2.a. Southern Madurai Block

The SMB consists of charnockites (massive/foliated), anatectic garnet-biotite gneisses, garnet-biotite-K-feldspar ± sillimanite (khondalite) gneisses, hornblende-biotite gneisses and quartzites as the major lithological components. Quartzo-feldspatic pegmatites, sapphirine-bearing granulites, calc-silicates and metacarbonates form the other common minor lithology (Figs 1, 2) (Tsunogae & Santosh, Reference Tsunogae and Santosh2006, Reference Tsunogae and Santosh2010; Endo et al. Reference Endo, Tsunogae, Santosh and Shaji2012; Plavsa et al. Reference Plavsa, Collins, Foden, Kropinski, Santosh, Chetty and Clark2012).

Fig. 2.

Simplified geological map of the study area marked in Figure 1 (after Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021) showing the Southern Madurai Block, Trivandrum Block and location of the Achankovil Shear Zone (AKSZ) proposed by Praharaj et al. (Reference Praharaj, Rekha and Bhattacharya2021). Location of the Thenmala Shear Zone (ThSZ) (Sacks et al. Reference Sacks, Nambiar and Walters1997) is marked. Squares (green) are the locations of monazite-dated samples and filled stars (yellow) are the locations of samples used in P-T calculations. White filled circles are the important key localities within the study area.

The regional strike of the supracrustal belts in the SMB is NE–SW, and that has been refolded about a N–S-trending axial plane (Drury et al. Reference Drury, Harris, Holt, Reeves-Smith and Wightman1984). According to Rajesh and Chetty (Reference Rajesh and Chetty2006), the SMB has a SE-plunging regional fold structure. Praharaj et al. (Reference Praharaj, Rekha and Bhattacharya2021) suggested that the S₂ gneissic layering in the SMB is steeply dipping and describes regional-scale steeply (westerly) plunging F₃ folds. The S₂ fabric is sigmoidally curved into the WNW-striking AKSZ as the shear zone is approached (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). At some parts of the SMB, gentle to open N- and S-closing folds are observed in the S₁–S₂ composite layering subparallel to the shear zone fabric, and these folds are progressively tighter to isoclinal near to the NNW- and NNE-striking melt-hosted (D₃) shears (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). Both the NNW as well as NNE-striking melt-hosted shears exhibit dextral and sinistral kinematics but no cross-cutting relations were noticed between the two sets of shears (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). The syn-D₃ generation of leucosomes along the shears was inferred from the tightening of folds in the S₁–S₂ composite layering adjacent to the melt-filled shears and the gradual disappearance radially away from the tip of the melt-filled shears (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021).

Enderbites, charnockites and alkali granites from the SMB show compositional variations from dioritic to granitic (Santosh et al. Reference Santosh, Hua, He, Li, Tsunogae, Shaji and Indu2017), whereas hornblende-biotite gneisses vary from granodioritic to tonalitic in composition (Chetty, Reference Chetty2017). The rocks are mostly peraluminous and magnesian in composition. The rock compositions vary from calc-alkaline series to shoshonite series and commonly show signatures of volcanic-arc evolution (Chacko et al. Reference Chacko, Ravindra Kumar, Meen and Rogers1992; Santosh et al. Reference Santosh, Hua, He, Li, Tsunogae, Shaji and Indu2017). Geothermobarometric calculations on charnockites suggest upper granulite-facies metamorphic conditions in the SMB with metamorphic P-T conditions ranging over 700–820°C and 5–7.5 kbar (Endo et al. Reference Endo, Tsunogae, Santosh and Shaji2012). However, other studies have reported UHT (> 900°C) metamorphism with high pressure (8–10 kbar) conditions along the northeastern parts (near Rajapalaiyam) of the SMB (Fig. 1) (Tateishi et al. Reference Tateishi, Tsunogae, Santosh and Janardhan2004; Braun et al. Reference Braun, Cenki-Tok, Paquette and Tiepolo2007; Tsunogae & Santosh, Reference Tsunogae and Santosh2010). Initial isobaric cooling followed by a near-isothermal decompression is suggested by Tateishi et al. (Reference Tateishi, Tsunogae, Santosh and Janardhan2004) for the sapphirine-bearing granulites of the SMB from Rajapalaiyam. Braun et al. (Reference Braun, Cenki-Tok, Paquette and Tiepolo2007) documented a clockwise P-T evolution divided into three parts: heating, cooling and decompression from the same sapphirine-bearing granulites. Tsunogae & Santosh (Reference Tsunogae and Santosh2006, Reference Tsunogae and Santosh2010) proposed an anticlockwise P-T trajectory with near-isobaric cooling for the stability fields of sapphirine + spinel + quartz assemblages of granulites from Rajapalaiyam.

Geochronological studies from the SMB revealed a polyphase Proterozoic metamorphic evolution with a widespread and major imprint of a Pan-African tectonothermal event and high-grade metamorphism (Santosh et al. Reference Santosh, Kagami, Yoshida and Nandakumar1992, Reference Santosh, Yokoyama, Biju-sekhar and Rogers2003, Reference Santosh, Tanaka, Yokoyama and Collins2005 b; Harris et al. Reference Harris, Santosh and Taylor1994; Bartlett et al. Reference Bartlett, Harris, Hawkesworth and Santosh1995, Reference Bartlett, Dougherty-Page, Harris, Hawkesworth and Santosh1998; Jayananda et al. Reference Jayananda, Janardhan, Sivasubramanian and Peucat1995; Bindu, Reference Bindu1997; Braun & Brocker, Reference Braun and Brocker2004; Ghosh et al. Reference Ghosh, de Wit and Zartman2004; Li et al. Reference Li, Santosh, Indu, Shaji and Tsunogae2016; Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021).

2.b. Trivandrum Block

The TB is mainly a garnet-biotite gneissic terrane with a large area covered by the khondalites (Figs 1, 2). The dominant lithologies in the TB are anatectic garnet-biotite ± graphite gneisses, garnet-bearing leptynitic gneisses, garnet-biotite-sillimanite ± cordierite (khondalite) gneisses, granitic and tonalitic orthogneisses, and non-garnetiferous and garnetiferous charnockites (quartzo-feldspathic gneisses with orthopyroxenes), together with minor amounts of quartzites and calc-silicates (Chacko et al. Reference Chacko, Ravindra Kumar and Newton1987; Santosh et al. Reference Santosh, Jackson and Harris1993, Reference Santosh, Morimoto and Tsutsumi2006; Bindu, Reference Bindu1997; Nandakumar & Harley, Reference Nandakumar and Harley2000; Satishkumar et al. Reference Satish-Kumar, Wada and Santosh2002; Shabeer et al. Reference Shabeer, Sajeev, Okudaira and Santosh2002; Endo et al. Reference Endo, Toshiaki, Santosh, Shimizu and Shaji2013; Sreejith & Ravindra Kumar, Reference Sreejith and Ravindra Kumar2013; Ravindra Kumar & Sreejith, Reference Ravindra Kumar and Sreejith2016). Khondalites, leptynites, calc-silicates and quartzites form the metapelitic suite, whereas charnockites, garnet-biotite gneisses and mafic granulites are mostly orthogneisses (Braun, Reference Braun2006; Tadokoro et al. Reference Tadokoro, Tsunogae and Santosh2008; Kröner et al. Reference Kröner, Santosh and Wong2012; Endo et al. Reference Endo, Toshiaki, Santosh, Shimizu and Shaji2013; Sreejith & Ravindra Kumar, Reference Sreejith and Ravindra Kumar2013; Harley & Nandakumar, Reference Harley and Nandakumar2014; Taylor et al. Reference Taylor, Clark, Fitzsimons, Santosh, Hand, Evans and McDonald2014; Kadowaki et al. Reference Kadowaki, Tsunogae, He, Santosh, Takamura, Shaji and Tsutsumi2019). The garnet-biotite gneisses commonly show arrested charnockitization (Ravindra Kumar et al. Reference Ravindra Kumar, Srikantappa and Hansen1985; Hansen et al. Reference Hansen, Janardhan, Newton, Prame and Ravindra Kumar1987; Raith & Srikantappa, Reference Raith and Srikantappa1993).

The gneissic rocks of the TB are characterized by NW–SE-trending, gently SW-dipping open upright folds that, on approach to the AKSZ, show increasing dip, and the dip direction changes to south (Drury et al. Reference Drury, Harris, Holt, Reeves-Smith and Wightman1984; Sacks et al. Reference Sacks, Nambiar and Walters1997; Rajesh et al. Reference Rajesh, Santosh and Yoshida1998; Cenki & Kriegsman, Reference Cenki and Kriegsman2005). The S₂ fabric in the TB dips at varying amounts, sub-horizontal to moderately inclined to the SW in the southern part, and steeply inclined towards the NE and SW as the NW–SE-striking sub-vertical AKSZ is approached (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). The poles of S₂ fabric falls on a well defined girdle showing NW–SE-trending upright asymmetric folds with sub-horizontal axes (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). The S₃-developed open folds become progressively tighter in close proximity to D₃ shear zones (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). As well as the AKSZ, numerous subordinate NW–SE-striking shear zones with both dextral and sinistral kinematics are present within a distance/zone of tens of kilometres in the TB (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). Two sets of melt-bearing, locally developed D₃ shears with sinistral NNE-striking shears and constant dextral NNW-striking shears are reported by Praharaj et al. (Reference Praharaj, Rekha and Bhattacharya2021), and it has been suggested that these leucosomes are coeval and conjugate in nature and formed along an extensional shear. In section view, a reverse sense of movement is observed with south-to-the-top translation (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021), which indicates that the TB is overriding the SMB.

The orthogneisses of the TB are metaluminous to peraluminous in composition and of two types: (1) granitic and mostly following an evolved ferroan character; and (2) tonalitic with a typically magnesian character (Chacko et al. Reference Chacko, Ravindra Kumar, Meen and Rogers1992; Braun, Reference Braun2006; Ravindra Kumar & Sreejith, Reference Ravindra Kumar and Sreejith2016; Nandakumar & Harley, Reference Nandakumar and Harley2018). Granitic gneisses and leucogranites are alkali-calcic to calc-alkalic and potassium-rich, whereas tonalitic varieties are mostly calcic in nature. The gneissic rocks (garnet-biotite gneiss and khondalite) yield mostly volcanic-arc signatures, and some are of island-arc affinities (Chacko et al. Reference Chacko, Ravindra Kumar, Meen and Rogers1992; Braun, Reference Braun2006; Ravindra Kumar & Sreejith, Reference Ravindra Kumar and Sreejith2016; Nandakumar & Harley, Reference Nandakumar and Harley2018). The earlier thermobarometric estimates suggested a temperature of between 650 and 750°C and a pressure of between 4.5 and 5.0 kbar for the TB (Chacko et al. Reference Chacko, Ravindra Kumar and Newton1987; Santosh et al. Reference Santosh, Jackson and Harris1993; Bindu, Reference Bindu1997). However, later studies also documented the UHT conditions (900–1000 °C) of metamorphism at parts of the TB (Fig. 1) (Nandakumar & Harley, Reference Nandakumar and Harley2000; Shimizu et al. Reference Shimizu, Tsunogae and Santosh2009; Endo et al. Reference Endo, Toshiaki, Santosh, Shimizu and Shaji2013). Overall, the pressure and temperature variations indicate upper amphibolite- to granulite-facies metamorphic conditions (Chacko et al. Reference Chacko, Ravindra Kumar and Newton1987; Braun et al. Reference Braun, Raith and Ravindra Kumar1996; Nandakumar & Harley, Reference Nandakumar and Harley2000). Although there is a slight difference in the pressure and temperature estimations in various studies from different rock types of the TB, a clockwise P-T path is generally accepted by different authors (Chacko et al. Reference Chacko, Ravindra Kumar and Newton1987; Morimoto et al. Reference Morimoto, Santosh, Tsunogae and Yoshimura2004; Tadokoro et al. Reference Tadokoro, Tsunogae and Santosh2008; Shimizu et al. Reference Shimizu, Tsunogae and Santosh2009; Blereau et al. Reference Blereau, Clark, Taylor, Johnson, Fitzsimons and Santosh2016; Kadowaki et al. Reference Kadowaki, Tsunogae, He, Santosh, Takamura, Shaji and Tsutsumi2019; Sorcar et al. Reference Sorcar, Joshi, Oliveira, Tomson and Nandakumar2020).

The TB was subjected to ductile deformation and metamorphism during the Pan-African orogeny (Choudhary et al. Reference Choudhary, Harris, Van Calsteren and Hawkesworth1992; Bartlett et al. Reference Bartlett, Dougherty-Page, Harris, Hawkesworth and Santosh1998; Braun & Kriegsman, Reference Braun and Kriegsman2003). Monazites from garnet-biotite gneisses, leucogranties and leucotonalites yielded ages ranging between 590 and 520 Ma (Braun & Brocker, Reference Braun and Brocker2004; Santosh et al. Reference Santosh, Maruyama and Sato2009; Taylor et al. Reference Taylor, Clark, Fitzsimons, Santosh, Hand, Evans and McDonald2014; Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). Metamorphic zircons from garnet-amphibole patch and metagranite suggested that the granulite-facies metamorphism has occurred at around 587–522 Ma (Taylor et al. Reference Taylor, Clark, Fitzsimons, Santosh, Hand, Evans and McDonald2014; Whitehouse et al. Reference Whitehouse, Ravindra Kumar and Rimsa2014). U–Th–Pb secondary-ion mass spectrometry (SIMS) geochronology revealed that magmatic crystallization in metagranite occurred at 1850 Ma (Whitehouse et al. Reference Whitehouse, Ravindra Kumar and Rimsa2014). U–Pb zircon geochronology and Hf isotope studies of the TB charnockites yielded 2.05–1.98 Ga ages and suggest a significant juvenile magma component in charnockite protoliths (Vijaya Kumar et al. Reference Vijaya Kumar, Rao, Plavsa, Collins, Tomson, Gopal and Babu2017). Granitoid gneisses and charnockites of the TB revealed protolith emplacement ages of 2100–1765 Ma (Kröner et al. Reference Kröner, Santosh and Wong2012, Reference Kröner, Santosh, Hegner, Shaji, Geng, Wong, Xie, Wan, Shang, Liu and Sun2015). Neoarchean–Mesoproterozoic (2.7–1.4 Ga) zircon ages were reported from granulite-facies metapelites and metagranitoids of the TB (Collins et al. Reference Collins, Santosh, Braun and Clark2007; Whitehouse et al. Reference Whitehouse, Ravindra Kumar and Rimsa2014; Kröner et al. Reference Kröner, Santosh, Hegner, Shaji, Geng, Wong, Xie, Wan, Shang, Liu and Sun2015).

2.c. Achankovil Shear Zone system

The AKSZ consists of cordierite-bearing granulites, garnet-biotite gneisses, migmatites, quartzites and charnockites with or without garnets (Santosh, Reference Santosh1987; Bartlett et al. Reference Bartlett, Dougherty-Page, Harris, Hawkesworth and Santosh1998; Rajesh & Chetty, Reference Rajesh and Chetty2006; Chetty, Reference Chetty2017). The rocks present within the high-strain zone are highly deformed and elongated parallel to the AKSZ (Rajesh & Chetty, Reference Rajesh and Chetty2006). Sacks et al. (Reference Sacks, Nambiar and Walters1997) suggested that the Thenmala shear zone (part of the AKSZ) is NW–SE-trending with a moderate to steep SW dip with kinematic indicators showing a predominantly dextral sense of shear with a minor components of reverse shear. In contrary, Rajesh et al. (Reference Rajesh, Santosh and Yoshida1998) interpreted the AKSZ as a WNW–ESE- to NW–SE-trending shear zone with a moderate dip to the SW and kinematic indicators suggesting a predominantly sinistral shear sense. A comprehensive structural analysis across the AKSZ by Praharaj et al. (Reference Praharaj, Rekha and Bhattacharya2021) supports the recent views that the AKSZ represents a high-T terrane boundary suture zone along which the TB and SMB were welded (e.g. Kröner et al. Reference Kröner, Santosh and Wong2012; Sreejith & Ravindra Kumar, Reference Sreejith and Ravindra Kumar2013; Ravindra Kumar & Sreejith, Reference Ravindra Kumar and Sreejith2016). According to Praharaj et al. (Reference Praharaj, Rekha and Bhattacharya2021), the AKSZ is not a single transpressive shear zone but essentially an anastomosing network of shear zones; the kinematics indicate that the main strand of the AKSZ is left lateral, although both dextral and sinistral kinematics are inferred in the subparallel shear zones occurring south of the AKSZ.

Pressure and temperature estimations from the AKSZ range over 5–7 kbar and 700–900 °C, respectively (Santosh, Reference Santosh1987; Chacko et al. Reference Chacko, Lamb and Farquhar1996; Nandakumar & Harley, Reference Nandakumar and Harley2000; Shabeer et al. Reference Shabeer, Satish-Kumar, Armstrong and Buick2005; Shimizu et al. Reference Shimizu, Tsunogae and Santosh2009; Collins et al. Reference Collins, Clark and Plavsa2014; Blereau et al. Reference Blereau, Clark, Taylor, Johnson, Fitzsimons and Santosh2016). A near-isothermal decompression path was deduced for garnet-cordierite gneisses by Santosh (Reference Santosh1987) and Sreejith & Ravindra Kumar (Reference Sreejith and Ravindra Kumar2012). Nandakumar & Harley (Reference Nandakumar and Harley2000) suggested a P-T path of initial isobaric cooling and decompression from 800 °C and 6 kbar to 750 °C and 4 kbar for charnockites and cordierite gneisses. Further, many studies proposed that the AKSZ attests a clockwise P-T evolution (Cenki et al. Reference Cenki, Braun and Brocker2004; Shabeer et al. Reference Shabeer, Satish-Kumar, Armstrong and Buick2005; Collins et al. Reference Collins, Clark and Plavsa2014).

Zircon geochronology indicates that the granulite-facies metamorphism in the AKSZ occurred at c. 513 Ma (Collins et al. Reference Collins, Santosh, Braun and Clark2007; Harley & Nandakumar, Reference Harley and Nandakumar2014). Zircons from cordierite gneisses gave a prominent age peak at 530 Ma (Bartlett et al. Reference Bartlett, Dougherty-Page, Harris, Hawkesworth and Santosh1998). Monazites from the garnet-biotite gneisses provided prominent c. 540 Ma ages (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021).

3.a. Southern Madurai Block

3.a.1. Garnet-biotite gneiss

Garnet-biotite gneisses of the SMB are mainly composed of garnet + biotite + quartz + plagioclase + K-feldspar assemblages (Fig. 4a). The tectonic fabric in garnet-biotite gneisses is defined by shape-preferred aggregates of biotite in a polycrystalline and polymineralic mosaic of dynamically recrystallized quartz and feldspar grains (Fig. 4a). Different dynamic recrystallization processes are indicated by sub-grain rotation recrystallization (Fig. 4b) and bulge nucleation as a result of grain boundary migration recrystallization (Fig. 4c). Quartz grains show chessboard twinning, undulose extinctions and vague sub-grain boundaries (Fig. 4b), whereas feldspars show weak undulose extinction (Fig. 4c). High-energy grain/phase boundaries are lined by discrete biotite flakes (Fig. 4c).

Fig. 4.

Representative microphotographs (a, d and e are plane-polarized images and b, c and f–i are crossed-polar images) from the Southern Madurai Block. (a) Garnet-biotite gneisses showing the S_2M fabric defined by the shape-preferred orientation of biotite. (b) Sub-grain rotation recrystallization (white arrow) in quartz, and (c) grain boundary migration between quartz and K-feldspar (black arrow) in garnet-biotite gneisses. (d) Charnockite with garnet, orthopyroxene, plagioclase, K-feldspar and quartz. (e) Orthopyroxenes in charnockite partially replaced by biotites. (f) Granoblastic texture with well equilibrated triple junctions among K-feldspar, quartz and plagioclase (white arrow) and (g) myrmekite texture (white arrow) in charnockites. Khondalite showing biotite and sillimanite enclosed within (h) first-generation cordierite (Crd1) and (i) second-generation cordierites (Crd2) surrounding garnet. Mineral abbreviations are after Kretz (Reference Kretz1983).

3.b. Trivandrum Block

3.b.1. Khondalite

Khondalites of the TB possess alternate garnet + biotite + sillimanite ± cordierite ± spinel-defined melanocratic and quartz + plagioclase + K-feldspar-defined leucocratic layers (Fig. 5a). Similar to the SMB khondalites, the khondalites of the TB also contain two generations of cordierites with a comparable growth history (Fig. 5b). Cordierites are commonly anhedral in shape, of variable sizes and contain numerous inclusion trails, for example, biotite, sillimanite, K-feldspar and plagioclase (Fig. 5a). Sillimanites are needle-like or tabular and aligned parallel to the gneissic fabric (Fig. 5a). Shape-preferred aggregates of early biotite grains are aligned parallel to the gneissic fabric, whereas the later individual biotite flakes lodged at grain and phase boundaries are randomly oriented (Fig. 5a). Subhedral to anhedral garnet porphyroblasts are common and possess inclusions of sillimanite, biotite, K-feldspar, plagioclase and quartz (Fig. 5a). Most of the quartz and K-feldspar grains show undulose extinction and the grain boundaries between them are curved with lobate and cuspate shapes, commonly showing bulge nucleation (Fig. 5c). Myrmekitic textures along the K-feldspar boundary are common in khondalites (Fig. 5d).

Fig. 5.

Representative microphotographs (e and g are plane-polarized images and remaining are crossed-polar images) from the Trivandrum Block. (a) Khondalite showing biotite and needle-like sillimanite enclosed within first-generation cordierite (Crd1) and (b) second-generation cordierites (Crd2) surrounding garnet. Images from khondalite showing (c) grain-boundary migration with lobate cuspate boundaries (white arrow) and (d) myrmekite texture between quartz and plagioclase (white arrow). (e) Charnockites showing orthopyroxene, clinopyroxene, hornblende, plagioclase, K-feldspar and quartz. (f) Orthopyroxene in charnockite partially replaced by biotite. (g) Garnet-biotite gneiss showing fabric-defining biotite and garnet porphyroblast. Images from garnet-biotite gneisses showing (h) granoblastic texture, (i) sub-grain rotation recrystallization and (j) quartz with undulose extinction. Mineral abbreviations are after Kretz (Reference Kretz1983).

3.c. Mineral reactions

The garnet-biotite gneisses, charnockites and khondalites from the SMB and TB show similar reaction textures. Three phases of mineral formation were identified in khondalites and these can be explained by the following possible reactions:

(1) $\begin{align}{\rm{Phase}}\,2&:\,{\rm{Bt}} + {\rm{Sil}} + {\rm{Pl}} + {\rm{Qtz}} + \left( {{\rm{Kfs}}} \right) \\ & \to {\rm{Grt}} + {\rm{Crd}}1 + {\rm{Melt}}\,({\rm{Spear}}\,\,et{\rm{ }}al.\,\,1999)\end{align}$

(2) $${\rm{Phase}}\,3:\,{\rm{Grt}} + {\rm{Sil}} + {\rm{Qtz}} \to {\rm{Crd}}2\,({\rm{Sorcar}}\,\,et{\rm{ }}al.\,\,2014)$$

In charnockites, biotite reacted with plagioclase and produced melt along with garnet, orthopyroxene and K-feldspar. It can be described by the following reaction:

(3) $\begin{align} {\rm{Bt}} + {\rm{Pl}} + {\rm{Qtz}} & \to {\rm{Grt}} + {\rm{Opx}} + {\rm{Kfs}} \\ & \quad \ +\, {\rm{melt}}\,({\rm{Tiwari}}\ \& \ {\rm{Sarkar}}\,\,2020) \end{align}$

The retrogression of orthopyroxene to fine-grained biotite can be described by the reaction:

(4) $${\rm{Opx}} + {\rm{Kfs}} + {{\rm{H}}_2}{\rm{O}} \to {\rm{Bt}}\,({\rm{Oziegbe}}\,et\,al.\,\,\,2020)$$

From the experimental results, Gardien et al. (Reference Gardien, Thompson and Ulmer2000) suggested that under fluid-absent conditions the melting of a quartz–biotite–plagioclase–K-feldspar assemblage could produce garnet + plagioclase II + melt with an increase in the modal amount of garnet as the pressure increases. This is consistent with our petrographic observations.

4.a. Southern Madurai Block

The rock-forming minerals of the major rock types, that is, four garnet-biotite gneisses (TV-9A, TV-32, TV-8 and TV-9D), one khondalite (PP-78) and two garnetiferrous charnockites (PP-77 and PP-90) were analysed to understand the mineral chemistry.

Garnet present in the SMB is a solid solution between pyrope and almandine, and the overall composition ranges from Alm_60–74Prp_26–40Grs_3–6Sps_2–5 in garnet-biotite gneisses and Alm_55–68Prp_32–45Grs₃Sps_3–6 in khondalites (online Supplementary Fig. S1a, available at http://journals.cambridge.org/geo). In charnockites, the average composition of garnets ranges from Alm_57–75Prp_25–43Grs_2–4Sps_3–6 (online Supplementary Fig. S1a). Irrespective of the rock types, the proportion of Fe, Mg, Ca and Mn in garnet is almost the same. There is no core-to-rim variation of compositions in garnet (online Supplementary Fig. S2, available at http://journals.cambridge.org/geo). The garnet shows a rim-wards depletion of Mg when it is associated with biotite.

The composition of biotite in garnet-biotite gneisses (K_0.86–0.92Mg_1.31–1.73Fe_0.76–0.97Al_1.17–1.43Si_2.67–2.83O₁₀(OH,F)₂) lies between phlogopite and annite, whereas in khondalites (K_0.90–0.96Mg_1.38–2.02Fe_0.54–0.98Al_1.29–1.39Si_2.66–2.75O₁₀(OH,F)₂) and charnockites (K_0.90–0.94Mg_1.57–1.90Fe_0.64–0.87Al_1.30–1.38Si_2.68–2.74O₁₀(OH,F)₂) the biotite grains are mostly phlogopite-rich. In garnet-biotite gneisses the X_Mg [=Mg/(Fe + Mg)] contents ranges from 0.40–0.68 (online Supplementary Fig. S1b). In khondalites the X_Mg contents ranges over 0.58–0.79, and in charnockites it ranges over 0.65–0.75 (online Supplementary Fig. S1b). TiO₂ content in biotite is variable and ranges over 3.52–7.74 wt% in khondalites and 5.41–7.28 wt% in garnet-biotite gneisses. Biotite in charnockites has a lower TiO₂ (5.06–6.09 wt%) content than khondalites and garnet-biotite gneisses.

Chemical variation of orthopyroxene in charnockites ranges over (Mg,Fe)_1.82–1.90Si_1.77–1.81O₆. The X_Fe [=Fe/(Fe + Mg)] in the range of 0.37–0.42 indicates that the orthopyroxenes in charnockites of the SMB are enstatite-rich (online Supplementary Fig. S1c). Orthopyroxenes present in charnockites have low Al content in the range of 0.33–0.37 and these orthopyroxenes are metamorphic in origin (Rajesh et al. Reference Rajesh, Santosh and Yoshikura2011).

Cordierite composition ranges over Mg_1.59–1.63Al_3.98–4.05Si_4.93–4.98O₁₈ in khondalites. X_Mg in cordierites ranges over 0.78–0.81. The cordierites present here are mostly anhydrous (online Supplementary Table S1).

In khondalites, charnockites and garnet-biotite gneisses, the plagioclase is typically oligoclase to andesine (Ab_59–71An_23–37Or_1–4) in composition (online Supplementary Fig. S1d). Alkali feldspar in garnet-biotite gneisses (Ab_10–22An_0–1Or_77–89) contain more albitic component compared with those in khondalites (Ab_3–10An_0–1Or_89–96) and charnockites (Ab_9-10An_0-1Or_90-91).

4.b. Trivandrum Block

The dominant rock types such as khondalite (PP-32, PP-41, PP-72 and TV-65), garnet-biotite gneiss (PP-34, PP-84, TV-60, TV-71 and TV-291), garnetiferous charnockite (TV-304) and non-garnetiferous charnockite (PP-51 and TV-69) were studied in detail to understand the mineral chemistry.

Garnet present here is also a solid solution between pyrope and almandine. In garnet-biotite gneisses, the average composition of garnets ranges over Alm_84–89Prp_11–16Grs_7–10Sps_1–2. The average composition of garnet in khondalites ranges over Alm_67–82Prp_18–33Grs_2–3Sps_1–2. In charnockites, the average composition of garnets ranges over Alm_76–84Prp_16–24Grs_3–4Sps_1–2. Garnets are mostly almandine-rich (online Supplementary Fig. S1e). Garnet porphyroblasts in contact with biotite shows a slight rim-wards depletion of Mg as a result of the Fe–Mg exchange between garnet and biotite. The garnet does not show any significant zoning in terms of Fe, Mg, Ca and Mn (online Supplementary Fig. S2).

The composition of biotite varies from phlogopite to annite in garnet-biotite gneisses (K_0.86–0.95Mg_0.86–1.92Fe_0.61–1.52Al_1.15–1.32Si_2.74–2.93O₁₀(OH,F)₂), whereas the biotite grains present in khondalites (K_0.87–0.95Mg_1.42–1.95Fe_0.56–0.97Al_1.18–1.31Si_2.77–2.86O₁₀(OH,F)₂) and charnockites (K_0.78–0.91Mg_1.24–1.73Fe_0.85–1.14Al_1.20–1.28Si_2.82–2.99O₁₀(OH,F)₂) are phlogopite-rich. The X_Mg value ranges over 0.60–0.78 in khondalites and 0.52–0.67 in charnockites (online Supplementary Fig. S1f). In garnet-biotite gneisses, the X_Mg value ranges over 0.37–0.76 (online Supplementary Fig. S1f). The TiO₂ content of biotite ranges over 3.04–6.01 wt% in khondalites and 3.01–6.71 wt% in garnet-biotite gneisses. The TiO₂ content in charnockites are slightly lower (1.11–5.46 wt%) than for khondalites and garnet-biotite gneisses.

The composition of orthopyroxene ranges over (Mg,Fe)_1.81–2.32Si_1.88–2.04O₆ and that of clinopyroxene over (Mg,Fe,Ca,Al)_1.91–2.30(Si,Al)_2.01–2.06O₆ in both garnetiferous and non-garnetiferous charnockites. The X_Fe value of orthopyroxenes ranges over 0.58–0.85, indicating that ferrosilite-rich orthopyroxenes are present in both types of charnockites, while in a few samples it varies between ferrosilite and pigeonite (online Supplementary Fig. S1g). Clinopyroxenes are mostly present in non-garnetiferous charnockites where the X_Fe content varies between 0.58 and 0.79, indicating the predominance of augite composition (online Supplementary Fig. S1g). The lesser amount of Al (0.02–0.03 apfu) in orthopyroxene is a characteristic of the charnockites of the SGT, and indicates a metamorphic origin (Rajesh et al. Reference Rajesh, Santosh and Yoshikura2011; Endo et al. Reference Endo, Tsunogae, Santosh and Shaji2012). The Al content (0.14–0.16 apfu) of orthopyroxene in garnetiferous charnockites is slightly higher compared with the non-garnetiferous charnockites.

Chemical composition of cordierite in khondalites ranges over Mg_1.32–1.39Al_3.89–3.97Si_5.03–5.07O₁₈. The X_Mg composition of cordierites are almost uniform and lie between 0.67 and 0.70. The X_Mg value in cordierite is slightly lower in the TB than in the SMB. Cordierites present here are hydrous (online Supplementary Table S1a).

Plagioclase feldspars are of oligoclase to andesine in composition in khondalites (Ab_62–74An_24–36Or_1–4), garnet-biotite gneisses (Ab_58–77An_22–41Or_1–2) and charnockites (Ab_76–77An_19–20Or_3–4) (online Supplementary Fig. S1h). Both perthites and antiperthites are present in different lithologies of the TB. Similar to the SMB, the K-feldspar in garnet-biotite gneisses (Ab_11–28An_1–3Or_69–88) are slightly more albitic than in khondalites (Ab_3–32An_0–4Or_64–96) and charnockites (Ab_3–31An_0–7Or_67–96) of the TB.

5.a. Thermobarometry

Four garnet-biotite gneisses (TV-9A, TV-32, TV-8 and TV-9D), one garnet-biotite-sillimanite ± cordierite gneiss (PP-78) and two garnetiferous charnockites (PP-77 and PP-90) were chosen for temperature and pressure estimations from the SMB. Five garnet-biotite gneisses (PP-34, PP-84, TV-60, TV-71 and TV-291), four garnet-biotite-sillimanite ± cordierite gneisses (PP-33, PP-41, PP-72 and TV-65) and one garnetiferous charnockite (TV-304) were selected from the TB. The sample locations are shown in Figure 2 and samples are described in Table 1.

Temperatures were estimated using garnet-biotite Fe–Mg exchange thermometric formulations proposed by Hodges & Spear (Reference Hodges and Spear1982), Hodges & Crowley (Reference Hodges and Crowley1985) and Bhattacharya et al. (Reference Bhattacharya, Mohanty, Maji, Sen and Raith1992). Non-ideal mixing conditions for all minerals are assumed for the calculations based on Hodges & Crowley (Reference Hodges and Crowley1985) and Bhattacharya et al. (Reference Bhattacharya, Mohanty, Maji, Sen and Raith1992), whereas ideal mixing for all minerals, except where Ti forms a major biotite component, is assumed for Hodges & Spear (Reference Hodges and Spear1982) computations. The garnet-cordierite thermometer by Bhattacharya et al. (Reference Bhattacharya, Mazumdar and Sen1988) was used to calculate the temperature from the samples where cordierite and sillimanite were present as extra phases. The garnet-orthopyroxene Fe–Mg exchange thermometer of Sen & Bhattacharya (Reference Sen and Bhattacharya1984), Harley (Reference Harley1984) and Lal (Reference Lal1993) was used to calculate the temperature in garnetiferous charnockites. In the garnet-orthopyroxene thermometer of Sen & Bhattacharya (Reference Sen and Bhattacharya1984), orthopyroxene is taken as ideal mixing and garnet is taken as non-ideal mixing. In Harley (Reference Harley1984), all minerals are considered as ideal mixing, while in Lal (Reference Lal1993) all minerals are considered as non-ideal mixing.

Pressure conditions were determined by using the GRIPS (garnet-rutile-ilmenite-plagioclase-quartz) barometer (Bohlen & Liotta, Reference Bohlen and Liotta1986) in garnet-biotite gneisses. In garnet-bearing charnockites, the GOPQ (garnet-orthopyroxene-plagioclase-quartz) barometer (Perkins & Newton, Reference Perkins and Newton1981; Bhattacharya et al. Reference Bhattacharya, Krishnakumar, Raith and Sen1991; Lal, Reference Lal1993) was used. GASP (garnet-sillimanite-quartz-plagioclase) (Ghent, Reference Ghent1976; Koziol & Newton, Reference Koziol and Newton1988), GRAIL (garnet-rutile-sillimanite-ilmenite) (Ghent & Stout, Reference Ghent and Stout1984) and garnet-cordierite-sillimanite-quartz (Carey, Reference Carey1995) barometers were used for the calculation of pressure in garnet-biotite-sillimanite ± cordierite gneisses. In the GASP barometer the a–X relations proposed by Newton & Perkins (Reference Newton and Perkins1982) and Ganguly & Saxena (Reference Ganguly and Saxena1984) were used for computing the activity of garnet, while the formulation of Newton & Haselton (Reference Newton, Haselton, Newton, Navrotsky and Wood1981) was used for computing the activity of anorthite in plagioclase. The P-T estimates are provided in Tables 4 and 5 where the temperatures are calculated at P = 7 kbar and pressures are estimated at T = 700°C.

Table 4.

Estimated temperature (°C) at P = 7 kbar using mineral thermometry. Sample locations are indicated in Figure 2. n – nearest neighbor mineral in matrix; r – rim of mineral. Mineral abbreviations after Kretz (Reference Kretz1983).

Table 5.

Estimated pressure (kbar) at T = 700°C using mineral barometry. Sample locations are indicated in Figure 2.

GASP – garnet-aluminosilicate-plagioclase-quartz; GRAIL – garnet-rutile-sillimanite-ilmenite-quartz; GRIPS – garnet-rutile-ilmenite-plagioclase-quartz; Grt-Crd – garnet-cordierite; GOPQ – garnet-orthopyroxene-plagioclase-quartz.

5.b. P-T path reconstruction

A P-T path can be obtained by integrating the outcomes of P-T pseudosection analysis, measured in zoning profiles in garnet and the different reaction textures involved (Spear, Reference Spear1992; White & Powell, Reference White and Powell2011). Since the garnets of granulite-facies rocks are devoid of zoning due to high-temperature diffusion, the P-T path is constructed by considering different textural relations and reactions among minerals. The phase equilibrium modelling (Holland & Powell, Reference Holland and Powell1998) was carried out using the Perple_X software (version 6.8.7) (Connolly & Petrini, Reference Connolly and Petrini2002; Connolly, Reference Connolly2005, Reference Connolly2009). The bulk-rock composition used for the construction of the P-T pseudosection was obtained using inductively coupled plasma – atomic emission spectroscopy (ICP-AES) at the Sophisticated Analytical Instrument Facility, IIT Mumbai, India (analytical conditions are described in Hrushikesh et al. Reference Hrushikesh, Prabhakar and Bhattacharya2019). The computed pseudosection is contoured with modal volume isopleths of garnet, cordierite and orthopyroxene to constrain the P-T paths. In all pseudosections, plagioclase and quartz are stable in all the fields. Rutile and ilmenite are considered as pure phases. The fluid, considered as pure H₂O, was constrained by the loss on ignition (LOI) in the computations.

Most of the existing P-T pseudosections from the SMB constructed on sapphirine-bearing granulites of Rajapalaiyam area (Fig. 1) provided UHT conditions (Tateishi et al. Reference Tateishi, Tsunogae, Santosh and Janardhan2004; Tsunogae & Santosh, Reference Tsunogae and Santosh2006, Reference Tsunogae and Santosh2010). However, the SMB lacks such information from the metapelitic gneisses, charnockites and garnet-biotite gneisses. In Contrast, in the TB the P-T paths are well documented from metapelitic gneisses and charnockites (Satish-Kumar & Harley, Reference Satish-Kumar and Harley1998; Morimoto et al. Reference Morimoto, Santosh, Tsunogae and Yoshimura2004; Tadokoro et al. Reference Tadokoro, Tsunogae and Santosh2008; Shimizu et al. Reference Shimizu, Tsunogae and Santosh2009; Blereau et al. Reference Blereau, Clark, Taylor, Johnson, Fitzsimons and Santosh2016; Wang et al. Reference Wang, Zhai, He, George, Sajeev, Jiao, Liu and Guo2021), whereas no data are available from garnet-biotite gneisses. In this case, an attempt is made to reconstruct the P-T psuedosection from the common lithologies of the SMB and TB. One sample each from the TB and SMB from khondalites (PP-33 and PP-78), garnet-biotite gneisses (TV-291 and TV-9A) and charnockites (TV-304 and PP-77) were used for the P-T psuedosection modelling (Figs 6, 7), as these three rocks form the major lithological components of both blocks.

Fig. 6.

NCKFMASH P-T pseudosections of khondalites (a and d), garnet-biotite gneisses (b and e) and charnockites (c and f). (a–c) Southern Madurai Block; (d–f) Trivandrum Block. The diagrams show stable mineral assemblages in the P-T window of 3–10 kbar and 450–900 °C. The molar proportions of element oxides used in the computations are shown in a box at the top of each diagram. Area marked with black bold line represents the texturally relevant stable phase assemblages used in reconstructing the P-T path. Red arrows show the possible clockwise P-T paths experienced by the rocks. Mineral abbreviations are after Kretz (Reference Kretz1983), except Fs – feldspar.

Fig. 7.

Summary diagram showing P-T path from the (a) Southern Madurai and (b) Trivandrum blocks. The texturally relevant stable phase assemblages from each P-T psuedosection and P-T paths (in red) are combined (shown with black bold line and red arrow in Fig. 6) to construct a common P-T path. Bold purple arrow shows the possible clockwise P-T path experienced by the rocks. P-T paths from previous studies are drawn for reference. T1: Tateishi et al. (Reference Tateishi, Tsunogae, Santosh and Janardhan2004); T2: Tsunogae & Santosh (Reference Tsunogae and Santosh2006); T3: Braun et al. (Reference Braun, Cenki-Tok, Paquette and Tiepolo2007); T4: Tsunogae & Santosh (Reference Tsunogae and Santosh2010); T5: Blereau et al. (Reference Blereau, Clark, Taylor, Johnson, Fitzsimons and Santosh2016); T6: Shabeer et al. (Reference Shabeer, Sajeev, Okudaira and Santosh2002); T7: Tadokoro et al. (Reference Tadokoro, Tsunogae and Santosh2008); T8: Kröner et al. (Reference Kröner, Santosh, Hegner, Shaji, Geng, Wong, Xie, Wan, Shang, Liu and Sun2015); T9: Chacko et al. (Reference Chacko, Ravindra Kumar and Newton1987); T10: Kadowaki et al. (Reference Kadowaki, Tsunogae, He, Santosh, Takamura, Shaji and Tsutsumi2019); T11: Wang et al. (Reference Wang, Zhai, He, George, Sajeev, Jiao, Liu and Guo2021). Relative stability fields of kyanite-sillimanite-andalusite are marked for reference.

The chemical system NCKFMASH (Na₂O-CaO-K₂O-FeO-MgO-Al₂O₃-SiO₂-H₂O) was chosen for P-T psuedosection construction. The calculation involves the following phases and phase components (in square brackets): garnet, Gt [alm, py, gr]; cordierite, hCrd [crd, fcrd, hcrd]; biotite, Bio [ann, phl, east]; mica [mu, pa, cel, ma]; pyroxene, Opx [mgts, en, fs]; spinel, Sp [sp, herc]; melt, Melt [fo8L, fa8L, kspL, h2oL, abL, sil8L, anL, q8L]; feldspar, Pl [abh, an]; K-feldspar, Kfs [san]; sillimanite [sill]; kyanite [ky]; and quartz [q] (mineral abbreviations as per Connolly, Reference Connolly2005). The details of the solution models are available in solution_model.dat; PERPLEX 6.8.7; database: hp62ver.dat. The solution models adopted are Gt(HP), Mica(CHA), Bio(TCC), feldspar, San, hCrd, Opx(HP), Sp (HP) and melt(HP). In the P-T psuedosection, the phase topologies are computed in the P-T window extending over 450–900 °C and 3–10 kbar (Fig. 6).

6.a. Southern Madurai Block

From the SMB, three garnet-biotite gneisses (TV-9C, TV-28 and TV-32) and one garnet-bearing charnockite (PP-77) were analysed (Fig. 2). In the garnet-biotite gneisses the monazite grains are subhedral to anhedral in shape and the length ranges over 80–200 µm (Fig. 8a, b). Compared with garnet-biotite gneiss, the monazite grains in charnockite are sparse and smaller in size with length ranging over 20–100 μm and anhedral in shape (Fig. 8c, d). In garnet-biotite gneisses, monazite grains are hosted within the cores as well as rims of garnet porphyroblasts, aligned parallel to or overgrowing the biotite and lodged at the margins of K-feldspars, plagioclase and quartz (Fig. 8a, b). In garnetiferous charnockite, monazites mostly occur within or at the periphery of K-feldspar, plagioclase and quartz; in rare cases, monazites are hosted within garnet, orthopyroxenes and biotite aggregates (Fig. 8c, d). Monazite grains of the SMB are invariably zoned from core to rim for Th and Y; the core is Y-rich and Th-poor, and the rim is Y-poor and Th-rich (Fig. 9a, b). U and Pb variations are either lacking or else exhibit significantly weaker chemical zoning relative to Th and Y.

Fig. 9.

Th and Y X-ray element maps and backscattered electron images of representative monazites from the (a, b) Southern Madurai Block, hosted in the matrix of garnet-biotite gneisses, and the (c–f) Trivandrum Block. Hosted within (c, d, f) garnet and (e) matrix of khondalites. White circles and keyed numbers are locations of analysed spots and spot ages (± 2σ error) in Ma, respectively.

A total of 23 spots were analysed on monazite grains from the garnet-biotite gneisses and 14 spots were analysed from garnet-bearing charnockite. Monazites from garnet-biotite gneisses of the SMB yielded one distinct age peak at 528 ± 5 Ma (Fig. 10a). Similarly, monazites from garnet-bearing charnockite yielded a prominent age peak at 518 ± 5 Ma (Fig. 10b). All the obtained chemical ages from the SMB were consistently late Neoproterozoic – Cambrian, irrespective of chemical zoning and textural association.

Fig. 10.

Probability density plots of monazite age data from different lithologies of the study area (sample locations in Fig. 2). Data from (a) garnet-biotite gneisses and (b) garnetiferous charnockites from the Southern Madurai Block and from (c) khondalites and (d) garnet-biotite gneisses from the Trivandrum Block. (e) Date from the Achankovil Shear Zone khondalite obtained from the present study (in red) and the monazite age data from garnet-biotite gneisses by Praharaj et al. (Reference Praharaj, Rekha and Bhattacharya2021) (in green) are plotted. Number of samples analysed (s) and number of spot ages (n). Errors are in ± 2σ.

6.b. Trivandrum Block

Three samples (Fig. 2) were selected for monazite analysis from the TB, of which two were khondalites (TV-65 and PP-113) and one was garnet-biotite gneiss (PP-10A). A total of 67 spots were analysed on monazite grains from khondalites and 31 spots were analysed from garnet-biotite gneiss. Monazites from khondalites are euhedral to subhedral in shape, larger in size (length c. 100–200 µm and width c. 50–100 µm), hosted within the cores and rims of garnet porphyroblasts, and occur along the flaky biotite and sillimanite and the margins of cordierite, plagioclase, K-feldspar and quartz (Fig. 8e, f). In the garnet-biotite gneiss, the monazite grains are mostly hosted within the cores and rims of garnets and fractures and cleavages of feldspars and quartz, overgrowing the fabric-defining biotite. They are subhedral to anhedral with long axis ranging over 100–250 µm and short axis ranging over 50–150 µm (Fig. 8g, h). Monazites from the TB were invariably zoned in Th and Y with patchy chemical zones (Fig. 9c–f), but the spot ages did not show any dependence on the chemical zoning (Fig. 9c–f).

Spot chemical ages of monazites from khondalites yielded a prominent age peak at 556 ± 4 Ma (Fig. 10c). Other less-prominent age populations are also present, for example, 1858 ± 27 Ma, 1263 ± 17 Ma and 982 ± 12 Ma (Fig. 10c). The probability density plot for garnet-biotite gneiss yielded one prominent age peak at 584 ± 7 Ma and three unmixed ages at 1373 ± 49 Ma, 1299 ± 34 Ma and 984 ± 15 Ma (Fig. 10d).

6.c. Achankovil Shear Zone system

Monazite grains from khondalites (TV-22) of the AKSZ (sample location in Fig. 2) are euhedral in shape, randomly distributed in garnet porphyroblasts, occur along the grain boundaries of plagioclase, quartz, K-feldspar and cordierites, and are overgrowing, aligned parallel to or included within biotite and sillimanite aggregates. Irrespective of their patchy and nebulous zoning in Th and Y, two monazite spot ages from the AKSZ gave a single prominent age peak at 507 ± 16 Ma (Fig. 10e).

7.a. Metamorphic P-T evolution

Khondalites, garnet-biotite gneisses and charnockites, the dominant lithologies from the TB and SMB, show evidence of high-T deformation in the form of dynamic recrystallization manifested by aggregates of xenoblastic minerals, triple junctions and granoblastic texture. Triple junctions sharing straight grain/phase boundaries indicate high-T annealing by grain-boundary area reduction (Altenberger, Reference Altenberger1992). Cuspate- and lobate-shaped grains were produced as a result of grain-boundary migration recrystallization with increasing temperature and decreasing flow stress (Passchier & Trouw, Reference Passchier and Trouw2005). These grain-size modifications in quartz and feldspars are mainly due to the deformation at high-grade conditions, possibly the result of solid-state diffusive mass transfer (Passchier & Trouw, Reference Passchier and Trouw2005). The textural relations imply that deformation at high-T induced grain-boundary-slip-accommodated dynamic recrystallization, and consequently randomized the grain shape and orientation. Polygonization or annealing recrystallization was limited in all rocks. Across both the SMB and TB, granulite-facies mineral assemblages retrogressed at a later stage.

There are no significant compositional variations in minerals present in different lithounits of the TB and SMB. Garnets are mostly almandine-rich and biotite compositions vary between phlogopite and annite in garnet-biotite gneisses, khondalites and charnockites of the TB and SMB. Orthopyroxenes in charnockites are ferrosilite-rich in the TB and enstatite-rich in the SMB. Plagioclase feldspars are oligoclase to andesine in composition in all three rock types of the TB and SMB.

In the SMB, the temperature conditions estimated by using garnet-biotite thermometry range over 600–780 °C in garnet-biotite gneisses and 600–650 °C in khondalites. The garnet-biotite thermometers provided slightly lower temperatures for khondalites compared with the garnet-biotite gneisses. The temperature of 830 °C yielded by the garnet-cordierite thermometer from khondalites is comparatively higher than the temperature obtained from garnet-cordierite granulite (650–740 °C) of the SMB (Tsunogae & Santosh, Reference Tsunogae and Santosh2010). The higher temperature indicates melting of phase-1 assemblages and the formation of garnet-cordierite (Crd1) assemblage in khondalites. Temperatures obtained from charnockites using garnet-orthopyroxene thermometry (740–860 °C) are comparable with the temperatures (710–820 °C) obtained by Endo et al. (Reference Endo, Tsunogae, Santosh and Shaji2012). Metamorphic pressures obtained from the SMB (5–8 kbar) by multiple barometers are consistent with earlier studies (7–8 kbar; Tateishi et al. Reference Tateishi, Tsunogae, Santosh and Janardhan2004; Tsunogae & Santosh, Reference Tsunogae and Santosh2010).

In the TB, the temperatures obtained from charnockites (720–880 °C) using the garnet-orthopyroxene thermometer is identical to that obtained in earlier studies (810–870 °C; Endo et al. Reference Endo, Toshiaki, Santosh, Shimizu and Shaji2013; Wang et al. Reference Wang, Zhai, He, George, Sajeev, Jiao, Liu and Guo2021). Temperatures estimated by garnet-biotite thermometry in garnet-biotite gneisses (615–770 °C) and khondalites (630–720 °C) are similar to that reported by Chacko et al. (Reference Chacko, Ravindra Kumar and Newton1987) (650–800 °C) using the same method. The 780 °C temperatures obtained from khondalites using garnet-cordierite (Crd1) thermometry is close to the c. 815 °C reported by Morimoto et al. (Reference Morimoto, Santosh, Tsunogae and Yoshimura2004), while it is slightly lower compared with the 860–960 °C reported by Shabeer et al. (Reference Shabeer, Sajeev, Okudaira and Santosh2002). Pressures estimated using various barometers range over 5–8 kbar and are consistent with the values reported by Shabeer et al. (Reference Shabeer, Sajeev, Okudaira and Santosh2002), Blereau et al. (Reference Blereau, Clark, Taylor, Johnson, Fitzsimons and Santosh2016) and Sorcar et al. (Reference Sorcar, Joshi, Oliveira, Tomson and Nandakumar2020), but higher than the 4.5–5.5 kbar reported by Bindu (Reference Bindu1997). The metamorphic temperatures obtained from different lithologies using various thermometers range over 600–860 °C for the SMB and 615–880 °C for the TB. Similar to temperature, the estimated pressures from both the TB and SMB are of same range (5–8 kbar).

The phase equilibrium modelling from the SMB and TB using khondalites, garnet-biotite gneisses and charnockites provided clockwise P-T paths that are consistent with the field and petrographic observations. With the exception of Braun et al. (Reference Braun, Cenki-Tok, Paquette and Tiepolo2007), all other researchers (Tateishi et al. Reference Tateishi, Tsunogae, Santosh and Janardhan2004; Tsunogae & Santosh, Reference Tsunogae and Santosh2006, Reference Tsunogae and Santosh2010) proposed an anticlockwise P-T trajectory for the sapphirine-bearing granulites from the SMB. The clockwise P-T path estimated by Braun et al. (Reference Braun, Cenki-Tok, Paquette and Tiepolo2007) from sapphirine-bearing granulites has a higher P-T range compared with this study (Fig. 7a). In the TB, all estimated clockwise P-T paths (with the exception of that in Chacko et al. Reference Chacko, Ravindra Kumar and Newton1987) were at a higher P-T range (Nandakumar & Harley, Reference Nandakumar and Harley2000; Shabeer et al. Reference Shabeer, Sajeev, Okudaira and Santosh2002; Cenki et al. Reference Cenki, Braun and Brocker2004; Collins et al. Reference Collins, Clark and Plavsa2014; Clark et al. Reference Clark, Healy, Johnson, Collins, Taylor, Santosh and Timms2015; Johnson et al. Reference Johnson, Clark, Taylor, Santosh and Collins2015; Blereau et al. Reference Blereau, Clark, Taylor, Johnson, Fitzsimons and Santosh2016; Wang et al. Reference Wang, Zhai, He, George, Sajeev, Jiao, Liu and Guo2021) than this study. Blereau et al. (Reference Blereau, Clark, Taylor, Johnson, Fitzsimons and Santosh2016) suggested a clockwise P-T path by combining garnet-biotite gneiss, charnockite and metapelitic gneiss from a single location where the charnockitization was in progress (Fig. 7b). The P-T paths obtained by Shabeer et al. (Reference Shabeer, Sajeev, Okudaira and Santosh2002), Tadokoro et al. (Reference Tadokoro, Tsunogae and Santosh2008), Kröner et al. (Reference Kröner, Santosh, Hegner, Shaji, Geng, Wong, Xie, Wan, Shang, Liu and Sun2015), Kadowaki et al. (Reference Kadowaki, Tsunogae, He, Santosh, Takamura, Shaji and Tsutsumi2019) and Wang et al. (Reference Wang, Zhai, He, George, Sajeev, Jiao, Liu and Guo2021) using different rock types such as Mg-Al metapelites, migmatitic pelitic granulite, charnockite and khondalite, respectively, are almost identical to or at a higher P-T range compared with this study. The similar clockwise P-T paths obtained from the SMB and TB represent the last deformation and metamorphic event shared between the SMB and TB during their accretion along the AKSZ. Most of the earlier metamorphic signatures were overprinted during this high-grade metamorphic event.

7.b. Timing of deformation and metamorphism

Dissolution and reprecipitation of monazites are common processes during deformation (Williams et al. Reference Williams, Jercinovic and Hetherington2007, Reference Williams, Jercinovic, Harlov, Budzyń and Hetherington2011) and most of the monazite grains from the SMB appeared to have been dissolved and reprecipitated during the last deformation that occurred during late Neoproterozoic – Cambrian time. Monazites from all the rock types of the SMB provide robust ages at c. 528 and 518 Ma, and older ages are completely absent. However, the monazites from the TB did not undergo complete reprecipitation except in parts. Monazites from different lithologies of the TB provided prominent peaks at c. 584 and 556 Ma with other less-prominent age populations at c. 1850, 1370, 1275 and 980 Ma. The age data obtained here are in good agreement with the available data (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). The tectonic significance of all the age populations could not be explained in relation to the deformation in the TB because of the coarse-grained nature of the rock and the centimetre-wide gneissic layers. In such coarse-grained rocks with wide gneissic layers, it is difficult to correlate the monazite growth with deformation and the associated metamorphic event.

Published Pb–Pb zircon age data suggest that during 800–700 Ma the SMB experienced widespread igneous activities and metamorphism (Ghosh et al. Reference Ghosh, de Wit and Zartman2004; Braun & Appel Reference Braun and Appel2006; Santosh et al. Reference Santosh, Maruyama and Sato2009, Reference Santosh, Hua, He, Li, Tsunogae, Shaji and Indu2017; Plavsa et al. Reference Plavsa, Collins, Foden, Kropinski, Santosh, Chetty and Clark2012; Li et al. Reference Li, Santosh, Indu, Shaji and Tsunogae2016), whereas these ages are not very common in the TB. In addition, the SMB lacks the prominent Palaeoproterozoic ages (c. 2.1 Ga) present in the TB (Santosh et al. Reference Santosh, Collins, Morimoto and Yokoyama2005 a; Kröner et al. Reference Kröner, Santosh, Hegner, Shaji, Geng, Wong, Xie, Wan, Shang, Liu and Sun2015). However, both terranes record strong imprints of late Neoproterozoic – Cambrian metamorphic ages (Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021). The difference in chronology also supports the fact that the TB and SMB evolved differently prior to their accretion along the AKSZ during late Neoproterozoic – Cambrian time.

7.c. Tectonic implications

The integrated structural, metamorphic and chronological analysis combined with the existing data allowed the tectonothermal evolution of the TB and SMB across the AKSZ to be determined. The structural studies reveal that the SMB granulite-facies rocks describe steeply plunging reclined F₂ tight folds with NNW/NW-trending axial planes, whereas the TB granulites show gently plunging open F₂ folds with moderately dipping WNW-trending axial planes in regional scale. The last deformation (D₃) shared by the TB and SMB during their accretion along the AKSZ produced NNW/NNE-trending subparallel sinistral and dextral S₃ shears on the SMB and TB. The garnet-bearing melt-hosted syn-D₃ conjugate shears attest to the prevalence of granulite-facies anatectic conditions during the D₃ deformation in both the SMB and TB. The metamorphic P-T conditions obtained from different lithologies of the SMB and TB provide similar temperature and pressure estimates (c. 600–880 °C and 5.0–8.0 kbar) as well as similar clockwise P-T paths with an initial prograde path followed by a later decompression path. The initial heating and loading could be the result of the thrusting, and the later decompression path could be the result of rapid uplift or crustal thinning. Post-collisional extension and associated mafic magmatism from a suite of gabbroic rocks were reported from the TB by Yang et al. (Reference Yang, Ganguly, Shaji, Dong and Nandakumar2019), and the change in the tectonic regime from a compressional to an extensional setting might have caused the rapid uplift and retrogression. The lack of early metamorphic signatures may be the manifestation of high-grade metamorphism and melting during the D₃ event. The monazite ages obtained from the TB are slightly older (c. 584 and 556 Ma) than the monazite ages obtained from the SMB (c. 528 and 518 Ma). The c. 584–556 Ma TB ages can be attributed to the initial thickening and high-grade metamorphism of the orogen, whereas the slightly younger ages (c. 528–518 Ma) from the SMB could be the result of the decompression of overthickened orogen. This diachroneity in ages was also well documented in the NMB (Plavsa et al. Reference Plavsa, Collins, Foden and Clark2015). The late Neoproterozoic – Cambrian monazite ages obtained from different lithologies of the TB and SMB are similar to the age population obtained from the AKSZ. These strong late Neoproterozoic – Cambrian imprints on monazites from all lithounits of the TB and SMB suggests that both these blocks were accreted along the AKSZ during the Pan-African Orogeny and most of the monazites reprecipitated during the last deformation (D₃) occurred at high-grade conditions associated with the accretion.

The results from our study reveal multiple deformation in the TB and SMB during the late Neoproterozoic – Cambrian period. Along with the structural discordance, the differences in chronology, metamorphic P-T range and geochemistry support the fact that the TB and SMB evolved separately before their accretion along the AKSZ during the Pan-African Orogeny. The data obtained here combined with the structural data from Praharaj et al. (Reference Praharaj, Rekha and Bhattacharya2021) suggest that the northern block (SMB) subducted below the block (Trivandrum and Nagercoil blocks) south of it (Fig. 11). The initial loading and heating and later decompression P-T path obtained from both the SMB and TB also support this. In the TB, the cooling and decompression path might have developed as a result of the post-collisional extension and crustal thinning as suggested by Yang et al. (Reference Yang, Ganguly, Shaji, Dong and Nandakumar2019). Figure 11 depicts a modified tectonic model based on the new structural and metamorphic data. The present observation is in very good agreement with the tectonic model proposed by Rajesh et al. (Reference Rajesh, Arai, Satish-Kumar, Santosh and Tamura2013), except for the orientation of the shear zone, and also with that proposed by Naganjaneyulu & Santosh (Reference Naganjaneyulu and Santosh2011). Rajesh et al. (Reference Rajesh, Arai, Satish-Kumar, Santosh and Tamura2013) suggested S-wards subduction of the SMB based on geochemical studies from spinel dunite present in the accretionary complex, and the occurrence of ultramafic cumulates as dismembered units at several localities. Based on seismic reflection and MT data, Naganjaneyulu & Santosh (Reference Naganjaneyulu and Santosh2011) suggested a tectonic model in which the predominantly S-dipping seismic reflectors beneath the SMB define a prominent S-dipping lithological layering, explained by imbricate structures or mega duplexes developed during the S-wards subduction of an oceanic plate. The present model corroborates the S-wards dipping model and deviates from the previously proposed N-wards subduction model, that is, the SMB is the overriding plate proposed by Santosh et al. (Reference Santosh, Hua, He, Li, Tsunogae, Shaji and Indu2017) and Yang et al. (Reference Yang, Ganguly, Shaji, Dong and Nandakumar2019) based on geochemical data. The arc magmatism signatures along the southern margin of the SMB (Santosh et al. Reference Santosh, Hua, He, Li, Tsunogae, Shaji and Indu2017) and the mafic magmatism associated with post-collisional extension in the TB (Yang et al. Reference Yang, Ganguly, Shaji, Dong and Nandakumar2019) were the main reasons for the N-wards subduction model. The new data from the present study provide an insight into the subduction–accretion–collision tectonics associated with the late Neoproterozoic – Cambrian evolutionary history of this region.

Fig. 11.

Schematic plate tectonic model across the Achankovil Shear Zone (modified after Rajesh et al. Reference Rajesh, Arai, Satish-Kumar, Santosh and Tamura2013) illustrating Neoproterozoic–Cambrian tectonic evolution of the Southern Madurai and Trivandrum blocks. Structural evidence in support of the S-wards subduction model is incorporated in the diagram (details in Praharaj et al. Reference Praharaj, Rekha and Bhattacharya2021).