Introduction

The tectonic configuration of collision zones is thought to be a product of long-lived accretion between distinct continental blocks in which crustal domains develop complex deformation structures and metamorphic history (Harris et al. 1994; Brown et al. 2006). Based on the following observations, such an intercontinental collision zone undoubtedly provides critical constraints on understanding the geodynamic behavior of accreted crustal fragments during the collision process. First, the lithodemic units in the collision zone preserve vital information about their deformation and metamorphism during burial and exhumation processes (Vernon 1978; Norris et al. 1990; Jamieson et al. 2002). The application of in situ geochronology on these lithologies (e.g., U–Th–(total) Pb monazite dating; Suzuki and Adachi 1991; Montel et al. 1996; Williams and Jercinovic 2002) provides better insights into the tectono-metamorphic evolution of polydeformed crustal provinces (e.g., England and Thompson 1984; Zhao and Cawood 1999). Second, detailed accretion-related tectonic processes documented from the collision zone have been proven essential prerequisite for reconstructing dispersed crustal blocks within supercontinent assembly (e.g., Kluth and Coney 1981; Beaumont et al. 1996; Cawood and Buchan 2007; Zhuang et al. 2015).

The Khammam Schist Belt (KSB), a part of the Proterozoic collision zone in southeastern India, is sandwiched between the Eastern Dharwar Craton (EDC) and the Eastern Ghats Mobile Belt (EGMB) (Fig. 1a; Ramam and Murty 1997; Hari Prasad et al. 1999; Okudaira et al. 2001). The belt is a collage of imbricated metasedimentary sequences that were accreted to the eastern margin of the EDC (Ramam and Murty 1997; Saha 2002). During this accretion and cratonization, the lithologies of the KSB and neighboring domains have been affected by multiple episodes of syn- to post-collisional deformation and metamorphism. These tectonic activities indicate that the belt preserves a protracted accretionary history along the eastern boundary of the EDC (Hari Prasad et al. 2000). A previous metamorphic study from the KSB amphibolites has identified peak P–T conditions of 8–12 kbar and 650–750 °C, which reveals the evidence for high-T metamorphism in the belt (Hari Prasad et al. 2000). The timing of this high-grade metamorphic event was determined at 824 ± 43 Ma using Sm–Nd mineral isochron (Okudaira et al. 2001). According to Hari Prasad et al. (2000), the KSB could represent a terrane evolved in an oceanic or continental margin island arc setting. However, these petrological and geochronological studies on the KSB were mostly retrieved from amphibolites and provide limited information on the pressure (P)–temperature (T)–time (t) history of the KSB. As a result, the tectonic evolution of the accretionary system in and around the KSB remains unclear. Therefore, a multidisciplinary approach that integrates structures and phase equilibria modeling with in situ geochronology is noteworthy to unravel the protracted accretionary history within the Khammam region of the southeastern Indian shield.

Fig. 1
figure 1

a Simplified tectonic map of central and southern India showing the distribution of major Archean cratons and Proterozoic mobile belts (modified after Narayanaswami 1966). The box represents the location of the Khammam region in southeastern India. b Generalized geological map of the Khammam region showing different lithological units with mesoscopic foliation patterns. Locations of the samples used for P–T estimation (KM-24, 29, and 26C) and monazite dating (KM-26C and 29) are also shown on the map. Equal-area projections show a summary of linear and planar data obtained from c metapelites of the KSB and d quartzofeldspathic gneisses of the EGMB

Full size image

In this study, we investigated various localities in and around the KSB, which comprise a wide variety of lithologies, including metapelites of the KSB and quartzofeldspathic gneiss of the EGMB (Fig. 1b). We present detailed petrological observations, pseudosection modeling, and U–Th–(total) Pb monazite ages to determine the metamorphic P–T–t history of these distinct lithodemic units. The specific issues addressed in this study include: (a) the relationship between tectonic fabrics and different metamorphic assemblages in the metapelites and quartzofeldspathic gneisses and (b) the possible timing of various deformation and metamorphic events. Further, we wanted to evaluate whether the KSB represents a possible extension of the Vinjamuru Domain that is located in the Nellore Schist Belt (Fig. S1b), as suggested by Sangam and Pavanaguru (2013). Finally, these new results from the Khammam region are combined with existing data to discuss its implications to counterparts in Columbia and Rodinia assemblies.

Geological background

The NNE–SSE trending Nellore–Khammam schist belt (NKSB) has traditionally been regarded as a collision orogenic belt that was sandwiched between the Archean EDC to the west and the Proterozoic EGMB to the east (Fig. S1; Leelanandam and Reddy 1988; Ramam and Murty 1997; Hari Prasad et al. 1999; Okudaira et al. 2001). The curvilinear NKSB extends for 600 km in length with variable width between 30 and 130 km (Fig. S1b; Okudaira et al. 2001). Based on lithological and geochronological data, the NKSB has been subdivided into the Nellore Schist Belt (NSB) and Khammam Schist Belt (KSB) in the south and north, respectively (Fig. S1b; Hari Prasad et al. 1999; Okudaira et al. 2001). The NSB comprises amphibolite facies Vinjamuru Domain and greenschist facies Udayagiri Domain (Fig. S1b). According to Sangam and Pavanaguru (2013), the upper amphibolite facies KSB rocks are considered the northern extension of the Vinjamuru Domain. The KSB lithologies form the basement to the Proterozoic-age Pakhal Group of rocks and Phanerozoic-age Gondwana sediments (Fig. 1b; Rajneesh Kumar et al. 1999). The eastern part of the KSB sporadically exposes parts of the Ongole Domain of the EGMB (Figs. 1b and S1b), typically dominated by high-grade enderbite and charnockite rocks (1.72–1.70 Ga; Kovach et al. 2001; Dobmeier and Raith 2003). Based on the existing geological and isotopic data, the NKSB and Ongole Domain together constitute the Krishna Province (Figs. 1 and S1), represents a discrete crustal province in the southern part of the EGMB (Dobmeier and Raith 2003; Sharma 2009). The present study area in the Khammam region is situated along the northwestern boundary of the Krishna Province (Figs. 1a and S1).

The KSB (Fig. 1b) comprises various lithological units, including metapelites, amphibolites, phyllites, and quartzites (Ramam and Murty 1997; Hari Prasad et al. 1999; Okudaira et al. 2001; Vijaya Kumar et al. 2006). The amphibolites are the most abundant lithology in the KSB and experienced three stages of deformation (D1, D2, and D3; Hari Prasad et al. 1999). The F1 and F2 folds were formed by two-stage (NE–SW and NW–SE directional) compressional events, while F3 folds were developed as a result of NW–SE shearing. In contrast, Rajneesh Kumar et al. (2000) identified four deformation events from the KSB lithologies. However, the metamorphic and geochronological evolution of the KSB is poorly investigated in previous studies. Hari Prasad et al. (2000) identified a high-T metamorphic event in the KSB amphibolites, which was dated using Sm–Nd mineral isochron at 824 ± 43 Ma (Okudaira et al. 2001). To determine the comprehensive geodynamic setting of this terrane, it is important to establish P–T–t constraints from all the involved lithologies during the accretion process. Hence, the lack of comprehensive information on metamorphic evolution and petrologically constrained age data from the KSB metapelites and associated lithologies allow us to revisit the KSB and its adjacent domains.

Field relations and structural setting

In this study, we focused on the Khammam region that is located in the western portion of the EGMB (Figs. 1b and S1b). The investigated lithologies include: (a) metapelites of the KSB, and (b) quartzofeldspathic gneisses of the EGMB (Figs. 1b, 2a–h). The sample locations are shown in Fig. 1b.

Fig. 2
figure 2

Field photographs showing lithological and structural relations from a–d the KSB metapelites and e–f quartzofeldspathic gneisses and charnockites of the EGMB. Photographs b, d, and f are in plan view, and the remaining images represent section view. In the plan view, photographs, pencil, marker, and hammerhead point north direction. a Tight isoclinal folds on early S0S foliation with the development of S1S axial planar fabric in schistose rocks. b The image shows possible recumbent geometry of F2S folds, which are interfolial to the S2S foliation. c Plan view showing spaced S3S crenulation cleavage on gently dipping S2S schistosity. Note that the F3S folds are plunging gently towards NNE. d The general appearance of garnet–kyanite schist within the supracrustal sequence. Note the presence of profuse garnet and kyanite porphyroblasts. e Banded foliation (S2G) in quartzofeldspathic gneiss, defined by sub-parallel leucosomes. The inset shows vestiges of folded S1G early foliation intrafolial to the S2G planes. f Development of shear foliation (S3G) in response to dextral shears within charnockites of the EGMB

Full size image

In the northeastern part of the Khammam region, metapelites of the KSB occur as discontinuous exposures within amphibolites of the KSB and migmatitic gneisses of the EDC (Fig. 1b). The contact relations between these lithologies are highly obscure, which might have shared a basement–cover relationship. The metapelites include garnet–biotite–muscovite ± chlorite schists, garnet–kyanite schists, and phyllites (Fig. 2a–d). These metapelites have recorded three episodes of deformation (D1S, D2S, and D3S), which resulted in the development of S1S, S2S, and S3S tectonic fabrics (Figs. 1c and 2a–c). Tight isoclinal folds on primary foliation characterize the D1S deformation in metapelites (Fig. 2a). The possible recumbent folding on early S0S//S1S foliation develops gently dipping S2S axial planar fabric during the D2S deformation (Fig. 2b). The NE–SW to NW–SE striking pervasive (< 30° dip) S2S foliation is defined by the shape-preferred orientation of muscovite + biotite + kyanite mineral aggregates (Fig. 2d). However, the existence of recumbent geometry for F2S folds is restricted to specific locations (Fig. 2b). During D3S deformation, the S2S schistosity is crenulated and produces NNE–SSW-trending steep S3S axial planar fabric with gently plunging (< 30°) F3S folds towards NNE (Figs. 1c, 2c).

The exposures of EGMB quartzofeldspathic gneisses in the study area are typically restricted to the eastern portion of the Khammam region (Fig. 1b). The rock is represented by sub-centimeter to centimeter-scale quartzofeldspathic layers with variable proportions of garnet, biotite, quartz, plagioclase, and sillimanite (Fig. 2e). These gneisses were migmatized, and leucosomes were often folded (Fig. 2e). The steeply (> 70°) dipping ENE–WSW-trending S2G foliation defines the pervasive tectonic trend in the quartzofeldspathic gneiss of the EGMB (Figs. 1d, 2e). The remnants of early S1G foliation are preserved intrafolial to the penetrative S2G foliation in low-strain zones of the D2G deformation (inset in Fig. 2e). The leucosome layers are transposed along the S2G foliation, suggesting that the possible partial melting in quartzofeldspathic gneisses was syntectonic to the D2G deformation. Further, the S2G foliations in associated charnockites are displaced by later dextral shears, producing N–S-trending and steeply dipping (> 70°) S3G foliation (Fig. 2f).

Methods

Electron probe microanalysis

Mineral compositions were determined using a CAMECA SX-FIVE electron probe microanalyzer (EPMA) equipped with five wavelength-dispersive spectrometers at the Department of Earth Sciences, Indian Institute of Technology Bombay (India). A focused beam with 15-kV accelerating voltage and 20-nA beam current was used during the silicate analyses. Calibration was performed using mineral standards including albite (Si, Na), diopside (Mg), Al2O3 (Al), K-feldspar (K), Fe2O3 (Fe), rhodonite (Mn), Cr2O3 (Cr), TiO2 (Ti), ZnS (Zn), BaSO4 (Ba), apatite (Ca, F) and NaCl (Cl). Each analysis was performed for ~ 2 min, in which peak and background counts for all elements were measured for 20 s and 10 s, respectively. The X-PHI method was used for correcting matrix absorption effects (Merlet 1992). For identifying zoning patterns, X-ray element mapping of garnet was performed with operation conditions of 15-kV accelerating voltage, 50-nA beam current, and 50–70 ms dwell time per pixel. The detailed mineral compositions of various lithologies are presented in Table S1.

EPMA monazite dating was performed with a focused beam using 20-kV accelerating voltage and 200-nA beam current. ThMa, UMb, and PbMa X-ray lines were calibrated using Th-glass, U-glass, and crocite standards. Simultaneously, two LPET crystals were employed in analyzing ThMa (160 s), UMb (160 s), and PbMa (240 s) X-ray lines. The error propagated during monazite dating primarily depends on the precise estimation of ThO2, UO2, and PbO compositions, and uncertainties associated with the calibration of other elements and monazite standards (Jercinovic et al. 2012; Prabhakar 2013). Since these elements were analyzed for long counting times, the sub-counting method (5 cycles) was employed to minimize the effects of sample surface damage on the peak-background measurements of ThMa, UMb, and PbMa (Spear et al. 2009). The measured intensity of PbMa was corrected for the interference of ThM2-O4, ThMζ1, ThMζ2, YLC2, YLC3, and LaLa intensities, whereas UMb was corrected for ThMc and ThM3-N4 interferences (Deshmukh et al. 2017). The resultant data were processed using the chi-square test to exclude inconsistent and anomalous counts and minimize error propagation on the final quantification of ThO2, UO2, and PbO concentrations (Prabhakar 2013). Further, empirical calibrations were typically associated with external uncertainties of 0.8%, which were propagated through the acquisition of ThO2, UO2, and PbO compositions (Dahl et al. 2005). Therefore, the counting errors obtained for monazite dates are additionally corrected for external uncertainties and are presented with 2σ errors (95% confidence level). The Steenkampskrall monazite standard (SHRIMP age: 1030 ± 6 Ma; Knoper et al. 2000) was used for monitoring the spectrometer and analytical drift during the monazite analysis.

High-resolution X-ray element maps were acquired to distinguish various compositional domains within the texturally constrained monazite grains. The methodology for identifying and retrieving meaningful populations from each compositional domain was adopted from Jercinovic and Williams (2005) and Williams et al. (2006). The monazite dates from each domain were processed and presented graphically using the Isoplot program (version 4.15; Ludwig 2003). The detailed monazite compositions along with spot dates (± 2σ) are given in Table S2.

Bulk rock analysis

Bulk rock compositions of metapelites and quartzofeldspathic gneiss were estimated using inductively coupled plasma atomic emission spectrometry (ICP-AES) at the Department of Earth Sciences, IIT Bombay. A solution was prepared for 0.25 g of the sample by using the fusion method (Sheth et al. 2011), in which the matrix comprises 0.75-g lithium metaborate (LiBO2) and 0.50 g of lithium tetraborate (LiB4O7). Geological Survey of Japan and United States Geological Survey reference materials (JG-2, JG-3, Jb-1b, and JGb-2) were used in the calibration, and rhyolite standard (JR-3) was reanalyzed at regular intervals to monitor the instrument drift. The whole-rock compositions of studied samples are given in Table S3.

Phase equilibria modeling

Pseudosection modeling was performed for the metapelite (KM-24 and KM-29) and quartzofeldspathic gneiss (KM-26C) samples from the Khammam region. The modeling was carried out in the NCKFMASHT system using the bulk composition obtained from ICP-AES analyses (Table S3). Fe2O3 was omitted in phase equilibria modeling due to the following reasons: (a) absence of Fe3+ rich phases (e.g., magnetite) in the equilibrium mineral assemblages, (b) Fe3+ content is negligible in key minerals, and (c) the occurrence of ilmenite in the equilibrium mineral assemblage implies low oxidation conditions (Diener and Powell 2010; Groppo et al. 2010). The consideration of MnO in the phase equilibria calculations shifts the garnet stability field towards lower-pressure conditions (e.g., Tinkham et al. 2001 and references therein). Besides, the effect of MnO on the stability (modal abundance) of other common minerals is negligible. Thus, a similar topology is observed in pseudosections with MnO as compared to those computed without MnO content (White et al. 2004). Therefore, the effect of MnO was ignored in the phase equilibria modeling of metapelites and quartzofeldspathic gneiss. The H2O content was adjusted so that the equilibrium assemblage will be stabilized with appropriate water content (Table S3). The effect of H2O on the modeling was determined by T–MH2O diagrams (Figs. S4a–c and S5b, c, e). Pseudosections were constructed using the Perple_X software program (version 6.8.6; Connolly 2005) with the internally consistent thermodynamic data set file (hp62ver.dat) after Holland and Powell (1998), which was updated in 2011. The activity–composition (a–x) solution models used in this study are garnet, biotite, white mica, cordierite, orthopyroxene and silicic melt (White et al. 2014), K-feldspar (Waldbaum and Thompson 1968), plagioclase (Newton et al. 1980), and spinel (White et al. 2002).

The P–T phase diagrams constructed using pseudosection modeling are typically associated with several notable errors, which limit our interpretation of results determined for complex mineral assemblages (Palin et al. 2016). The choice of bulk composition is one of the important factors in estimating the best quality results from the pseudosection modeling of a heterogeneous sample. Often, the precise estimation of bulk composition is challenging, and most times, it represents a poor analogue of composition in which the equilibrium mineral assemblages were stabilized (Connolly and Petrini 2002). Further, the uncertainties associated with the a–x relationships of end-member minerals were propagated through to the final P–T results, together with bulk-composition-related errors (Holland and Powell 2011 and references therein). In view of the limitations mentioned earlier, bulk compositions acquired using ICP-AES analyses (in this study) were correlated with those calculated using mineral modes from thin sections. The difference between these results is negligible. Besides, appropriate solution models with updated thermodynamic data sets were considered during pseudosection computations.

Results

Sample description, petrography, and mineral compositions

In this section, we describe sample locations, metamorphic textures, and mineral assemblages (Figs. 1, 3, 4, 5) from garnet-bearing metapelites (KM-29 and KM-24) of the KSB and quartzofeldspathic gneiss (KM-26C) of the EGMB. Mineral abbreviations used in this section are after Whitney and Evans (2010).

Fig. 3
figure 3

Detailed petrographic and microtextural relations from metapelites of the KSB. a Garnet porphyroblast in sample KM-29 consists of biotiteI, ilmenite, and rutile inclusions. Note that the matrix foliation (S2S) defined by muscovite and quartz aggregates wraps around garnet porphyroblast. b Large kyanite grains are oriented along with the S2S fabric and contain inclusions of biotite and muscovite. c Kyanite- and muscovite-defined foliation (S2S) wrapping the garnet porphyroblast. Note the bending of kyanite grains (arrow) while wrapping around the garnet. d Two types of biotite grains are identified in the matrix foliation of metapelites. BtII defines the S2S fabric, whereas BtIII replaces muscovite and K-feldspar in the matrix. e Small garnet crystals (~ 500 µm) associated with S2S foliation, and f elongated kyanite crystals oriented parallel to the matrix foliation (S2S) in sample KM-24

Full size image
Fig. 4
figure 4

Petrographic and microtextural relations from quartzofeldspathic gneiss (KM-26C) of the EGMB. a Garnet porphyroblast with inclusions (arrows) of an early assemblage consisting of ilmenite, biotite, and quartz. b Growth of sillimanite (arrows) and biotite along the grain boundaries of garnet porphyroblasts. c Large ribbon-shaped quartz aggregates within the K-feldspar matrix and quartz-defined leucosomes layer (dashed lines). d Thin films of optically continuous K-feldspar filling the grain boundaries (red arrows, enlarged view). Note the cuspate to lobate K-feldspar enclosing anhedral quartz (yellow arrow), indicating partial melting in sample KM-26C

Full size image
Fig. 5
figure 5

Schematic diagram describing the growth of various minerals in relation to the deformation and metamorphic events from the samples a KM-29 and b KM-26C

Full size image

Sample KM-29

Sample KM-29 was collected from an isolated outcrop ~ 20 km southwest of the Kothagudem town (Fig. 1b). The sample contains garnet porphyroblasts within a matrix defined by biotite + kyanite + muscovite + K-feldspar + quartz + ilmenite (Fig. 3a–d). The garnet porphyroblasts (3–6 mm) occur commonly as subhedral grains with inclusions of biotite, muscovite, ilmenite, rutile, and quartz (Fig. 3a). Rutile is only observed as inclusions within the garnet porphyroblasts. The wrapping of muscovite-defined S2S foliation around garnet suggests that these porphyroblasts were formed pre-tectonic to the D2S deformation (Fig. 3a). Inclusion minerals are completely absent in the rim portion of the garnet (Fig. 3a). The garnet exhibits weak core–rim compositional zoning (Prp10–11Alm85–88Grs1.7–2.6Sps2.1–2.6) with an increase of grossular content towards the rim along with slight variations in pyrope, almandine, and spessartine contents (Table S1; Fig. S2a). These minor compositional variations from the core to the rim of garnet porphyroblast suggest long-lasting peak metamorphism at temperature conditions ≥ 600 °C, which significantly homogenizes prograde zoning patterns of the garnet core through intracrystalline diffusion (cf. Spear and Florence 1992; La Roche et al. 2015). This is also evident from the flattened compositional profile of garnet (Fig. S2a) in sample KM-29. The kyanite porphyroblasts usually consist of quartz, muscovite, and biotite inclusions (Fig. 3b), aligned parallel to the S2S fabric, and show slight bending as it wraps the garnet porphyroblast (Fig. 3b, c). All these textures indicate that the growth of kyanite postdates the garnet and syntectonic with the D2S deformation.

Biotite occurs as inclusions within garnet cores (BtI), and as large discrete grains (BtII) interleaved with matrix muscovite and quartz (Fig. 3a, b, d). Both these varieties of biotite crystals exhibit distinct composition. The inclusion-type biotite (BtI) grains are characterized by XFe = 0.58–0.59, TiO2 = 4.0–5.0 wt%, while the matrix biotite (BtII) grains contain slightly higher XFe (0.64–0.65) with lower TiO2 (3.0–4.0 wt%) content (Fig. 3a, b, d; Fig. S3a; Table S1). Locally, fine-grained biotite aggregates replacing matrix minerals and garnet rims (Fig. 3a, d) are designated as retrogressive biotite (BtIII). BiotiteIII has XFe values ranging from 0.63 to 0.64 with TiO2 content of 2.4–2.7 wt% (Fig. S3a; Table S1). Muscovite grains in the matrix are compositionally homogeneous with Na and Si contents variable between 0.06–0.08 apfu (atom per formula unit) and 3.00–3.10 apfu. K-feldspar grains in the matrix have a uniform composition with Xor value ranging from 0.84 to 0.89 (Table S1).

According to the aforementioned petrographic observations, inclusion-bearing garnet core along with inclusion minerals were stable during the early-stage prograde metamorphism, while the inclusion-poor garnet core and syn-D2S kyanite were considered to be part of the peak metamorphic assemblage (Fig. 5a). The prograde (MPR) and peak (MP) metamorphic assemblages are interpreted to be garnet (inclusion-rich core) + biotiteI + muscovite + ilmenite + rutile + quartz and garnet (inclusion-poor core) + kyanite + biotiteII + muscovite + ilmenite + quartz + K-feldspar, respectively (Figs. 3a, 5a). The post-peak retrogressive event (MPP) is characterized by the replacement of garnet, kyanite, and muscovite with biotiteIII along their grain boundaries, and diffusive compositional adjustment (change in the grossular content) of garnet towards the rim (Fig. 3a, b, d, Fig. S2a).

Sample KM-24

Sample KM-24 was collected from a flat-lying exposure located ~ 20 km east of the Khammam town (Fig. 1b). This sample consists mainly of garnet, biotite, kyanite, muscovite, plagioclase, quartz, and ilmenite (Fig. 3e–f). The growth of fine-grained (< 1 mm), subhedral, and inclusion-free garnet porphyroblast within the matrix foliation dominated by biotite and quartz suggests that the garnet has grown syntectonic with the S2S foliation (Fig. 3e). The garnet (Prp8–11Alm82–84Grs3.5–3.9Sps3.5–4.2) is enriched in almandine content with a low concentration of pyrope, grossular, and spessartine and exhibits uniform composition throughout the grain. Kyanite is mostly oriented along the S2S foliation (Fig. 3f), which advocates the syn-S2S growth of kyanite crystals. Biotite in the matrix has a uniform composition with XFe of 0.61–0.64 and TiO2 of 1.9–2.9 wt%. The Na content in the matrix muscovite (0.10–0.19 apfu) is slightly higher than muscovite from sample KM-29 (Fig. S3b), while plagioclase grains are sodic with Xab of 0.80–0.84 (Table S1). Based on the textural relations, the sample comprises the peak metamorphic assemblage (MP) of garnet + kyanite + biotite + muscovite + ilmenite + plagioclase + quartz (Fig. 3e–f).

Sample KM-26C

Sample KM-26C was collected from an isolated outcrop located in the easternmost part of the Khammam region (near Kallur village; Fig. 1b). This sample is characterized by subidioblastic garnet porphyroblasts within the matrix composed of K-feldspar, quartz, and biotite (Fig. 4a–d). The garnet porphyroblasts are chemically homogeneous (Prp14–16Alm80–81Grs3.5–3.9Sps0.5–0.8) and often contain inclusions of biotite, quartz, ilmenite, and K-feldspar (Fig. 4a). Biotite grains as inclusions (BtI) in garnet and those in the matrix (BtII) have a similar composition with XFe between 0.61 and 0.65. Matrix K-feldspar has higher Xor content (0.83–0.90) than K-feldspar composition (Xor = 0.71–0.80) in the leucosome portion (Table S1). Sillimanite occurs as acicular crystals, associated with the garnet along its grain boundaries (Fig. 4b). The leucosomes (Fig. 2e), represent the migmatized portion of sample KM-26C, are characterized by K-feldspar and quartz minerals with cuspate to lobate outlines (Fig. 4c, d).

Based on the above observations, the inclusion minerals along with the garnet (MPR: garnet (inclusion-rich core) + biotiteI + ilmenite ± K-feldspar) constrain the early prograde event in the sample KM-26C (Fig. 5b). The partial melting in quartzofeldspathic gneiss (Fig. 4a, b) marks the peak metamorphism (MP: garnet (inclusion-poor core) + K-feldspar + biotiteII + quartz + melt), which might have homogenized the prograde garnet composition due to high-grade metamorphism (Fig. S2c). The post-peak metamorphism (MPP) is marked by sillimanite growth along the subhedral margins of garnet, and minor compositional readjustments in biotiteII and garnet rim compositions (Fig. 4b and Fig. S2c).

Conventional thermobarometry

The equilibrium P–T conditions of various metamorphic episodes in metapelites (KM-24 and KM-29) and quartzofeldspathic gneiss (KM-26C) were determined using the application of mineral thermobarometry (Table 1). The temperature conditions were estimated using the garnet–biotite exchange thermometer (Ferry and Spear 1978; Bhattacharya et al. 1992; Holdaway 2000) and Ti-in-biotite thermometer (Henry et al. 2005). The pressure conditions were calculated using the garnet–biotite–plagioclase–quartz (GBPQ; Hoisch 1990), garnet–biotite–aluminum silicate–quartz (GBAQ; Wu 2017), and garnet–aluminum silicate–plagioclase–quartz (GASP; Holdaway 2001) barometers.

Table 1 Summary of conventional thermobarometry results obtained for metapelites (KM-24 and 29) and quartzofeldspathic gneisses (KM-26C) in the Khammam region
Full size table

Based on the aforementioned textural relations, it is evident that the prograde garnet compositions (and inclusion minerals) in metapelites and quartzofeldspathic gneiss were homogenized due to intracrystalline diffusion induced during prolonged high-temperature peak metamorphism (cf. La Roche et al. 2015). Therefore, it is difficult to constrain the prograde metamorphic conditions using the compositions of garnet core and entrapped biotite inclusion. We have paired garnet core (distant from inclusions) and matrix biotite compositions, and garnet rim and biotite in contact with garnet compositions to determine the peak and post-peak P–T conditions, respectively. In sample KM-29, the mineral pairs represented by garnet core (inclusion-poor)–matrix biotite (BtII) and GBAQ assemblages recorded P–T conditions of 617– 657 °C at 7.5 kbar (Table 1). These P–T conditions represent the peak metamorphic episode (MP) for sample KM-29. The P–T conditions for post-peak metamorphism (MPP) were retrieved using garnet rim-biotite (BtIII) mineral pairs and GBAQ matrix mineral assemblages at 566–613 °C and 7.2 kbar. Similarly, the garnet and matrix minerals (biotite, kyanite, plagioclase, and quartz) yielded peak P–T conditions (MP) of 612–639 °C at 6.0–7.2 kbar for sample KM-24 (Table 1). The calculations for garnet core (inclusion-poor)–matrix biotite (BtII) pairs in sample KM-26C yielded peak (MP) temperature conditions of 720–836 °C. The peak pressure conditions could not be determined due to the lack of suitable minerals assemblages. The P–T estimates using garnet rim–biotite (BtII) in contact with garnet and GBAQ mineral assemblages yielded slightly lower temperature conditions of 716–820 °C at ~ 6.9 kbar (Table 1). These P–T values represent near-peak re-equilibrium conditions that stabilized sillimanite in the matrix assemblage.

P–T pseudosection modeling

Sample KM-29 (garnet–kyanite–K-feldspar schist)

The peak mineral assemblage (MP) preserved in sample KM-29 comprises garnet core (inclusion-poor) + muscovite + biotiteII + K-feldspar + kyanite + quartz + ilmenite. In excess H2O fluid conditions, the peak mineral assemblage is not stabilized in the P–T frame of NCKFMASHT pseudosection. To determine the actual H2O content, a T–MH2O diagram was computed at 7.5 kbar, which is in the range of P–T estimates of conventional thermobarometry (7.2–7.5 kbar; Table 1). The H2O content was considered between anhydrous (MH2O = 0) conditions and compositions with a maximum amount of H2O (MH2O = 1) that is represented by the loss-on-ignition data (~ 7 mol%). In T–MH2O diagram, the peak assemblage is stable in the temperature conditions of ~ 480–720 °C at MH2O values of 0.10–0.65 (Fig. S5b). At MH2O > 0.65, kyanite is not stable in the predicted assemblage and stabilizes mineral assemblages with excess H2O. Plagioclase is stable at lower temperatures (< 480 °C), which is not a part of the peak mineral assemblage. To estimate appropriate MH2O content, a series of pseudosections were constructed at MH2O = 0.27, 0.40, and 0.60 (Fig. S4a–c). Finally, we have considered the MH2O value of 0.55 (3.64 mol%; Fig. S5b) in the pseudosection modeling, which gives the best approximation of calculated modal abundances of garnet (13–15 vol%), kyanite (2–3 vol%) and muscovite (22–24 vol%) as compared to observed values (Fig. S5a–b).

The P–T pseudosection for sample KM-29 was constructed in a P–T range of 4–9 kbar and 500–800 °C (Fig. 6a). The field-3 with garnet + muscovite + biotite + K-feldspar + kyanite + quartz + ilmenite marks the observed peak metamorphic assemblage (MP), which stabilizes in a wide P–T window (6.0 to > 9.0 kbar and 500–700 °C; Fig. 6a). The intersection of Xalm (Fe/[Fe + Mg + Ca]) = 0.87, Xprp (Mg/[Fe + Mg + Ca]) = 0.12 and Xgrs (Ca/[Fe + Mg + Ca]) = 0.018 isopleths of garnet, which represent inclusion-poor garnet core composition, constrains the peak (MP) metamorphic conditions at 7.4–7.8 kbar and 600–640 °C. The intersection of garnet rim isopleths (Xalm = 0.89, Xprp = 0.10 and Xgrs = 0.024) retrieves P–T conditions for the post-peak adjustments (MPP) at 6.1–6.6 kbar and 590–625 °C (Fig. 6a). These P–T estimates are in good agreement with the peak (617–657 °C at 7.5 kbar) and post-peak (566–613 °C and 6.0–7.2 kbar) metamorphic conditions calculated using the application of conventional thermobarometry (Table 1). On combining the P–T conditions of peak (MP) and post-peak (MPP) metamorphic episodes, a small decompression path (ΔP ~ 1.4 kbar; calculated from average values) can be observed for the sample KM-29 (Fig. 6a). Although the pressure difference in this P–T path is closer to the typical uncertainty on pressure estimates (i.e., 1 kbar) in pseudosection calculations, it should be noted that the grossular content increases (1.8–2.4 mol%) towards the garnet rim. By considering these factors, the decompression path (MP–MPP =  ~ 1.3 kbar; Fig. 6a) constrained for sample KM-29 appears to be valid and formed as a result of post-peak adjustments of garnet in response to decompression.

Fig. 6
figure 6

Perple_X-generated phase diagrams constructed in the NCKFMASHT system for samples a KM-29 and b KM-24. Bulk compositions (mol%) used in pseudosection calculations are given in the top parts of pseudosections. a A P–T pseudosection constructed for sample KM-29, modeled with compositional isopleths of garnet. The arrow indicates the inferred P–T trajectory from the core to the rim portion of the garnet (MP to MPP). b The modeled P–T pseudosection for KM-24 contoured with the garnet compositional isopleths. The thick red line in both pseudosections marks the solidus, while the green (in a) and blue (in a and b) thick lines demarcate the limits of plagioclase and garnet-bearing assemblages, respectively. The thick orange lines indicate the stability fields of andalusite, kyanite, and sillimanite. Yellow shades represent mineral isopleth intersections for the peak (MP) and post-peak (MPP) metamorphic assemblages

Full size image

Sample KM-24 (garnet–kyanite–plagioclase schist)

A T–MH2O diagram was constructed at 7.0 kbar to determine the appropriate H2O content for the pseudosection modeling of sample KM-24. In T–MH2O diagram, the peak assemblage (garnet + kyanite + plagioclase + muscovite + biotite + quartz + ilmenite) is stable at < 670 °C with MH2O values between 0.75–0.85 (Fig. S5c). At MH2O > 0.85, kyanite is not stable in the predicted assemblage, and equilibrate mineral assemblages with excess H2O. K-feldspar is stable at lower MH2O values (< 0.65), which is not a part of the peak mineral assemblage. Based on these constraints, we have considered an average MH2O value of 0.80 (4.16 mol%; Fig. S5c) for the pseudosection modeling of sample KM-24. The pseudosection constructed for sample KM-24 shows the stabilization of peak mineral assemblage (MP: garnet + kyanite + plagioclase + muscovite + biotite + quartz + ilmenite) for a wide range of P–T conditions at 5.0–7.8 kbar and 500–670 °C (field 2; Fig. 6b). The garnet porphyroblasts are compositionally homogeneous and not diagnostic in estimating the prograde evolution (Fig. S2b). The intersection of Xalm = 0.85, Xprp = 0.11 and Xgrs = 0.32 isopleths of unzoned garnet constrain the P–T conditions of peak metamorphism (MP) at 6.6–6.8 kbar and 580–610 °C (Fig. 6b). The estimated peak metamorphic conditions are consistent with the results of conventional thermobarometry (612–639 °C at 6.0–7.0 kbar; Table 1), and also similar to those constrained for the sample KM-29 (Fig. 6a).

Sample KM-26C (quartzofeldspathic gneiss)

The observed peak metamorphic assemblage (MP: garnet core (inclusion-poor) + K-feldspar + quartz + ilmenite + biotiteII + melt) is stable over a wide range of H2O (Fig. S5e). At 7 kbar, which represents the minimum pressure estimated using the conventional thermobarometry approach, the peak assemblage is stable over a temperature range of 750–910 °C with an MH2O value of > 0.15 (Fig. S5e). An MH2O value of 0.69 (2.55 mol%) is considered as appropriate water content to stabilize the peak assemblage (Fig. S5e). At this MH2O value (0.69), the observed modal abundances of garnet (20–23 vol%) and K-feldspar (40–45 vol%) are comparable with the measured mineral modes (Fig. S5d–e).

Since the sample KM-26C is migmatized, the bulk composition obtained using ICP-AES analysis represents the residual bulk composition after melt migration (or loss). This residual composition can only constrain the peak and post-peak conditions of the rock (Korhonen et al. 2013). The P–T pseudosection for the sample KM-26C (gneiss) was modeled in a P–T range of 4–10 kbar and 600–1000 °C using the residual bulk composition (Fig. 7a). The peak mineral assemblage (field 14) is stable at variable pressures and 750–850 °C (Fig. 7a). The intersection of garnet isopleths (Xalm vol% = 0.803, Xprp = 0.160, and Xgrs = 0.039) yielded the peak metamorphic conditions at 790–810 °C and 7.9–8.1 kbar (Fig. 7a). The sillimanite-bearing post-peak assemblage is represented by the field with garnet + K-feldspar + sillimanite + quartz + ilmenite + biotite (field 7) assemblage, which is stable at P–T conditions of ~ 7 kbar and 740–750 °C (Fig. 7a).

Fig. 7
figure 7

a The P–T pseudosection for KM-26C was computed using the residual bulk composition, and the intersection of garnet isopleths constrains the peak metamorphic conditions (MP; yellow triangle). The post-peak P–T path is drawn to the sillimanite-bearing field (indicated by dotted arrow MP–Mpp). b The pseudosection was derived after two-stage melt reintegration (12 vol%) procedure (see text; Fig. S6). The pseudosection is contoured with vol% of garnet, K-feldspar, and biotite to show the temperature dependency on their mineral modes. The thick red line in the pseudosection represents the solidus curve. The thick blue in b and orange line (a and b) demarcates the stability of garnet and sillimanite-bearing assemblages, respectively

Full size image

The prograde evolution of a migmatized rock can be determined by recalculating the plausible protolith composition, which involves reintegrating melt lost from the system with the residual bulk composition (White et al. 2004: Diener et al. 2008). The experimental evidence of Rosenberg and Handy (2005) suggests that melt migration starts when the partial melting exceeds 7 vol% and some amount of melt may be retained in the grain boundaries, which can be approximated to ~ 1 vol% (Sawyer, 2001), resulting in the net loss of 6 vol% from the system in a single prograde melting event. For sample KM-26C, the down temperature melt reintegration procedure (Korhonen et al. 2013) was applied to reconstruct plausible protolith bulk composition. The first stage melt reintegration (by adding ~ 6 vol%, melt) was performed in the field of peak assemblage garnet + K-feldspar + quartz + ilmenite + biotite + melt (field 14: Fig. 7a) at 800 °C and 8.0 kbar and the new bulk composition obtained after reintegration was used to compute a new pseudosection (Fig. S6a). Similarly, two-stage melt reintegration steps were performed by adding a total of 12 vol% of melt to the system to obtain the plausible protolith bulk composition with H2O-saturated solidus (Figs. 7b and S6b). The pseudosection computed after reintegrating the melt with 12 vol% stabilized H2O-saturated solidus at 660–680 °C with the plausible prograde assemblage of garnet + K-feldspar + quartz + ilmenite + biotite ± muscovite ± plagioclase (Fig. 7b). Although muscovite and plagioclase are not observed in the sample, these minerals might have consumed on account of prograde heating and onset of the melting process (Fig. S6c).

The contoured modal isopleths of garnet, K-feldspar, and biotite in the pseudosection are steep and temperature dependent (Fig. 7b). Petrographic observations suggest that prograde garnet growth in the leucosome was contemporaneous with the partial melting and the consumption of inclusion phases like K-feldspar, biotite (± muscovite), and ilmenite (Figs. 4a and S6c). Therefore, a near-isobaric prograde P–T path (MPR–MP: ~ ∆T 100 °C) is inferred from the field of plausible prograde assemblage (field 10) to the peak P–T conditions (field 15) by considering the modal variations of garnet, K-feldspar and biotite (Figs. 7b). However, the prograde arm of this P–T path is not fully reliable due to the high uncertainty associated with pressure estimates of the prograde assemblage. The P–T estimates from peak to post-peak assemblage (MP–MPP) define a decompression path, which was consistent with the growth of fine-grained sillimanite and biotite in the matrix adjacent to garnet porphyroblast (Figs. 4b, 7a). These estimates infer a possible clockwise P–T path for the quartzofeldspathic gneiss, which involves a prograde melting (represented by dashed lines) event, followed by a decompression event on account of melt crystallization (Fig. 7a, b).

Monazite geochronology

In situ U–Th–(total) Pb monazite dating was performed on two samples, including a metapelite (KM-29) from the KSB and a quartzofeldspathic gneiss (KM-26C) from the EGMB. X-ray element maps (Th, U, and Pb) were combined with BSE images of monazites to identify discrete compositional domains (CD) (cf. Jercinovic and Williams 2005; Williams et al. 2006). These compositional domains are labeled as CDS-I, II, and III for sample KM-29 and CDG-I, II, III, IIIA, IV, and V for sample KM-26C (Figs. 8 and 9). The textural setting and compositional variation of analyzed monazite grains in these samples are summarized in Figs. 8, 9, and S7. The ThO2 vs. Age, [Sm2O3 + Gd2O3] vs.[ThO2], and Th/U vs. Age plots for various compositional domains in samples KM-29 and KM-26C are shown in Fig. S8.

Fig. 8
figure 8

a–e Textural occurrence and compositional variation of monazite grains in the metapelite (KM-29) of the KSB. f Probability density plot of monazite age data showing three distinct peak ages (1215 ± 13 Ma, 1137 ± 18 Ma, and 849 ± 06 Ma) for various compositional domains (CDS-I, CDS-II, and CDS-III) identified in the monazite grains. The spot ages given in the right panel are in ‘Ga’

Full size image
Fig. 9
figure 9

a–d Textural occurrence and compositional variation of monazite grains in the quartzofeldspathic gneiss (KM-26C) of the EGMB. Note that these monazites are identified with five distinct compositional domains. e Probability density distribution of monazite age data showing five different age peaks (2689 ± 35 Ma, 2494 ± 14 Ma, 1621 ± 14 Ma, 1581 ± 15 Ma; 1438 ± 25 Ma and 869 ± 17 Ma) from corresponding compositional domains (CDG-I; CDG-II; CDG-III; CDG-IIIA, CDG-IV, and CDG-V). The spot ages given in the right panel are in ‘Ga’

Full size image

KSB metapelite (KM-29)

Monazite in sample KM-29 mostly occurs in association with the matrix minerals, aligns parallel to the S2S and S3S matrix foliations, and is rarely observed as an inclusion within the garnet porphyroblasts (Figs. 8a–e and S7a–e). The matrix monazite is generally grown as elongated grains (Figs. 8b–e and S7a), whereas inclusion-type monazite is anhedral to subhedral shape with sub-rounded edges (Figs. 8a and S7b–c). Th and Y monazite zoning patterns reveal three distinct compositional domains, CDS-I, II, and III. The CDS-I is represented by high-Th and low-Y (ThO2 = 7–11%, UO2 = 0.0–0.4%, Y2O3 = 0.8–1.2%) monazite cores (Fig. 8a), which are overgrown by relatively low-Th and comparable-Y CDS-II (ThO2 = 4–7%, UO2 = 0.2–0.9%, Y2O3 = 1.0–1.1%), and low-Th and high-Y CDS-III (ThO2 = 5–6%, UO2 = 0.9–1.9%, Y2O3 = 0.3–3.1%) domains (Fig. 8a–c). The integration of spot ages from chemically distinct CDS-I and CDS-II domains have yielded weighted mean ages of 1215 ± 13 Ma (MSWD = 0.85; n = 11) and 1137 ± 18 Ma (MSWD = 1.0; n = 7), respectively (Fig. 8f). Analyses from the monazite outer rims (CDS-III) yielded a weighted mean age of 849 ± 06 Ma (MSWD = 1.3; n = 39).

EGMB quartzofeldspathic gneiss (KM-26C)

The subhedral to anhedral monazite grains from sample KM-26C are commonly hosted within garnet porphyroblasts, and K-feldspar, plagioclase, and quartz within the matrix (Fig. 9a–d). X-ray elemental maps of monazite grains can distinguish five discrete compositional domains, CDG-I, II, III, IIIA, IV, and V in sample KM-26C (Fig. 9a–d). The CDG-I domain is commonly preserved in the core of monazite-1 and is characterized by low-Th, -U, and -Pb domain (ThO2 = 4–6%, UO2 = 0.10–0.19%, PbO = 0.57–0.81%; Fig. 9a). This domain is overgrown by relatively high-Th, -U and -Pb domain (CDG-II: ThO2 = 8–13%, UO2 = 0.25–0.50%, PbO = 1.0–1.7%). The monazite spot ages from CDG-I yielded the oldest ages with a weighted mean age of 2689 ± 35 Ma (MSWD = 1.3; n = 11). The CDG-II domain yielded a weighted mean age of 2494 ± 14 Ma (MSWD = 1.1; n = 24) (Fig. 9e). The CDG-III with relatively high-Th and -U, and low-Pb (ThO2 = 7–11%, UO2 = 0.01–0.37%, PbO = 0.55–0.85%) is observed in the core portion of Mnz-3 (Fig. 9c). The CDG-IIIA domain is characterized by relatively low-Th, U and Pb domain (ThO2 = 4–6%, UO2 = 0.01–0.40%, PbO = 0.3–0.5%, Y2O3 = 0.0–0.4%) preserved in the Mnz-2, Mnz-4 and also as outer rim overgrowing CDG-II in Mnz-1 (Fig. 9a, b, d). Despite the variation in chemical compositions, the CDG-III and CDG-IIIA domains have shown a considerable overlap in their ages with weighted mean ages of 1621 ± 14 Ma (MSWD = 0.87; n = 13) and 1581 ± 15 Ma (MSWD = 0.79; n = 21), respectively (Fig. 9e). Further, these domains are overgrown by CDG-IV (ThO2 = 1–3%, UO2 = 0.1–0.5%, PbO = 0.20–0.28%, Y2O3 = 0.0–0.2%) and CDG-V (ThO2 = 2–6%, UO2 = 0.15–0.30%, PbO = 0.10–0.23%, Y2O3 = 0.0–0.2%) compositional domains (Fig. 9c). The CDG-V is mostly preserved as thin rims over pre-existing chemical domains (Fig. 9b) and demarcates the final-stage monazite growth in the gneisses. Spot analyses provide weighted mean ages of 1438 ± 25 (MSWD = 0.73; n = 11) and 869 ± 17 Ma (MSWD = 1.15; n = 09) for the CDG-IV and CDG-V domains, respectively (Fig. 9e).

Discussion

Metamorphic P–T history and interpretation of monazite ages

The P–T histories and monazite age data described from the KSB metapelites and EGMB quartzofeldspathic gneisses are hitherto unreported from the Khammam region (Figs. 1 and S1), barring those documented by Hari Prasad et al. (2000). The results acquired from petrological observations indicate that the KSB metapelites have recorded three stages (MPR, MP, and MPP) of metamorphism. The garnet zoning patterns suggest that the inner core was completely homogenized due to the long-lived peak metamorphism (Fig. S2a, b). Therefore, the inner core compositions are not reliable for estimating the prograde P–T conditions (La Roche et al. 2015). The prograde metamorphic stage probably initiated outside the P–T field of garnet and continued to stable after crossing the ‘garnet-in’ curve. Based on the wrapping S2S foliation around garnet, the prograde metamorphism (MPR) should have occurred inter-tectonic between D1S and D2S deformation events. The peak metamorphism (MP), accompanied by syn-D2S garnet (inclusion-poor core) and kyanite growth, is equilibrated at the P–T conditions of 7.4–7.8 kbar and 600–640 °C (Fig. 6a). The subsequent post-peak metamorphism (MPP) at 6.1–6.6 kbar and 590–625 °C resulted in the diffusional adjustments of garnet rim and growth of biotiteIII (Fig. 6a). The slight fall of pressure and temperature conditions from MP to MPP suggests a decompression path, which could have caused the exhumation of metapelites from ~ 26 km to ~ 22 km depth (Fig. 10c). The application of the melt reintegration procedure for quartzofeldspathic gneiss KM-26C constrains a near-isobaric prograde path (~ ∆T 100 °C; MPR–MP; Figs. 7b and S6a, b). The subsequent partial melting, which resulted in the high-grade peak metamorphism (MP) at 7.9–8.1 kbar/790–810 °C (Fig. 7a), is contemporaneous with the D2G deformation (Fig. 2e). We assume that the increase of temperature from MPR to MP was caused by magmatic underplating, which occurred during the formation of the Ongole Domain of the EGMB (Sarkar and Schenk 2014; Sarkar et al. 2015). The formation of fine-grained sillimanite and biotite, around garnet (Fig. 4b), could be due to the post-peak readjustments (MPP) at ~ 7 kbar/740–750 °C (Fig. 7a).

Fig. 10
figure 10

Schematic diagram summarizes the various tectonothermal episodes and corresponding monazite ages identified in the a metapelite (KM-29) and b quartzofeldspathic gneiss (KM-26C). The top panels of these figures characterize compositional variation within different monazite age domains, while the bottom panels highlight tectonic inferences suggested for each age population. c A P–T diagram summarizes various tectonothermal episodes and corresponding monazite ages for the samples KM-24, 29, and KM-26C in the Khammam region

Full size image

In metapelites, the Y budget of the monazite is mainly governed by the crystallization of garnet during sub-solidus metamorphism (Spear and Pyle 2010). Therefore, in the absence of other Y-bearing phases like xenotime, the Y partitioning between garnet and monazite can be used as a proxy to understand the relative timing of garnet growth and resorption (Mahan et al. 2006; Dumond et al. 2015; Rahimi and Massonne 2018). Monazites in KSB metapelites exhibit distinct chemical domains that have grown during different stages of metamorphism and deformation (Figs. 8 and S7). U–Th–(total) Pb dating of monazite cores with low-Y contents yielded two distinct age populations at 1253–1196 Ma (CDS-I) and 1178–1102 Ma (CDS-II) (Fig. 8a, b). The relatively older CDS-I is exclusively seen as inclusion within the pre-D2S garnet, whereas (Figs. 8a, and S7b, c) CDS-II is noted as monazite overgrowth around CDS-I, occupying core region in the matrix, and as discrete grains aligned parallel to the S2S foliation (Figs. 8a, b, and S7d). In sample KM-29, xenotime is not stable in the equilibrium mineral assemblage, and therefore, Y preferentially partitions into garnet than monazite. Hence, the growth of the Y-poor core and rim (CDS-I and II) of monazite could have formed simultaneously during the development of garnet porphyroblast. Based on petrographic textures (Figs. 3a and S7a), it is evident that the garnet growth initiated pre-tectonic to S2S foliation. The above observations constrain the minimum timing for the prograde metamorphism (MPR) at 1215 ± 13 Ma, where garnet growth was initiated during late-D1S to early-D2S deformation (Figs. 5a and 10a). The timing of peak metamorphism (MP) is constrained at 1137 ± 18 Ma (Fig. 10a), which was consistent with the growth of CDS-II monazite age domain and syn-D2S kyanite (Fig. S7d). The much younger age at 849 ± 6 Ma, obtained from high-Y monazite rims (CDS-III) and discrete matrix monazite grains overgrowing and crosscutting the S2S foliation, at places aligned parallel to the S3S foliation, represents the possible timing of the D3S deformation (Figs. 8d, e, 10a, and S7e).

In quartzofeldspathic gneiss (KM-26C) of the EGMB, monazites hosted in the garnet show two distinct resorbed domains, CDG-I and CDG-II, which yielded weighted mean ages of 2689 ± 35 Ma and 2494 ± 14 Ma, respectively (Fig. 9a, d). These Neoarchean ages perhaps demonstrate the two separate reworking events that affected the pre-existing Archean crust (> 2700 Ma). The overgrowth of high-Th and low-Y domains (CDG-III) with relatively high Th/U ratio (48–220) on highly resorbed monazite cores (CDG-II) indicates the relative timing of partial melting in the gneiss. Since the Th and U preferentially fractionate into the melt, the monazite growing from this melt preferentially incorporates Th into its crystal structure (Xing et al. 2013), while U is retained in the melt (Keppler and Wyllie 1990), which is consistent with high Th/U ratio in domain CDG-III (Fig. S8c). This CDG-III domain is also characterized by relatively low-Sm and Gd (Sm2O3 = 1.2–2.0% and Gd2O3 = 0.3–0.50%) compared to younger compositional domains (CDG-IV and CDG-V; Fig. S8d). This low-Sm and Gd is attributed to the growth of garnet that preferentially sequesters these elements during high-grade metamorphism (cf. Bea and Montero 1999). The CDG-III compositional domains consistently yield ages between 1659–1575 Ma and characterize the timing of the partial melting and garnet growth. Thus, the timing of initial partial melting and garnet growth can be constrained at 1621 ± 14 Ma (MSWD = 0.87; n = 13). The relatively younger ages of 1631–1533 Ma (weighted average: 1581 ± 15 Ma; MSWD = 0.79; n = 21) were acquired from Th- and U-depleted domains (CDG-IIIA), usually seen as overgrowing CDG-II and as independent grains within the matrix (Fig. 9a, b, d). This domain is characterized by a relatively high concentration of Sm and Gd (Fig. S8d), suggesting its growth from relatively Th-poor residual melt after forming the CDG-III domain. At this stage, garnet growth might have ceased, and the monazite formed from residual melt enriched in Sm and Gd (Sm2O3 = 2–3% and Gd2O3 = 0.50–1.54%) as compared to CDG-III domain (Figs. 10b and S8d). Based on the monazite compositional variation and garnet growth, we interpret that transition from the prograde to the peak metamorphism (MPR to MP) in quartzofeldspathic gneiss would have occurred between 1621 ± 14 Ma and 1581 ± 15 Ma (Fig. 10c).

The significance of 1479–1377 Ma (CDG-IV) age population is not clear, as these domains do not show any relation with other monazite domains and tectonic fabrics (Fig. 9c). However, the higher abundance of Sm and Gd in the CDG-IV domain (Sm2O3 = 3–4% and Gd2O3 = 0.60–3.27%; Fig. S8d) indicating the breakdown of garnet and releasing the MREE’s and HREE’s, leaving monazite domains enriched in these elements. Therefore, we speculate that 1438 ± 25 Ma age (MSWD = 0.73; n = 11) obtained from the CDG-IV domain represents the timing of post-peak re-equilibration and the stabilization of sillimanite needles in the matrix (Fig. 10b). The domain CDG-V with relatively low Sm + Gd (Sm2O3 = 2–4% and Gd2O3 = 0.6–1.57%) is seen overgrowing CDG-III, IV (Fig. 9b). The Neoproterozoic ages from this domain 903–819 Ma are similar to those ages recorded from the pelites of KSB, which may correspond to a localized D3G shearing event (Figs. 2f, 10b).

Geodynamic implications

The results obtained in this study, together with previous data from the Ongole and Vinjamuru Domains (Gupta et al. 1984; Dobmeier et al. 2006; Dharma Rao and Reddy 2009; Ravikant et al. 2010; Vijaya Kumar et al. 2010; Dharma Rao et al. 2011a; Chatterjee et al. 2016; Sarkar and Schenk 2016; Hrushikesh et al. 2019), and Napier–Rayner Complex of East Antarctica (Grew and Manton 1979; Clarke 1988; Grew et al. 2001; Harley 2003; Halpin et al. 2005, 2007; Harley et al. 2013), suggest the following regional and geodynamic implications (Fig. S9). The oldest Archean dates of 2785–2445 Ma and Mesoproterozoic high-grade metamorphic events (1659–1575 Ma and 1631–1533 Ma) are recorded from the EGMB quartzofeldspathic gneisses (Fig. 9), which are similar to those documented in the Ongole Domain (~ 2600–2400 Ma and ~ 1650–1550 Ma; Simmat and Raith 2008; Henderson et al. 2014). This suggests the possible westward extension of the Ongole domain. Henderson et al. (2014) performed detrital zircon analysis from the Ongole Domain and demonstrated that the Neoarchean ages were possibly derived from the Napier complex of East Antarctica, which indicates the timing of UHT metamorphism in the Oygarden Group (Clarke 1988; Asami et al. 2002). These authors also postulated that the formation of the Ongole Domain occurred in the near vicinity of the Napier Complex during Late Paleoproterozoic to Early Mesoproterozoic (Henderson et al. 2014). This inference is based on the sporadically preserved similar ages in the Rayner–Napier complex on account of felsic magmatism and UHT metamorphism (Kelly et al. 2002; Halpin et al. 2005, 2013; Horie et al. 2016). Therefore, our new data from the Khammam region support that the Neoarchean components in the Ongole Domain were possibly derived from the Rayner–Napier Complex of East Antarctica.

Further, the high-T metamorphic events recorded in the quartzofeldspathic gneiss (1659–1575 Ma and 1631–1533 Ma; Fig. 9) are temporally consistent with partial melting (1620–1600 Ma) and crustal reworking (1550–1540 Ma) events identified in the Ongole Domain (cf. Sarkar et al. 2014, Sarkar et al. 2015). According to Sarkar et al. (2015), the heat required for the high-T and UHT metamorphism would have been derived from the magmatic underplating beneath the Ongole magmatic arc in a subduction zone setting. The 1479–1377 Ma tectonic overprint in the quartzofeldspathic gneiss may be related to the terminal rifting of the Columbia supercontinent that led to the dispersal of continental blocks with simultaneous alkaline magmatism in southeastern India and East Antarctica (Fig. S9; Dobmeier and Raith 2003; Simmat and Raith 2008; Upadhyay 2008; Dharma Rao et al. 2011a; Sarkar et al. 2015; Subramanyam et al. 2016; Sain et al. 2017). However, due to the absence of alkaline magmatism within the Khammam region, the significance of the Late Mesoproterozoic (1438 ± 25 Ma) event is unclear.

In contrast, KSB metapelites experienced Late Mesoproterozoic metamorphic episodes at 1253–1196 Ma and 1178–1102 Ma (Fig. 10). The tectonic scenario of these metamorphic events can be best explained as a result of emplacement of the Chimalpahad Anorthosite Complex (Fig. 1b; Sm–Nd whole-rock age: 1300–1100 Ma; Dharma Rao et al. 2011b) that favored an increase in temperature. Such evidence of arc magmatism and associated metamorphism indicates prolonged subduction in the Khammam region. Further, it has been interpreted that the tectonothermal events between 1253 and 1102 Ma are due to protracted accretion between SE India and East Antarctica during the formation of the Rodinia supercontinent (Fig. S9). This conclusion is based on the development of compressional fabrics and high-T metamorphism in the KSB metapelites (Figs. 10 and S9). On integrating published age data from the Vinjamuru domain metapelites (1650–1550 Ma, 1500–1400 Ma, and 1380–1250 Ma; Hrushikesh et al. 2019), it becomes evident that the KSB metapelites have experienced distinct metamorphic history (Figs. 10 and S9). The disparity in the tectonothermal histories reveals that the KSB, Vinjamuru Domain, and Ongole Domain evolved distinctly during Late Paleoproterozoic to Late Mesoproterozoic (Fig. S9).

The youngest metamorphic event (903–813 Ma; this study) retrieved from the monazite rims not only attributed to the timing of N–S directed shearing (this study; Figs. 2f, 10 and S9) but also demarcate the shearing event in the adjacent Ongole and Vinjamuru domains (Sarkar et al. 2014; Hrushikesh et al. 2019). On a regional scale, this Neoproterozoic shearing event appears contemporary with the 870–780 Ma sinistral shearing and anatexis in the Eastern Indian Tectonic Zone (EITZ) that formed during the southward movement of India after its collision with East Antarctica at ~ 1000 Ma in the Rodinia assembly (Chatterjee et al. 2010). Therefore, we assume that the shearing event related to the formation of EITZ extends westward into the Ongole Domain and NKSB (Figs. 10 and S9). Nevertheless, such a presumption requires a detailed petrochronological analysis to determine the effect of 870–780 Ma anatexis in various crustal terranes in southeastern India.

Conclusions

Pressure–temperature–time estimates, described in this study using an integrated approach involving microtextural relationships, conventional thermobarometry, pseudosection modeling, and monazite dating, provide deeper insights into the polycyclic tectonothermal evolution of southeastern India. The studied EGMB quartzofeldspathic gneisses in the Khammam region have experienced high-T metamorphism with the peak P–T conditions of 7.9–8.1 kbar and 790–810 °C. In contrast, the garnet–kyanite schists investigated from the KSB witnessed MP–HT conditions of 7.4–7.8 kbar and 600–640 °C. The high-T metamorphism of quartzofeldspathic gneisses (EGMB component) is explained by the partial melting beneath the crust during the formation of the Ongole Domain at ca. 1659–1533 Ma. Whereas, the MP–HT metamorphism in KSB pelites (1178–1102 Ma) was related to the collision between Indian and East Antarctica in Rodinia assembly. We suggest that the Neoarchean components retrieved from the quartzofeldspathic gneisses were probably derived from the Napier Complex of East Antarctica. All the lithologies in the Khammam region witnessed a Neoproterozoic shearing event (903–813 Ma), which is possibly contemporaneous with the formation of EITZ and might have developed as a result of post-collisional (c. 1.0 Ga) adjustments between India and East Antarctica after the Rodinia assembly. Based on the absence of this shearing event in EDC rocks, we suggest that the effect of this shearing is terminated within the KSB. Therefore, the Khammam region with protracted tectonothermal history (2785–2445 Ma and 1659–813 Ma) is an analogue for a hot collision zone, which evolved as a composite terrane during meso- to neoproterozoic time.