Abstract
The Southern Granulite Terrane of India exposes remnants of an interbanded sequence of orthoquartzite–metapelite–calcareous rocks across the enigmatic Palghat–Cauvery Shear Zone (PCSZ), which has been interpreted as a Pan-African terrane boundary representing the eastward extension of the Betsimisaraka Suture Zone of Madagascar. Zircon U–Pb geochronology of metasedimentary rocks from both sides of the PCSZ shows that the precursor sediments of these rocks were sourced from the Dharwar Craton and the adjoining parts of the Indian shield. The similarity of the provenance and the vestiges of Grenvillian-age orogenesis in some metasedimentary rocks contradict an interpretation that the PCSZ is a Pan-African terrane boundary. The lithological association and the likely basin formation age of the metasedimentary rocks of the Southern Granulite Terrane show remarkable similarity to the rock assemblage and timing of sedimentation of the Palaeoproterozoic to Neoproterozoic shallow-marine deposits of the Purana basins lying several hundred kilometres north of this terrane. Integrating the existing geological information, it is postulated that the shallow-marine sediments were deposited on a unified land-mass consisting of a large part of Madagascar and the Indian shield that existed before Neoproterozoic time, part of which was later involved in the Pan-African orogeny.
Supplementary material: Details of the zircon U–Pb LA-MC-ICP-MS analyses of samples are available at http://www.geolsoc.org.uk/SUP18793.
The Proterozoic sedimentary sequences consisting of sandstone, shale and limestone that cover a vast expanse of the crystalline basement of peninsular India are known as the Purana sedimentary rocks (Fig. 20.1, reviewed in Malone et al. 2008). Detailed facies analyses suggest that the Purana sediments were deposited in an intracontinental shallow-marine environment (Basu et al. 2008; Malone et al. 2008; Bickford et al. 2011a, b). U–Pb isotope analysis of detrital zircon grains in the clastic sediments suggests that the sediments of the Purana basins were sourced from the exhumed metamorphic rocks of the Indian shield (reviewed in Basu et al. 2008; Malone et al. 2008). The Southern Granulite Terrane (SGT) that lies several hundred kilometres south of the present-day exposure limit of the Purana rocks shows detached outcrops of interbanded quartzite, metapelite (sensu stricto) and marble+calc-silicate rocks (Fig. 20.2). The temporal and genetic relations of this metasedimentary rock ensemble, which is generally considered to be the metamorphosed equivalent of the Purana sandstone–shale–limestone sequence (reviewed in Sengupta et al. 2009), are not understood. As metamorphic overprints obliterated most of the sedimentary features, U–Pb isotope analyses of detrital zircon grains and their in situ overgrowths can provide valuable information on the provenance, timing of sedimentation and metamorphism of the sedimentary sequences of the SGT. This information may also help to establish whether these metasedimentary rocks are temporally and genetically related to the Purana sediments exposed further north. U–Pb isotope data from detrital zircon grains also help in tracing the loci of trans-continental orogenic belts and sutures that stitched together disparate continental fragments to form supercontinents (e.g. Collins et al. 2003; de Waele et al. 2011). The Mozambique suture, along which the supercontinent Gondwanaland may have been amalgamated, is a case in point. Based on contrasting sedimentary sources of detrital zircon populations in a suite of high-grade metapelitic rocks, Collins et al. (2003) argued that the Neoarchaean Antananarivo block and the Meso- to Neoarchaean Antongil block of Madagascar represent opposite sides of the Mozambique Ocean that closed during the Pan-African orogeny. The junction between the two blocks, the Betsimisaraka Suture Zone (BTSZ), has been extended to merge with the Palghat Cauvery Shear Zone (PCSZ) of the SGT (Collins et al. 2003, 2007a, b; see Fig. 20.1). One implicit assumption in this correlation is that the crustal blocks that lie to the north (the Salem Block, Clark et al. 2009a) and to the south (Madurai Block, Ghosh et al. 2004) of the PCSZ represent disparate crustal terranes that were juxtaposed only during the Late Neoproterozoic orogeny. Extant geological and geochronological studies present conflicting views on the status of the PCSZ. Vestiges of mafic–ultramafic rock associations were interpreted to represent slices of Neoproterozoic ocean crust (Sajeev et al. 2009; Yellappa et al. 2012). Inferred high-pressure to ‘eclogite’ facies metamorphism of late Neoproterozoic age (Sajeev et al. 2009) and crustal-scale shear deformation along the PCSZ (Chetty 1996) have been cited as supportive evidence of the geodynamic model proposed by Collins et al. (2003). On the other hand, continuation of 2.5 Ga old felsic crust across the so-called PCSZ (Braun & Kriegsman 2003; Ghosh et al. 2004; for review of references up to 2003; Bhaskar Rao et al. 2003; Brandt et al. 2014), lack of evidence in favour of crustal-scale shear deformation in many parts of the PCSZ (Mukhopadhyay et al. 2003) and the occurrence of middle Neoproterozoic plutonism and metamorphism on both sides of the PCSZ are in gross disagreement with proposed continuation of the BTSZ of Madagascar along the PCSZ. With this backdrop we present laser ablation-inductively coupled plasma mass spectrometry (LA-ICPMS) U–Pb isotope dates of zircon populations from three key rock samples from the SGT. Two of them are quartzose–metasedimentary rocks collected from the northern part of the PCSZ (Fig. 20.2). The third sample is a porphyritic granite that intruded and was subsequently deformed and metamorphosed together with an intercalated sequence of orthoquartzite–metapelite–metacarbonate lying to the south of the PCSZ (i.e. on the Madurai Block, Fig. 20.2). The geochronological data of the zircon populations from the three key rock samples are compared with zircon U–Pb ages reported from south of the PCSZ, the adjoining parts of the Indian shield and northern Madagascar. The motivation for this is (a) to test if the metasedimentary rocks within the SGT are temporally and genetically related to the Purana sediments lying further north, (b) to put constraints on the provenance of the sedimentary protoliths, the timing of formation and inversion of the sedimentary basins that developed across the PCSZ, (c) to test the hypothesis that the PCSZ is a Neoproterozoic terrane boundary and represents the eastward continuation of the BTSZ and (d) to compare the crustal architecture and age provinces of northern Madagascar with those within the SGT with the goal of understanding the disposition of different crustal domains in the Indo-Madagascar landmass prior to the Pan-African orogeny.
Reconstructed east Gondwanaland showing the positions and broad tectonic architecture of India, Madagascar, Sri Lanka and Antarctica (modified after Ghosh et al. 2004; de Waele et al. 2011; Turner et al. 2014). Neoproterozoic metamorphic episodes include one or both of Grenvillian and Pan-African metamorphism. MG, Marwar Group; ADFB, Aravalli–Delhi Fold Belt; VB, Vindhyan Basin; DB, Deccan Basalt; BuC, Bundelkhand Craton; ChB, Chhattisgarh Basin; SC, Singhbhum Craton; BaC, Bastar Craton; EDC, Eastern Dharwar Craton; WDC, Western Dharwar Craton; CG, Closepet Granite; EGP, Eastern Ghats Province; CB, Cuddapah Basin; KP, Krishna Province; SB, Salem Block; CSS, Cauvery Shear System; MB, Madurai Block; TB, Trivandrum Block; BM, Bemarivo domain; An, Antongil Craton; Ant, Antananarivo Craton; Ms, Masora Craton; Ik, Ikalamavony subdomain; A-M, Anaboriana–Manapotsy; (a) Itremo Group; (b) Sahantaha Group; (c) Maha Group; Vo, Vohibory domain; And, Androyen domain; Ano, Anoysen domain; WC, Wanni Complex; HC, Highland Complex; VC, Vijayan Complex; NC, Napier Complex; RC, Rayner Complex; LHC, Lützow–Holm-Bay Complex. (1) Moyar–Attur Shear Zone (MSZ); (2) Bhavani Shear Zone (BSZ); (3) Moyar–Bhavani Shear Zone (MBSZ); (4) Palghat–Cauvery Shear Zone (PCSZ); (5) Karur–Kambam–Painavu–Trichur Shear Zone (KKPTSZ); (6) Achankovil Shear Zone (ACSZ); (7) Central Indian Shear Zone (CITZ); (8) Son–Narmada South Fault; (9) Son– Narmada North Fault; (10) Betsimisaraka Suture Zone (BTSZ); (11) Ranotsara–Bongolava Shear Zone (RBSZ).
General geological map of southern peninsular India showing the inferred shear zones. Also shown in the inset map are sample locations of this study as well as those of Collins et al. (2007b) and Kooijman et al. (2011). This map is modified after Ghosh et al. (2004). (1) Nagercoil Hills; (2) Cardamom Hills; (3) Biligirirangan Hills; (4) Nilgiri Hills; (5) Shevaroy Hills; (6) Kollimalai Hills; (7) Palni Hills; (8) Coorg Hills. All other acronyms are as in Figure 20.1.
Geological background
The Precambrian shield of peninsular India consists of Palaeo- to Neoarchaean nuclei that are girdled by Proterozoic fold belts/granulite terranes (Fig. 20.1; reviewed in Naqvi & Rogers 1987; Ramakrishnan & Vaidyanathan 2008). A large part of the metamorphosed basements is covered by Meso- to Neoproterozoic unmetamorphosed sedimentary sequences consisting of sandstone–shale–carbonate associations which, in the Indian literature, are known as the rocks of Purana basins (Holland 1909). The Vindhyan, Chhattisgarh and Cuddapah basins are examples of the most extensive Purana basins (Fig. 20.1). Recent studies on several Purana basins have revealed that, barring a few parts where sedimentation continued until late Neoproterozoic time, sedimentation in these basins ended before 900 Ma (reviewed in Basu et al. 2008; Malone et al. 2008; Bickford et al. 2011a; Turner et al. 2014). U–Pb ages of detrital zircon grains in the clastic sediments of the Vindhyan and Chhattisgarh basins suggest that sediments were sourced from the adjoining metamorphic terranes of the Indian shield (Patranabis-Deb et al. 2007; Malone et al. 2008; Bickford et al. 2011a; Turner et al. 2014). Sedimentological attributes of the Purana sediments suggest their deposition in a shallow-marine environment (reviewed in Patranabis-Deb et al. 2007; Malone et al. 2008). The southern margin of the Cuddapah Basin defines the southernmost limit of the preserved Purana basins (Fig. 20.1). The SGT girdles the southern edge of the Palaeo- to the Mesoarchaean Western and Neoarchaean Eastern Dharwar Craton (Fig. 20.1). The geological history of the SGT is complex and is succinctly reviewed in a number of publications (Braun & Kriegsman 2003; Ghosh et al. 2004; Plavsa et al. 2012; Brandt et al. 2014). The major magmatic and metamorphic episodes in the SGT and the adjoining crustal domains of the Indian shield are graphically presented in Figure 20.3. On the basis of the extant geological information, the SGT can be divided into three roughly east–west elongated blocks (Fig. 20.2). In this communication the term ‘block’ is used in a loose sense to imply a segment of continental crust having characteristic geological features. These ‘blocks’ may or may not be bound by terrane boundaries or shear zones. The northernmost part, the Salem Block, is composed dominantly of Neoarchaean felsic orthogneisses (charnockite/enderbite/hornblende–biotite gneiss) and contains enclaves of mafic–ultramafic rocks and minor banded iron formation. The rocks of the Salem Block underwent high-grade metamorphism during the earliest Palaeoproterozoic time (Clark et al. 2009a; Peucat et al. 2013). The southernmost part is composite in nature and can be subdivided into the northern Madurai Block and the southern Trivandrum Block that are presumed to be separated by the Achankovil Shear Zone (Fig. 20.2; reviewed in Braun & Kriegsman 2003). In contrast to the Salem Block, the Madurai Block and the Trivandrum Block show abundant metasedimentary rocks that were affected by high-grade metamorphism during Late Neoproterozoic time (Ediacaran–Cambrian, reviewed in Braun & Kriegsman 2003; Ghosh et al. 2004; Collins et al. 2007b among others). A c. 300 km-long and 70 km-wide zone that separates the Salem Block from the Madurai Block is composed of a plethora of rocks that are akin to both the Salem Block and the Madurai Block and have a protracted magmatic, deformation and metamorphic history ranging from c. 2900 to c. 500 Ma (reviewed in Chetty 1996; Mukhopadhyay et al. 2003; Ghosh et al. 2004; Dutta et al. 2011; Plavsa et al. 2012; Brandt et al. 2014). It has been suggested in many studies that a number of crustal-scale shear zones (collectively termed the Cauvery Shear System, CSS) have dissected the rocks of this narrow belt (Fig. 20.2; reviewed in Chetty 1996; Dutta et al. 2011; Plavsa et al. 2012; Brandt et al. 2014). The PCSZ, which marks the southern boundary of the CSS, is perceived to be a terrane boundary along which the Madurai and Trivandrum blocks were amalgamated with the Salem Block, and remaining part of the CSS during the Pan-African orogeny (reviewed in Collins et al. 2007a, b; Yellappa et al. 2012). An interlayered suite of metasedimentary rocks consisting of orthoquartzite–metapelite–marble–calc-silicate rocks occurs as detached outcrops in a country of Neoarchaean to Palaeoproterozoic felsic orthogneisses (reviewed in Meissner et al. 2002; Ghosh et al. 2004; Collins et al. 2007a, b; Sengupta et al. 2009; Kooijman et al. 2011; Plavsa et al. 2012; Brandt et al. 2014). The metasedimentary rocks occur on both sides of the PCSZ (Fig. 20.2). U–Pb ages of zircon in a few metasedimentary rocks suggest that the latest deformation and metamorphism of the metasedimentary rocks occurred during Pan-African time (c. 650–520 Ma, Raith et al. 2010; Kooijman et al. 2011). Petrological studies on the metasedimentary rocks, albeit from only two localities near the PCSZ, demonstrate that the metasedimentary packets were buried to great depth (corresponding to >9 kbar) and were metamorphosed between 700 and 800 °C (Sengupta et al. 2009; Raith et al. 2010). The northernmost outcrop of the metasedimentary rocks of the SGT lies several hundred kilometres south of the unmetamorphosed sedimentary sequence of the Cuddapah Basin (Figs 20.1 & 20.2). In a reconstruction of Gondwanaland (Fig. 20.1), the Western Dharwar Craton and the SGT are juxtaposed against the Mesoarchaean Antongil–Masora and the Neoarchaean Antananarivo blocks of Madagascar, respectively. On the basis of contrasting sources of protoliths of the metasedimentary rocks overlying the Meso- and Neoarchaean blocks of Madagascar, it is assumed that a suture zone (BTSZ) exists between these two Archaean blocks (reviewed in Collins et al. 2003). The BTSZ is considered as a late Neoproterozoic suture zone along which the Mozambique Ocean might have been consumed (Collins et al. 2003, 2007a). Based on a few U–Pb ages of detrital zircon grains in metasedimentary rocks of the SGT, Collins et al. (2007b) postulated that the PCSZ represents the eastward continuation of the BTSZ (Fig. 20.1; reviewed in Collins et al. 2003). The idea proposed by Collins et al. (2007a) has been supported (Chetty 1996; Sajeev et al. 2009; Yellappa et al. 2012) and contradicted (Ghosh et al. 2004; Plavsa et al. 2012; Brandt et al. 2014) by subsequent studies. Lack of convincing field evidence in favour of crustal-scale shear zones in the eastern part of the PCSZ (Mukhopadhyay et al. 2003) also contradicts the model of Collins et al. (2007a).
Timing of magmatic and metamorphic events in the Cauvery Shear System, Madurai Block, Western Dharwar Craton, Eastern Dharwar Craton, Krishna Province and Eastern Ghats Province. References: (1) Ghosh et al. (2004); (2) Bhaskar Rao et al. (1996); (3) Clark et al. (2009b); (4) Upadhyay et al. (2006); (5) Kooijman et al. (2011); (6) Mohan et al. (2013); (7) Plavsa et al. (2012); (8) Raith et al. (1999); (9) Saitoh et al. (2011); (10) Santosh et al. (2012); (11) Teale et al. (2011); (12) Yellappa et al. (2012); (13) Sato et al. (2012); (14) Sato et al. (2011a); (15) Sato et al. (2011b); (16) Kovach et al. (1998); (17) Ramakrishnan & Vaidyanathan (2008); (18) Dobmeier & Raith (2003); (19) Bose et al. (2011); (20) Rekha et al. (2013).
Analytical methods, description of samples
Zircon grains were separated from the studied rock samples employing standard procedures (crushing, sieving, magnetic and heavy liquid separation). Representative zircon grains were then hand-picked using a binocular microscope. The separated zircon fractions included all sizes and morphologies and for each sample over 100 grains were selected and mounted on one-inch epoxy discs. The internal structures of the grains were documented by cathodoluminescence and back-scattered electron imaging using a Jeol 8900 JXA Superprobe. The U–Pb isotope ratios of the grains were measured using a Thermo-Fisher Element 2 sector field ICP-MS coupled to a New Wave UP193HE ArF excimer laser system at the Institut für Mineralogie, Westfälische Wilhelms-Universität, Münster, Germany. Details of the analytical techniques are given in Kooijman et al. (2012). Repeat in-run measurements of the 91500 standard zircon (Wiedenbeck et al. 1995) yielded an external reproducibility (2σ, n=20) of 1.5% for 206Pb/238U and 2.8% for 207Pb/206Pb. All uncertainties are reported at the 2σ level.
Three rocks, two metapelites and one porphyritic granite, were chosen for this study. The metapelite samples were collected close to the northern margin of the PCSZ (Fig. 20.2), whereas the porphyritic granite that intruded and co-folded with the orthoquartzite–pelite–carbonate unit of Alambadi was collected south of the PCSZ (Fig. 20.2). In the following section we present a brief description of the field features and highlight the morphology and zoning patterns of the zircon grains recovered from the three rocks.
Metapelite UNG64 (N11°27.960′E77°27.960′)
The sample was collected from a metasedimentary enclave within the virtually undeformed Pan-African (c. 560 Ma, U–Pb zircon date, Brandt et al. 2010) alkaline granite complex of Sankagiri (Fig. 20.4a). This is a garnet–biotite–kyanite schist that is profusely veined by quartz (Fig. 20.4b). Intercalated marble and calc-silicate rocks were deformed and metamorphosed together with the metapelite. Petrological study demonstrated that the metasedimentary sequence was buried to a depth corresponding to 9 kbar (700–750 °C), followed by uplift along a steeply decompressive pressure–temperature path (Sengupta et al. 2009). Contact metamorphism by the pegmatitic granite caused replacement of bladed kyanite by fibrous sillimanite (Sengupta et al. 2009).
Field features of the studied samples. (a) Enclave of metapelite (UNG64) within the pegmatoidal granite from Sankagiri. (b) Syn-metamorphic quartz veins in the kyanite–biotite–garnet-bearing metapelite enclave in sample UNG64. (c) Quartz-rich bands intercalated with pelitic schist showing F2 folds (sample N58). (d) Schistosity defined by biotite-rich layers in the pelitic layers of sample N58. Note the compositional banding parallel to schistosity. (e) Porphyritic granite with euhedral to subhedral megacrysts of K-feldspar set in a fine- grained matrix of quartz–biotite–plagioclase (A72). Note a raft of calc-silicate rock within the porphyritic granite. (f) Stretched K-feldspar augen defining a foliation that is folded by F2 fold in A72.
Morphology and zoning patterns of zircon
The zircon grains have rounded to irregular outlines but a few elongated grains (width–length ratio c. 1:2) show a subhedral to euhedral shape. The oscillatory zoning in most grains shows planar truncations, suggesting that they are detrital grains (Fig. 20.5a). Overgrowths on oscillatory zoned detrital grains show higher luminescence and a ‘patchy’ internal structure (Fig. 20.5a). They are very thin along the prisms but rather broad at the crystal ends. At the immediate interface with the detrital grains, very thin high and low luminescence zones are developed. Ragged interfaces indicate minor resorption of the detrital grains through dissolution by pore fluids. The patchy zoning of the rims argues against their crystallization from a melt for which the rock lacks evidence.
Cathodoluminescence (CL) images of zircon grains from the studied samples: (a) UNG64 (metapelite), (b) N58 (psammitic paragneiss) and (c) A72 (porphyritic granite). Numbers to the upper left of grains indicate grain number. Ages younger than 1000 Ma are cited as 206Pb/238U ages, whereas for older zircon domains, the 207Pb/206Pb ages are given.
Psammitic paragneiss N58 (N11°5.949′E78°8.964′)
This metasedimentary rock sample was collected from a roadside exposure near the northern bank of the Cauvery River (Fig. 20.2). A compositional banding is manifested by centimetre- to decimetre-thick quartz-rich (psammitic) bands that are intercalated with quartz-poor pelitic schist (Fig. 20.4c, d). In the psammitic layers, a prominent schistosity subparallel with the compositional banding is defined by preferred orientation of biotite. The schistosity and the compositional banding show outcrop-scale folds (Fig. 20.4c). Porphyroblasts of kyanite and garnet are developed within the metapelitic layers. The pelitic bands contain very thin (<1 cm-thick) leucosomes that never exceed more than 15 vol% of the rock.
Morphology and zoning patterns of zircon
Zircon grains show diverse morphologies and internal structures (Fig. 20.5b). Most of the grains are rounded to anhedral with a few grains showing euhedral prismatic shape. Core regions of both euhedral and anhedral grains show fine oscillatory zoning that is truncated by grain outline as well as highly luminescent, structureless, discontinuous rims (Fig. 20.5b). In a few euhedral grains, the oscillatory zoned cores are truncated and overgrown by thin low-luminescence mantles and the latter are truncated by high-luminescence rims (Fig. 20.5b). Featureless low-luminescence domains on high-luminescence cores (with or without oscillatory zoning) are present in a few grains. The morphology of the oscillatory and high-luminescence cores and their truncation by high-luminescence rims suggests that the cores are detrital grains on which the luminescent rims grew in situ during metamorphism.
Porphyritic granite A72 (N11°45.202′E78°3.428′)
The sample was collected from a porphyritic granite which intruded the folded metasedimentary (orthoquartzite–shale–carbonate) unit near Alambadi, south of the supposed PCSZ (Fig. 20.2). Rafts of calc-silicate rocks and quartzite in the granite further corroborate the intrusive relation with the metasedimentary unit (Fig. 20.4a). The granite and the metasedimentary country rocks were subsequently deformed and metamorphosed. A distinct L-S fabric and augen-structure was developed in the granite (Fig. 20.4e, f), although in low-strain domains a magmatic flow structure defined by euhedral K-feldspar phenocrysts in a finer-grained matrix of plagioclase, quartz and biotite may still be preserved. The L-S fabric is folded and the geometry of the folds is similar to the regional F2 folds defined by the adjoining metasedimentary package (Mukhopadhyay et al. 2003; our data).
Morphology and zoning patterns of zircon
Zircon grains exhibit a variety of morphology with most grains showing euhedral prismatic shapes (Fig. 20.5c). Aspect ratios of these grains are c. 3:1. The morphology of euhedral grains is consistent with their crystallization from a melt phase, presumably during magmatic crystallization of the protolith of the rock. The euhedral shape is less distinct and the aspect ratio varies considerably in other grains (Fig. 20.5c). Distinct oscillatory zoning patterns largely preserve the magmatic crystal shape and growth structures. A few grains show featureless low-luminescence cores. Highly luminescent and featureless overgrowths are restricted to the ends of the crystals and rarely truncate the oscillatory zoning (Fig. 20.5c).
Results
In this section 206Pb/238U and 207Pb/206Pb dates of the different textural domains of the representative zircon grains from each sample are presented. Details of the analyses are represented in Figure 20.6. The U–Pb data of zircon grains in each sample are plotted in concordia diagrams. Only concordant to near-concordant ages (<10% discordance) were considered for provenance analysis and bracketing the age of sedimentation. Furthermore, 207Pb/206Pb ages are used for spots that yielded ages older than 1000 Ma whereas 206Pb/238U ages are used for spots that yielded ages younger than 1000 Ma. The program Isoplot-3.00 (Ludwig 2003) was used for statistical analysis of the age data and construction of Concordia plots.
(a, b) Concordia diagrams showing the results of U–Pb analyses of zircon grains from metasedimentary rocks. Histograms plus probable peaks of the age cluster of the detrital zircon grains of <10% discordance are shown in the inset: (a) sample UNG64; (b) sample N58. (c) Concordia diagram showing the results of U–Pb analyses of zircon grains from the A-type granite sample A72. Histogram plus probable peaks of the age cluster of the crystallization age of zircon grains showing <10% discordance are shown in the inset. (d) The age histograms plus probability curve for all Pan-African metamorphic zircon rims from all the three samples UNG64, N58 and A72. Two distinct age populations with peaks at 572±2 and 639±4 Ma are evident.
Metapelite (sample UNG 64)
A total of 61 spots were analysed on 23 zircon grains; 40 out of 61 analyses show <10% discordance. The truncated oscillatory detrital grains show an age spectrum ranging from 2509 to 836 Ma (Fig. 20.5a). Near-concordant ages of detrital zircons cluster in two distinct populations: 2448±7 and 897±20 Ma (Fig. 20.6a). The ages of thin overgrowths that presumably date the in situ growth of metamorphic zircon fall within 633–487 Ma.
Psammitic paragneiss (sample N58)
In sample N58 69 spots were analysed on 34 grains. Out of 69 analyses, 46 dates show <10% discordance while the remaining analyses show 30–89% discordance. The detrital cores yield a spectrum of ages (3030–2230 Ma for <10% discordant grains; Figs 20.6 & 20.5b). Compared with the metapelite UNG 64, the detrital zircon population of this sample does not contain Neoproterozoic components and shows a tight cluster around 2395±6 Ma (81% of the analyses, Fig. 20.6b). Three-spot analyses of the highly luminescent rims yielded 206Pb/238U ages in the range of 567–525 Ma (two reversely concordant and one <10% discordant).
Porphyritic granite (sample A72)
A total of 81 spots were analysed on 27 grains out of which only eight analyses show >10% discordance and are not considered for discussion. The oscillatory zoned domains give ages tightly clustering between 790 and 760 Ma with an average of 768±3 Ma. The rare dark luminescent cores give slightly older dates (803–815 Ma). The three overgrowths on the oscillatory zoned domains give an average age of 570±19 Ma.
Discussion
Timing of sedimentary basin formation on the SGT and the temporal relation with the Purana basins of the Indian shield
Constraining the time of sedimentation of the metamorphosed orthoquartzite–shale–carbonate sequence is not straightforward. Nevertheless, the age of the oldest metamorphic event that affected the sediments and the youngest age of detrital zircon grains provide the temporal window within which the sedimentation and hence the basin formation took place (Collins et al. 2003, 2012; Kooijman et al. 2011; Sato et al. 2011a, b, 2012). Figure 20.7 presents the concordant to near-concordant ages of detrital and metamorphic zircons from two metasedimentary samples (UNG64 and N58) from the CSS. Also included in Figures 20.7 and 20.8 are published ages of metamorphic and detrital zircon grains from metasedimentary sequences of northern Madagascar, the CSS, the Madurai Block, the Trivandrum Block and the Vindhyan basin – the largest Purana basins in the Indian shield. The detrital and metamorphic zircon populations in sample UNG 64 indicate deposition of the sedimentary suite in the narrow time span between 836 and 633 Ma (Fig. 20.6a). This age bracket of sedimentary basin formation is in good agreement with the age bracket of sedimentation (689–513 Ma) of the protoliths of some metapelites from the Madurai and Trivandrum blocks (Fig. 20.7; Collins et al. 2007b). Owing to the paucity of younger detrital zircon grains, a wide time bracket for sedimentary basin formation is obtained from the zircon data of two samples that lie close to the PCSZ (N58: 2280–567 Ma; Ayy-1, Raith et al. 2010: 1992–615 Ma). The four metasedimentary samples from the Madurai Block analysed by Kooijman et al. (2011) suggest that sedimentation in these areas occurred within a broad time span of c. 1750–990 Ma (Fig. 20.7). A similar wide time bracket for sedimentary basin formation (1900–515 Ma) is also reported by Collins et al. (2007b) from the Trivandrum Block (Fig. 20.7). All of these observations support the view that sedimentation of the protoliths of the metasedimentary rocks that developed on the Archaean/Palaeoproterozoic gneissic basement of the CSS and Madurai Block occurred over a protracted period of time. The intercalated sequence of quartzite–metapelite–metacarbonate rocks that are exposed in the CSS and Madurai Block is likely to represent the metamorphosed equivalents of platformal shallow-marine sandstone–shale–carbonate suites. The precursor sediments of these sequences, therefore, bear a resemblance to the Purana basin fills in terms of broad depositional environment, the age populations of detrital zircons and the timing of sedimentary basin formation (see Figs 20.6 & 20.7). The southernmost exposure limit of the Purana basins, the Cuddapah Basin and its equivalents occurs several hundred kilometres north of the northernmost exposure limits of the Proterozoic metasedimentary rocks of the SGT (Fig. 20.1). It seems, therefore, likely that the Purana basins extended further south, at least up to Tirunelveli, where exposures of interbanded quartzite, metapelite and marble are widespread (Fig. 20.7). In this scenario, the southern part of the vast shallow-marine intracontinental deposits of the extended Purana basins were deformed and metamorphosed during the Pan-African orogenesis (cf. Sengupta et al. 2009; Raith et al. 2010; Kooijman et al. 2011; Sato et al. 2011a; Brandt et al. 2014).
Zircon U–Pb geochronology of detrital zircon and the metamorphic overgrowths in the SGT, Vindhyan (Purana) basins and central and northern Madagascar. References: (1) Raith et al. (2010); (2) Kooijman et al. (2011); (3) Collins et al. (2007b); (4) Brandt et al. (2011); (5) Turner et al. (2014); (6) Cox et al. (2004); (7) Cox et al. (1998); (8) de Waele et al. (2011); (9) Tucker et al. (2007, 2011); (10) Collins et al. (2003); (11) Fitzsimons & Hulscher (2005).
Geological and geochronological attributes of the crystalline basement and sedimentary cover in the Indo-Madagascar land mass. Note that the area bound by thick dashed lines in the Indo-Madagascar land mass shows remarkable similarity in terms of geological evolution at least from 2.5 Ga. Because of lack of information on rocks lying south of that line the sign ‘??’ is used. All the abbreviations are the same as Figure 20.1. References: (1) de Waele et al. (2011); (2) Cox et al. (2004); (3) Collins et al. (2007a, b); (4) Kooijman et al. (2011); (5) this study; (6) Plavsa et al. (2012); (7) Ghosh et al. (2004); (8) Raith et al. (1999); (9) Braun & Kriegsman (2003); (10) Collins et al. (2003); (11) Sato et al. (2011a, b); (12) Teale et al. (2011).
Provenance of precursor sediments of the studied metasedimentary rocks
Although the inferred precursors of the metasedimentary units that occur across the PCSZ resemble the stable shelf deposits of the Purana basin fills, superimposed deformation and metamorphism have obliterated most sedimentary structures (e.g. palaeoslope markers) of the former rocks. Nevertheless an attempt has been made to identify the likely hinterlands of the metasedimentary assemblage integrating the U–Pb/Pb–Pb ages of the detrital zircon grains and geochronological information of the crystalline basement rocks from the adjoining parts of the Indian shield and from the continents that were once contiguous within the reconstructed Gondwanaland (Fig. 20.3). Concordant and nearly concordant 206Pb/238U and 207Pb/206Pb ages of detrital zircon populations in the spatially adjoining samples UNG64 and N58 can be divided into four age clusters: 3030–2600, 2550–2450, 2400–2200 and 1092–836 Ma. Zircon grains of the oldest population, although few in number and restricted to sample N58, are likely to have their source in the Western Dharwar Craton and the CSS where rocks of this age are reported (Fig. 20.3). The Vindhyan sedimentary sequence is considered to be representative of the Purana basins and it contains detrital zircons that exhibit ages ranging from c. 3.3 to 1.0 Ga. Therefore, the possibility cannot be completely ruled out that these oldest zircon populations, at least part of them, could also have been derived from recycling of the Purana sedimentary sequences. It is intriguing that sample UNG64, which marks the northernmost outcrop of the quartzite–metapelite–marble ensemble in the SGT, does not contain any contribution from the Western Dharwar Craton (or from Purana sedimentary rocks), whereas the quartzite sample MR-48 of Kooijman et al. (2011) that comes from an area south of the PCSZ has this older component. Sampling bias and varied drainage patterns may explain this diversity of detrital zircon population. Detrital zircons of the 2550–2450 Ma age cluster are most likely derived from the Eastern Dharwar Craton and the CSS where rocks of this age range are reported (Fig. 20.3). Identification of the source of zircon grains of the 2400–2200 Ma age cluster is not straightforward. Extant geochronological information, although sparse, suggests that, barring one locality from Rajasthan, where 2240 Ma old felsic rocks are found (Khairmalia felsites; Malone et al. 2008), evidence for 2400–2200 Ma old igneous or partial melting events that led to zircon growth are hitherto not reported from the Indian shield. Intriguingly, 2400–2200 Ma-old detrital zircons are abundant in the Vindhyan sedimentary sequence that covers the northern Indian shield (Turner et al. 2014; Figs 20.1 & 20.6). This observation raises three possibilities: (1) the zircon grains in the Vindhyan sequence are derived from the Aravalli Craton; (2) rocks of this age range have not yet been dated from the Indian shield; or (3) rocks of this age range are lying beneath the Phanerozoic volcanic and sedimentary rock cover (see Fig. 20.2). In recent communications, de Waele et al. (2011) and Tucker et al. (2011) argued that c. 2300–2200 Ma-old detrital zircon grains that are present in the metasedimentary rocks of northern Madagascar were sourced from the Aravalli Craton. It is, therefore, possible that the detrital zircon grains from the third cluster could have been sourced from the Aravalli Craton, although recycling of Purana sedimentary components remains an alternative possibility. Evidently, a more detailed study is warranted to support or reject the other two possibilities. The Neoproterozoic detrital zircon population defining the fourth age cluster (1092–836 Ma) could have been derived from several areas within India and adjoining Sri Lanka, namely the Madurai Block, the Wanni and Vijayan Complexes of Sri Lanka, and the Eastern Ghats Belt, where crustal growth during this period has been documented (Fig. 20.3 and the references cited therein). It is, however, not known if all of these terranes were exhumed at this time to provide detrital components of the sedimentary precursor of sample UNG64.
Timing of metamorphism
Concordant to near-concordant (<10% discordance) ages of zircon rims that developed on detrital (in samples UNG64 and N58) or magmatic zircon grains (sample A72) are likely to date the metamorphism that affected these rocks. In view of overlapping 206Pb/238U ratios in zircon rims in all the three samples, the data are combined for statistical analysis. The data define two distinct age populations that fall within tight clusters at 639±4 and 527±2 Ma (Fig. 20.6d), supporting the view that the metasedimentary rocks and the porphyritic granitoids underwent two distinct metamorphic episodes. Two phases of metamorphism in the SGT across the Ediacaran–Cambrian boundary are also reported by Raith et al. (2010; 615±11 and 529±9 Ma) and Kooijman et al. (2011; clustering at 590 and 520 Ma). The older phase of metamorphism is also supported by the occurrence of the metapelite (sample UNG64) as enclave within c. 560 Ma-old alkaline granite in the Sankagiri area, and the thermal effects caused by the granite host (Brandt et al. 2010). In view of this observation, the 639±4 Ma age is interpreted to date the metamorphism and coeval deformation of the garnet+kyanite-bearing metapelite enclave (sample UNG64, Sengupta et al. 2009). Effects of Ediacaran or Cambrian metamorphic events are also widespread throughout the SGT (reviewed in Braun & Kriegsman 2003; Ghosh et al. 2004; Santosh et al. 2006; Collins et al. 2007a; Clark et al. 2009b; Kooijman et al. 2011; Brandt et al. 2014).
Can the PCSZ be a Neoproterozoic terrane boundary?
Several recent studies have argued that the PCSZ represents a Neoproterozoic terrane boundary along which the southern Madurai Block was welded to the northern Salem Block during the Pan-African orogenesis (e.g. Collins et al. 2003; Sajeev et al. 2009; Yellappa et al. 2012). In the context of this suggestion the following observations are relevant:
Orthoquartzite–shale–carbonate sequences of shallow-marine origin occur as deformed and metamorphosed keels in the orthogneiss basement on both sides of the PCSZ.
Age populations of detrital zircons in quartzites and metapelites indicate a similar provenance of the clastic material, with a substantial contribution from the Western Dharwar Craton (and/or Purana sedimentary sequence).
The c. 955 Ma metamorphism of the metasedimentary packages exposed south of the supposed PCSZ suggests that sedimentation was completed before this time. This then supports the interpretation that the Madurai Block, which hosts these metasedimentary rocks, was connected to the Dharwar Craton (the source of >3.3 Ga old detrital zircon grains) before c. 955 Ma (Fig. 20.8).
The 800–700 Ma old magmatic rocks (A-type porphyritic granite, massif-type anorthosite, carbonatite–alkaline rocks, reviewed in Braun & Kriegsman 2003; Ghosh et al. 2004; Brandt et al. 2014) that represent magmatism in an extensional setting are present on both sides of the PCSZ (Fig. 20.8). Taken together these observations contradict the interpretation that the PCSZ is a Neoproterozoic terrane boundary. Our study, therefore, supports the view of Mukhopadhyay et al. (2003), Ghosh et al. (2004) and Plavsa et al. (2012) that no significant break in terms of geological and geochronological characters exists across the PCSZ.
Implications of this study in the context of the Indo-Madagascar connection
In the reconstructed Gondwanaland the Indian shield is juxtaposed against Madagascar (Fig. 20.8). In this reconstruction the Palaeo- to Mesoarchaean Western Dharwar Craton faces the Antongil–Masora Block of the same age, whereas the Salem Block, the CSS and the Madurai Block face the Neoarchaean Antananarivo Block of east Madagascar (Fig. 20.8, reviewed in Rekha et al. 2013). Highlighting contrasted age populations of detrital zircons in metasedimentary units lying across the BTSZ (>3300 Ma ages in the eastern part and 2300–1800 Ma ages in the western part of the BSTZ), Collins et al. (2003) considered the BTSZ to represent a fossil ‘suture zone’ along which the Mozambique Ocean closed during the Pan-African orogeny. The implication of this model is that the western Antananarivo Block and eastern Mesoarchaean Antongil–Masora Block represent two sides of the Mozambique Ocean and the two blocks were juxtaposed only during the late Neoproterozoic time. The model of Collins et al. (2003), however, has been recently challenged by Tucker et al. (2011) and de Waele et al. (2011) on the grounds that Mesoarchaean detrital zircon populations are also present in the metasediments that developed on the western Antananarivo Block (the Itremo unit). According to these authors a coherent landmass, including the Antongil–Masora–Antananarivo blocks, existed well before Neoproterozoic time. This observation thus contradicts the interpretation that the BTSZ represents a Neoproterozoic terrane boundary. The Antananarivo Block of Madagascar shares many geological features with the crust exposed in the Salem Block, the CSS and the Madurai Block (Fig. 20.8). These include:
A c. 2.5–2.6 Ga orthogneiss basement is common to all of these terranes (Ghosh et al. 2004; de Waele et al. 2011; Plavsa et al. 2012; Brandt et al. 2014).
The Neoarchaean crust of the Antananarivo Block, the CSS and the Madurai Block was covered by shallow-marine sandstone–shale–carbonate sequences that were deposited possibly in two cycles, between 1800–800 and 800–620 Ma (Figs 20.7 & 20.8). Part of the detrital components in these sediments (now deformed and metamorphosed) was sourced from a unified Mesoarchaean landmass (Western Dharwar Craton–Antongil–Masora) lying to the north (Fig. 20.8).
c. 2300–1800 Ma detrital zircon populations are common in the sedimentary sequences that developed on the gneissic basement of the Antananarivo Block, the CSS (samples UNG64 and N58) and the Madurai Block (Fig. 20.7).
c. 800–700 Ma rift-related magmatism (A-type granitoids, alkaline rocks and massif-type anorthosite) and associated thermal metamorphism are common to all of the blocks (Collins et al. 2003; Ghosh et al. 2004; de Waele et al. 2011; Kooijman et al. 2011; Brandt et al. 2014; and this study).
The Proterozoic basement, the sedimentary cover rocks and the c. 700–800 Ma magmatic rocks of the Antananarivo Block, the CSS and the Madurai Block became intensely deformed and metamorphosed during the c. 600–500 Ma Ediacaran/Cambrian tectonothermal events (Figs 20.7 & 20.8).
In view of the marked similarity in geological histories of northern Madagascar and the SGT, it seems likely that the Precambrian rocks of the Antananarivo Block, the Salem Block, the CSS and Madurai Block formed a unified landmass (Greater Dharwar Craton) at least before 800 Ma. If the c. 955 Ma metamorphic impressions on some metasedimentary units of the Madurai Block are considered, this unified Indo-Madagascar landmass seems to have existed before early Neoproterozoic time. The presence of c. 2300–1800 and >3000 Ma-old detrital zircon populations in the metasedimentary units of the SGT as well as northern Madagascar supports the proposition of Tucker et al. (2011, 2014) and de Waele et al. (2011) that the Western Dharwar Craton and adjoining parts of the Indian shield provided detritus for the Proterozoic sedimentary basins developed on the unified Indo-Madagascar landmass. If this model remains valid, a large part of the Neo- and Meso-Archaean crusts of central and northern Madagascar (at least up to the Ranotsara–Bongolava Shear Zone, Fig. 20.8) and the SGT of India would represent one continuous strand line of the perceived Mozambique Ocean. The southern limit of this shallow-marine sedimentary sequence should be marked by metasedimentary rocks where the ‘detrital footprint’ of Mesoarchaean rocks would be recognized (Fig. 20.8). At present this signature is traceable up to the Sirumalai hills in the Madurai Block (Fig. 20.8). Geological evidence for the closure of this ocean and, hence, the location of the late Neoproterozoic suture should be searched in areas lying further south of this unified crustal block. Detailed geological and geochronological studies on rocks from the southern part of the Madurai Block and the Trivandrum Block are required to ascertain if the Mozambique Ocean suture passed through the SGT.
Acknowledgments
P. S., M. T., P. C. and S. S. acknowledge the financial support from the Council of Scientific and Industrial Research New Delhi. S.S. and P.S. also acknowledge the financial assistance from the Center of Advance Studies, Department of Geological Sciences, Jadavpur University and University Potential for Excellence (Phase II), Jadavpur University. D. M. acknowledges support through the Honorary Scientist Project of the Indian National Science Academy. The fieldwork of M. R. was financially supported through RWE-grant PN 71734. E. K. and the analytical work were supported by a Leibniz Award (DFG, Germany) to K. M. We thank Dr U. K. Bhui (Pandit Deendayal Petroleum University, India) and Dr U. Dutta (Indian School of Mines, India) for their help in the field and many stimulating discussions. We sincerely thank Professor A. Hofmann and an anonymous reviewer for their critical but very constructive comments, which have improved the clarity of the manuscript. We also thank Professor R. Mazumder for inviting us to contribute to this Geological Society of London Memoir and for his editorial remarks.
- © 2015 The Geological Society of London