwar and Bastar cratons cannot be constrained. Based on geological and structural analysis, it was suggested that suturing occurred subsequently to Mulug Group deposition (1565 Ma), but prior to Albaka and Sullavai Group deposition (Chaudhuri et al. 2012). The age of the Sullavai Group has recently been shown to be younger than 710 Ma (Joy et al. 2015). The 1000 Ma event reported herein is thus best interpreted as the age of ‘Grenvillian’ orogeny by which the Dharwar and Bastar cratons were reassembled.

Table 8 Results of U–Pb analysis of the zircons from sample EHV591 from the Khammam belt of NKSB.
Full size table

It has been previously demonstrated that the NKSB records multiple deformation and metamorphism cycles, as well as granitic and alkaline intrusive magmatism (Saha et al. 2015 and references therein). Our NKSB samples represent unique and contrasting geochronological signatures to the samples from the KGB including the granite. The samples EHV594 and EHV591 show presence of Archaean zircons. The 3100–3250 Ma ages, which are interpreted as relict zircons (either representing protolith or detrital zircons as the nature of the protolith of the gneisses sampled, whether igneous or sedimentary has not been established during this study) of the Khammam schist belt of the NKSB (figures 6 and 7).

For sample EHV594, the prominent age range is \(\sim \)1800 Ma. This age range is not very apparent in EHV591, where 2400 Ma age is most prominent. Sample EHV594 also shows the 2400–2700 Ma age range, though not as pronounced as recorded in EHV591 (figure 6). The 1900 Ma age has also been reported previously in the Sm–Nd isochron age of the gabbro intrusions (Vadlamani 2010) and the \(^{207}\)Pb/\(^{206}\)Pb zircon evaporation ages of andesitic rocks (Vadlamani et al. 2012) in the Vinjamuru Group, and the U–Pb xenotime age from the Udaigiri Group (Das et al. 2015). The KOC with an age range of 1850–1900 Ma (Vadlamani 2010; Vijaya Kumar et al. 2010) is postulated to represent orogenic accretion that led to Columbia supercontinent assembly (Saha et al. 2015).

Fig. 6
figure 6

U–Pb concordia diagram of zircons from the Khammam schist belt of NKSB.

Full size image
Fig. 7
figure 7

Summary of age groups obtained in the study with regional events. The height of the individual bars is proportional to the number of grains within that age group.

Full size image

Both EHV591 and EHV594 record a ca. 500 Ma age from a few zircon grains. The 500 Ma event is also recorded in the amphibolites from the Khammam schist belt (Okudaira et al. 2001), the mylonites of the Vinikonda Granite (Dobmeier et al. 2006) and from pegmatites in the Nellore schist belt (Ghosh et al. 1994). Combining the results from the two samples (EHV591 and EHV594), we interpret that the Khammam belt of NKSB represents Archaean protolith (>2600 Ma), which has been affected by four major tectonothermal events which has affected/reset the U–Pb geo-chronometer in zircons. The first event at 2500–2600 Ma could be associated with the first amalgamation of the NKSB with the Dharwar craton, as indicated by the granite event in the KGB and the widespread granitic event in the Dharwar craton (Jayananda et al. 2013). This event is also represented by granite magmatism and zircon rim ages in the KGB (Santosh et al. 2004). We propose that the terranes were juxtaposed at 2600–2500 Ma to form an integral part of proto-India. The 1800–1900 Ma event identified in this study, combined with previously reported events from the Nellore schist belt (Vadlamani 2010; Vadlamani et al. 2012; Das et al. 2015), support another continental accretion event in the eastern part of the southern Indian craton as proposed by Saha et al. (2015). The \(\sim \)1600 Ma event is considered to represent suturing of the Ongole domain with the Nellore schist belt (Bose et al. 2011; Vijaya Kumar et al. 2011; Henderson et al. 2013) and this event is evident in sample EHV594 (figure 6). The \(\sim \)1000 Ma event from the NKSB (Yoshida et al. 1996; Okudaira et al. 2001) was not identified in the present study. However, zircons from the NKSB show evidence of the \(\sim \)500 Ma Pan-African event that led to the formation of Gondwana (figures 6 and 7).

7 Conclusions

The northern part of the Nellore–Khammam schist belt known as the Khammam schist belt, and the Karimnagar granulite belt, which are juxtaposed at a high angle to each other are unique in terms of their tectonothermal evolution, as recorded by U–Pb zircon age patterns.

Southern India has been affected by multiple events of rifting and collision-accretion throughout its Proterozoic history (see also Basu and Bickford 2015; Saha et al. 2015). The present study indicates that the Khammam schist belt and the northern part of the NKSB probably formed part of cratonic India by about 2500 Ma. The Dharwar and Bastar cratons amalgamation was initiated at \(\sim \)2600 Ma with the formation of the granulites and culminating with the emplacement of the granites at \(\sim \)2400 Ma. The schist belt has experienced another orogenic accretion at about 1900 Ma probably assigned to the formation of Columbia. The Dharwar and Bastar cratons separated to form the P–G valley at \(\sim \)1600 Ma along with the metamorphic resetting of zircons within the KGB and BGB granulites and gneisses. During this same time period (\(\sim \)1600 Ma), at an orthogonal direction to P–G valley, accretion tectonics would have been ongoing whereby the Ongole domain was accreted to the NKSB (Henderson et al. 2013). The P–G valley area inverted to be under a compressional regime, resulting in closure of the basin and the reassembly of the Dharwar and Bastar cratons at about 1000 Ma, which could be linked to the formation of Rodinia. The NKSB was affected by the late Pan-African tectonothermal event at ca. 500 Ma that possibly marked the final episode of its assembly with the Eastern Ghats province of the EGMB, since then forming part of the present day Indian subcontinent.

Our study highlights that plate accretion and rifting events in the eastern part of the Dharwar craton and between the Dharwar and Bastar cratons have been repeated through time with highly complex events. Southern India’s present tectonic configuration is the result of multiple tectonic events between 2500 and 500 Ma, which occurred in different areas and at different times, which can potentially be linked to the formation and dispersal of three supercontinents.