Geochemistry of metasedimentary rocks of the Sonakhan and Mahakoshal greenstone belts, Central India: Implications for paleoweathering, paleogeography and mechanisms of greenstone belt development

Hamidullah Wani • M. E. A. Mondal • Iftikhar Ahmad

Abstract A comparative study of the Precambrian Sonakhan (SGB) and Mahakoshal (MGB)greenstones belts of Central India has been undertaken to decipher their provenance, paleoweathering, paleogeography, and tectonics. As compared to the Upper Continental Crust (UCC), the MGB samples are enriched while the SGB samples are depleted in mafic elements indicating the presence of mafic rocks in the source of the MGB. This is complemented by the Ni–Cr diagram. The REE concentrations, LREE fractionated patterns and negative Eu anomalies of the MGB and SGB samples indicate derivation of sediments from a highly fractionated granitic source. Since MGB samples also contain the geochemical signature of mafic rocks, it is, therefore proposed that the MGB clastic load were derived from two sources(mafic + felsic) with arc character. This is attested by Cr and Zr relationships, and LILE enrichment, and HFSE depletion. These features suggest that the SGB developed as autochthonous while the MGB developed as an allochthonous belt. The chemical alteration indices such as chemical index of alteration (CIA), plagioclase index of alteration(PIA),and index of compositional variability for MGB samples indicate that they were dominantly derived as the first cycle (with minor recycled) sediments from bimodal sources (dominantly continental arcs) by intense chemical weathering as compared to the SGB samples,which were derived from felsic sources (dominantly cratonic rocks), and partly by recycling through a low chemical weathering. The CIA and PIA values of the samples reveal a change in the climatic conditions from Late Archean to Late Paleoproterozoic. Such change is interpreted in terms of migration of the Indian plate from high latitudes in the Late Archean to lower latitudes during the Late Paleoproterozoic. This is consistent with the paleomagnetic data that placed India in the configuration of 2.45 Ga Ur and 1.78 Ga Columbia supercontinents.

Keywords Central Indian Shield · Sonakhan and Mahakoshal greenstone belts · Paleoweathering and paleogeography · Greenstone belt development

1 Introduction

Precambrian greenstone belts are archives of the Earth’s early lithospheric history (DeWit and Ashwal 1995);hence,their study is important for understanding the crustmantle processes, and the tectonic settings that were prevalent in the Precambrian. Besides their tectonic significance,greenstone belts are also economically important as a repository for syngenetic mineralization, and epigenetic deposits (Gally et al. 2007). The greenstone belts have complex geological histories, where rocks belonging to multiple tectonic histories occur in close association with one another. The greenstone belts are made up of interlayered volcanic ultramafic–mafic rocks (komatiites,komatiitic basalts,and tholeiites) that constitute 10%–50%of the belt; and sedimentary rocks (turbidites, graywackes,meta-argillites, quartzites, cherts, and banded iron formations, which are intruded by tonalite-trondhjemite-granodiorite plutonic rocks (TTG) (DeWit and Ashwal 1995;Hunter and Stowe 1997). The greenstone belts are generally metamorphosed to greenschist to amphibolite facies and are strongly folded and sheared (Bedard et al 2013).The origin of greenstone basins remains unclear, but it is believed that magmatism had played a major role in their evolution (DeWit and Ashwal 1995; Hunter and Stowe 1997). However, it is not certain whether this magmatism was plate tectonic driven or plate independent. According to Condie (1981) and De Witt and Ashwal (1995), greenstone belts are zones of variably metamorphosed ultramafic to mafic, some also felsic volcanic rock sequences with associated sedimentary rocks that occur within Archean and Proterozoic cratons between granite and gneiss bodies.Those supracrustal rocks have been interpreted as having formed at ancient oceanic spreading centers and island arcs. These belts are a relict witness of the early Earth crustal history, and each belt has its own geological,structural, and stratigraphic characteristics, but all are economically important due to their Au, Cu, Zn, and Pb potential wealth. The Indian peninsula contains several greenstone belts, thus providing the opportunity to understand their evolution. The paper seeks the comparison of the geochemistry of meta-sedimentary rocks that occur in the Sonakhan and Mahakoshal greenstone belts of neighboring cratons of the Central Indian Shield to understand whether their origin is similar or not. The metaclastic sediments of the greenstone belts across the world, especially in the Indian peninsula have not been studied in detail and compared to unravel the mechanism of their genesis.The Indian peninsula contains several Precambrian greenstone belts.Those belts have the potential to enhance the global database on the Precambrian greenstone development, and to provide valuable information regarding their provenances, weathering history, and paleogeography. The geochemistry of the clastic rocks of the greenstone belts has not been studied as much as that of volcanic rocks to understand their genesis. In this paper, an attempt has been made to compare the geochemistry of the clastic metasedimentary rocks of the Late Archean to the Early Paleoproterozoic Sonakhan greenstone belt (hereinafter referred to as SGB) of the Bastar craton and the Late Paleoproterozoic Mahakoshal greenstone belt (hereinafter referred to as MGB) of the Bundelkhand craton to understand the origin and evolution (Fig. 1). Geochemistry and mineralogy of clastic rocks have been applied to determine the compositions of the rocks in the source area (Taylor and McLennan 1985, McLennan et al.1993; Wani and Mondal 2010; Balestra et al. 2019; Corrado et al. 2019),infer the paleoclimate and weathering processes that affected the continental landmasses (Nesbitt and Young 1982; Fedo et al. 1995; Aldega et al. 2020), evaluate the secondary processes such as hydraulic sorting (McLennan et al.1993),delineate the tectonic setting of the sedimentary basin (Bhatia 1983; Bhatia and Crook 1986; Wani and Mondal 2011) and trace the evolutionary history of crust and mantle (Condie 1993).

Fig. 1 Map showing the distribution of the Proterozoic basins including the Mahakoshal and Sonakhan greenstone belts on the Indian craton.Proterozoic basins shown are:Ampani(A),Bhima(B),Cuddapah (C), Chattisgarh (Ch), Indravati (I), Kaladgi (K), Khariar(Kh), Pranhita–Godavari (PG), and Vindhyan (V). CITZ = Central Indian Tectonic Zone; NNSL = North Narmada-Son Lineament(Fault); SNSL = South Narmada-Son Lineament (Fault); PGR-Pranhita–Godavari rift; CR-Cambay rift; MR-Mahanadi rift

2 Geological setting

The Central Indian shield is the amalgamation between the northerly Bundelkhand block and the southerly Bastar block, which are separated by the east–west trending Narmada-Son lineament or Central Indian Tectonic Zone(CITZ) (Fig. 1). The SGB, which occurs in the Bastar craton,covers an area of about 1200 km2extending NNWSSE for about 40 km from Sonakhan (21°23′35′′N,82°48′55′′E) in the north, to Remra (21°17′N: 82°46′E)in the south,and having a maximum width of 40 km in the central part (Fig. 1). The Sonakhan Group (SGB) comprises a bimodal volcanic-sedimentary sequence that unconformably overlies the basement gneissic complex locally known as the Baya Gneiss.In the north and east,the entire greenstone succession is covered by the Meso-Neoproterozoic Chhattisgarh Basin. The Sonakhan Group has been divided, from base to top into three formations(Das et al. 1990): (1) Baghmara Formation, (2) Arjuni Formation, and (3) Bilari Formation (Fig. 2, Table 1). The Baghmara Formation consists dominantly of meta-ultramafites, meta-basalt, metagabbro, pyroclastics of intermediate to basic composition, ignimbrite, rhyolite, acid tuff,pebbly tremolite-actinolite schist, and banded iron formation. In the Baghmara Formation, larger bodies of sills trending NW–SE occur. The Arjuni Formation constitutes a thick sedimentary pile interspersed with minor volcanics and overlies the Baghmara Formation. The polymictic conglomerate member of the Arjuni Formation is called the Jonk conglomerate (Das et al. 1990). The Jonk Conglomerate is dominantly clast-supported and has been suggested to be of fluvio-glacial origin. It contains mixtures of boulders and large-sized pebbles and blocks of granite,gneiss, acid volcanic rocks, porphyries, amphibolite,metabasalt, quartzite, vein quartz, BIF, jasper, schist, and phyllite. The overlying Bilari Formation includes both felsic and mafic intrusive and extrusive bodies,and like the Arangi mafic volcanics,meta-basalt,and pyroclastics along with rhyolite.The gneissic rocks,which form the basement for the SGB,have ages ranging from 3.5 to 2.6 Ga(Sarkar et al.1990a,b).The granitoids that intrude the gneisses and the greenstone sequence yielded ages ranging from 2.6 to 1.5 Ga (Sarkar et al. 1990a). Thus, the SGB is considered to be Late Archean to Early Paleoproterozoic in age. This inference is further supported by the stratigraphic position as the SGB is overlain by the Mesoproterozoic Chhattisgarh Supergroup (1.4-1.0 Ga; Patranabis-Deb et al. 2007;Das et al. 2009; Bickford et al. 2011).

The MGB has an ENE–WSW direction with a length of 600 km that stretches from Barmanghat to Rihand Dam(Fig. 1). In the north, the MGB is separated from the Vindhyan basin by the Son Narmada North Fault (SNNF),and its southern boundary is marked by the Son Narmada South Fault (SNSF), which separates the MGB from the Proterozoic granites of the CITZ (Ramakrishnan and Vaidyanadhan 2008).The major geological units within the CITZ include metamorphosed supracrustal belts, metamorphosed mafic and ultramafic rocks, metacarbonates,iron, and manganese formations, TTG gneisses,charnockites, and related arc magmatic suites, exhumed high-pressure and ultrahigh-temperature metamorphic belts, and post-collisional K-rich granites(Acharyya 2003;Bhandari et al. 2011). The close association of metamorphosed mafic and ultramafic rocks, together with belts of iron and manganese formations in the CITZ, suggests a typical subduction–accretion setting where the Bastarcraton is believed to have been subducted under the hinterland of Bundlekhand craton similar to those described from present-day active continental margins (Nakagawa et al. 2009). MGB is made up dominantly of quartzites,carbonatites, chert, banded iron formations, greywacke–argillite, and mafic volcanic rocks. The regional metamorphic grade of MGB rocks is generally between greenschist to amphibolite facies (P = 8 kPa, T = 520 °C;Deshmukh et al. 2017). The belt is divided, from base to top, into three distinct units: the Saleemanabad Formation,Parsoi Formation, and Dudhmaniya Formation (Roy and Devarajan 2000; Fig. 2; Table 2). The mafic–ultramafic and alkaline igneous rocks that are exposed in the MGB are intruded within the Saleemanabad Formation. However,the Parsoi and the Dudhmaniya formations do not contain intrusive rocks. The MGB has been affected by three phases of deformation with an overall ENE–WSW trend.The first phase is characterized by upright isoclinal folds with steep southward dipping axial planes and occurred between 1800 and 1900 Ma (Deshmukh et al. 2017). The second phase occurred between 1700 and 1800 Ma and shows vertical to inclined, E–W striking folds with axial planes dipping to the south and a pronounced crenulation cleavage(Deshmukh et al.2017).The third phase occurred between 1700 and 1600 Ma and is marked by broad folds and N–S striking axial planes (Ramakrishnan and Vaidyanadhan 2008; Deshmukh et al. 2017). The duration of each was about 100 Ma. The granitic rocks of the TTG complex yield Rb–Sr ages up to ca.2200 Ma(Sarkar et al.1995). Since the TTG complex is the basement for the MGB, therefore, its supracrustals and intrusions are younger than 2200 Ma. The intrusive granites like the Jhigradandi pluton and the granitoids present in the eastern part of the MGB have been dated between 1813 ± 65 and 1709 ± 40 Ma (Pandey et al. 1998). Hence, the MGB is considered to be Late Palaeoproterozoic in age.

Table 1 Stratigraphic succession of the Sonakhan greenstone belt(Saha et al. 2000)

3 Methodology

Twenty-six fine to medium-grained metaclastic rock samples, consisting of eleven from the SGB and fifteen from the MGB were selected for major, trace, and rare earthelement analysis. Approximately 1 kg of each fresh rock sample was crushed and pulverized to a fine powder (200 mesh size) in an agate mill. Major elements of the SGB samples were analyzed on SIEMENS-SRS-3000 sequential X-Ray Fluorescence Spectrometer (XRF) on fused discs and fixed with polyvinyl alcohol at the Wadia Institute of Himalayan Geology(WIHG),Dehra Dun,India.The major elements of the MGB samples were analyzed on a Philips Magi XPRO PW2440, using pressed powder discs pasted with polyvinyl alcohol at the National Geophysical Research Institute (NGRI), Hyderabad, and calculated in weight percent(wt.%).Details of the analytical techniques along with precision and accuracy are described by Saini et al. (1998). Trace elements including rare earth elements(REE) of both the MGB and SGB samples were analyzed at the National Geophysical Research Institute (NGRI),Hyderabad by inductively coupled plasma mass spectrometer (ICP-MS, Perkin Elmer SCIEXELAN DRC II)and calculated in ppm (parts per million). Details of the analytical techniques along with precision and accuracy are described by Roy et al.(2007).The analytical precision and accuracy of the major oxide data are better than 1.5%.The precisions achieved for ICP-MS analyses were ＜5%Relative Standard Deviation with comparable levels of accuracy. International standards were used for calibration and testing accuracy.In this study,we have synthesized the major and trace element data of the metasedimentary samples from the studies by Wani and Mondal (2016) and Mondal et al. (2018) in addition to new data.

Table 2 Stratigraphy of the Mahakoshal greenstone belt (Roy and Devarajan 2000)

4 Results

4.1 Mineralogy

Since the MGB rocks have undergone greenschist to amphibolite facies metamorphism, therefore, they do not reflect the original mineralogy of the studied samples. The metaclastic rocks of MGB are composed of quartz, chlorite, biotite, epidote, and opaque phases (Fig. 3a, b). The SGB metaclastic rocks are composed of quartz, chlorite,and muscovite. However, some of the samples contain well-preserved original minerals. These samples are arkosic and contain clasts of quartz,fresh plagioclase,and a few rock fragments of metamorphic origin. The plagioclase grains make up about 20% of the rock (Fig. 3c, d).

4.2 Major element geochemistry

The major elements and trace elements including the rare earth elements (REEs) data of the samples are given in Table 3. The major element composition of the SGB and MGB samples is quite variable. In general, samples of the MGB are characterized by lower mean concentrations of CaO (0.54 wt.%) and Na2O (0.18 wt.%), and higherconcentrations of Fe2O3t(6.62 wt.%), MnO (0.08 wt.%),TiO2(0.85 wt.%), MgO (5.44 wt.%), K2O (4.11 wt.%) as compared to the SGB samples (CaO = 1.37 wt.%, Na2-O = 2.49 wt.%, Fe2O3t= 4.14 wt.%, MnO = 0.05 wt.%,TiO2=0.44 wt.%, MgO = 1.5 wt.%, K2O = 3.63 wt.%;Table 3). However, the SiO2, Al2O3,and P2O5concentrations do not show large variations between the MGB samples (SiO2= 63.44, Al2O3= 15.6, P2O5= 0.12 wt.%)and SGB samples (SiO2= 66.41, Al2O3= 15.5,P2O5= 0.17 wt.%). The difference in major-element compositions between the MGB and SGB samples has proved to be significant for MgO, Na2O, TiO2,and P2O5using the Student’s t-test at better than 95% confidence level (p ＜0.05) (Table 3).

Table 3 Major and trace elements of SGB and MGB samples

Table 3 continued

Table 3 continued

Table 3 continued

4.3 Trace elements

The mean values of transitional elements of the MGB samples have higher concentrations of Sc (11.73 ppm), V(103.6 ppm),Cr(82.3 ppm)and Ni(53.12 ppm),but lower of Co (11.2 ppm) in comparison to SGB (Sc = 7.33 ppm,V = 43.45 ppm, Cr = 21.85 ppm, Ni = 9.67 ppm, Co =17.59 ppm;Table 3).This difference in transition element concentrations between the MGB and SGB samples has proved to be significant for V, Cr, Ni, and Co using the Student’s t-test at a better than 95% confidence level(p ＜0.05)(Table 3).As compared to the SGB samples,the MGB samples have higher concentrations of LILEs like Cs(6.98 ppm) and Pb (24 ppm) but have lower of Sr(47.02 ppm), Ba (414.9 ppm), Th (16.5 ppm), and U(4.1 ppm). The Rb (121.9 ppm) concentrations are almost similar between the SGB and the MGB samples. HFSEs such as Zr(8941 ppm) and Hf (172.9 ppm)concentrations are higher in the MGB samples while Y (22.65 ppm), Nb(11.16 ppm), and Ta (1.53 ppm) are lower. The difference in LILEs and HFSEs between the SGB and the MGB samples has proved to be significant for Sr, Cs, Ba, U, Pb,Zr,Hf,and Ta using the Student’s t-test at better than 95%confidence level (p ＜0.05) (Table 3).

4.4 Rare earth elements (REE)

Fig. 4 Chondrite normalized REE diagram (Sun and McDonough 1989)for the Sonakhan samples,Mahakoshal samples,and average of Sonakhan and Mahakoshal samples

The total rare earth elements(∑REE)concentration in the samples from the SGB have a higher mean value(201.8 ppm) and a lower mean value in the MGB samples(175.7 ppm). On the chondrite normalized REE diagram(Fig. 4; normalizing values from Sun and McDonough(1989), the LREE patterns of the SGB and the MGB samples are fractionated with LREE enrichment (La/Yb)n = 9.97 for the SGB samples and 10.59 for the MGB samples,flat HREE(Gd/Yb)n = 1.33 for the SGB and 1.51 for the MGB samples.The samples from both belts have a negative Eu/Eu* anomaly, which is 0.76 for the SGB and 0.53 for the MGB samples. Overall the concentrations of REE of the SGB and MGB samples are almost similar.However,a difference for Eu,Tb,and Dy has proved to be significant using the Student’s t-test at better than 95%confidence level (p ＜0.05) (Table 3).

5 Discussion

5.1 Source rock characteristics

In comparison to Upper Continental Crust (UCC) (Taylor and McLennan 1985), samples of the MGB are rich in MgO and Fe2O3t, and Co, Sc, V, Ni, and Cr (Fig. 5). The major elements as MgO,Fe2O3t,and compatible elements,such as transition elements are characteristic of mafic source rocks (Feng and Kerrich, 1990). The proportions of mafic elements, especially MgO, Fe2O3t, Ni, and Cr,indicate the presence of olivine and pyroxene in the source rocks. This is evident from the significant correlation between MgO and Fe2O3t(r = 0.89). In comparison to UCC,the SGB samples do not have a high concentration of these mafic elements (except for Co) (Fig. 5). The difference in mafic components between the MGB and SGB samples is consistent with the Student’s t-test which has proved to be significant (＞95%) for MgO, V, Cr, Ni, and Co (Table 3).

On the UCC-normalized diagram, both the MGB and SGB samples are rich in Th and U. However, in terms of HFSEs such as Nb and Ta, SGB samples show slightly higher concentration, while MGB samples show depletion(prominent negative anomalies) in these elements. The enrichment of LILEs and depletion of HFSEs,especially in Nb and Tb,are associated with arc rocks,hence,indicating an arc character for the source of MGB samples (Eissen et al.1991;Fig. 5).Both the MGB and SGB samples show slight LREE and HREE enrichment in comparison to UCC,indicating a source similar to UCC(Fig. 6).This is further attested by the strong LREE enrichment [(La/Yb)n = 9.97 in the SGB samples and 10.59 in the MGB samples], and slight HREE enrichment on the chondrite normalized REE diagrams [(Gd/Yb)n = 1.33 of the SGB and 1.51 of the MGB samples]. Furthermore, both the SGB and MGB samples show a significant negative Eu anomaly (Eu/Eu* = 0.76 for the SGB and 0.53 for the MGB). It is interesting to note that the REE patterns and Eu anomalies of the MGB and SGB samples are similar to each other and suggest that the same mineral phases control REE in both groups of samples derived from highly differentiated sources such as granites. However, from the foregoing discussion, it becomes clear that the MGB samples also carry a mafic signature given by the amounts of Fe2O3t,MgO, Ni, Cr, Co, and V. The samples also show an arc signature given by their high concentrations in LILEs and depletion in HFSEs. Therefore, it can be considered that the MGB received the clastic load from a mixed source of arc affinity with a major contribution from mafic rocks and less from felsic rocks, whereas the clastic load of the SGB derived from a felsic source only. The preservation of a felsic signature rather than a mafic signature in the MGB samples reflected by REE characteristics is actually because their budget in clastic sedimentary rocks is chiefly controlled by granitoids, which mask the contribution of mafic–ultramafic components (Jahn and Condie 1995).This conclusion is consistent with the enrichment of both Cr and Zr in the MGB samples since these two elements monitor chromite and zircon contents, respectively(Wronkiewicz and Condie 1989). The difference between Cr and Zr concentration of SGB and MGB samples is also depicted by the Student’s t-test at ＞95% confidence level(p ＜0.05).The MGB samples are enriched in both Cr and Zr (82.63 and 8941 ppm), in comparison to the SGB ones(Cr = 21.85 ppm and Zr = 663.6 ppm). On the Ni–Cr diagram (Taylor and McLennan 1985; Fig. 7), the MGB and SGB samples plot along a linear trend. It is evident from Fig. 7 that the SGB samples, which plot at the lower end of the Ni–Cr trend(post-Archean field)may have been derived from a felsic source, while the clastic load of the MGB is enriched in both Ni and Cr in comparison to the SGB one’s plot at the upper end of the Ni–Cr trend (Post Archean to Neoarchean field). This complements the enrichment of MgO and Fe2O3tin the MGB samples,which reflects the mafic nature of source rocks.

Fig. 5 UCC normalized (Taylor and McLennan 1985) major and trace element diagrams for Sonakhan samples Mahakoshal samples, and average of the Sonakhan and Mahakoshal samples

5.2 Classification of SGB and MGB sediments and sedimentary sorting

Fig. 6 UCC normalized (Taylor and McLennan 1985) REE (rare earth elements) for the Sonakhan samples Mahakoshal samples, and average of the Sonakhan and Mahakoshal samples

To classify the terrigenous sands, Pettijohn et al. (1972)developed a diagram based on log (Na2O/K2O) versus log(SiO2/Al2O3) (Fig. 8A). According to this classification,the samples of the MGB are mostly litharenites while the ones of the SGB are greywackes, which indicates a compositional immaturity of the sediments.The proportions of the major elements of the MGB and SGB samples reflect the relative proportions of quartz to feldspars and clays/phyllosilicates, and the majority of the samples plot along the expected mixing trend on the SiO2versus Al2O3plot (Fig. 8b). Furthermore, the plot of SiO2versus Al2O3(Fig. 8b) can better be used to understand the effect of weathering and sedimentary sorting. It is clear from the plot that the SGB samples (except two samples) cluster in the middle of the trend line indicating a compositional homogeneity with a small effect of weathering and sorting.However, the MGB samples which plot all along a linear trend, indicate that most of the samples were affected by weathering and only three samples with SiO2value ＞72%have been affected by sorting.

Fig.7 Ni–Cr distributions in the Sonakhan and Mahakoshal samples(after Taylor and McLennan 1985)

5.3 Sediment recycling and tectonics

The Index of Compositional Variability, ICV = (Fe2O3-+ K2O + Na2O + CaO + MgO + TiO2)/ Al2O3can be used to evaluate the original composition of the sediments(Cox and Lowe 1995;Cox et al.1995;Mondal et al.2012).The ICV is a measure of the abundance of alumina relative to other major constituents of rock except for SiO2. Nonclay minerals have a higher ratio of the major cations to Al2O3than clay minerals.Thus the non-clay minerals have a higher ICV. The ICV value decreases in the order of pyroxene and amphibole (10–100), biotite (8), alkali feldspar (0.8–1), plagioclase (0.6), muscovite and illite (0.3),montmorillionite (0.15–0.3), and kaolinite (0.03–0.05)(Cox et al. 1995). Hence, immature shales with a high percent of non-clay silicate minerals show ICV values greater than one. These shales are often found in tectonically active settings as first-cycle deposits (Van de Kamp and Leake 1985). In contrast, more mature mudrocks with mostly clay minerals generally have ICV values lower than one(Cox et al.1995).Such shales ought to form in cratons of quiescent tectonic environments (Weaver 1989). However,they have also been found as the first cycle material in intensely weathered zones(Barshad 1966).The ICV values of the SGB samples range from 0.55 to 1.11 with an average of 0.88 suggesting that most of the samples are compositionally mature and were likely dominated by recycling, but this is not the case. Since these samples are greywacke which is texturally and compositionally immature rocks, and, therefore, the reasons for having low ICV values must be different.The petrographic evidence shows that some of the samples have preserved the originalmineralogy and contain abundant fresh plagioclase grains(up to 20%) therefore, lowering their ICV values to ≤1(Fig. 3).Furthermore,the presence of abundant plagioclase grains in the metagreywackes indicates that these rocks are arkosic. There are also some metamorphic rock fragments in these metagreywackes. The rock fragments are formed by continental block faulting of basement rocks(Dickinson 1985). The fresh nature of plagioclase grains suggests that the source rocks were not affected by severe chemical weathering.The ICV value of MGB samples shows a wide variation ranging from 0.65 to 3.1 with an average value of 1.24 and was dominated by first cycle sediments of mafic sources and, therefore, indicate tectonically active settings of the basin. This inference is further supported by the litharenite nature of the MGB samples. Dickinson (1985)reported that the litharenites derived from orogens, where stratified rocks are deformed, uplifted and eroded, and deposited in a foreland basin. According to Dickinson(1985), litharenites are also formed during back-arc thrusting where the arcs are uplifted and eroded. Since the Central Indian Tectonic Zone, suggests a typical subduction–accretion setting where the Bastar craton is believed to have been subducted under the hinterland of the Bundlekhand craton,therefore,such inference can be taken into consideration. The abundance of the trace elements as Sc,La, Zr, Cr, Th is characteristic for different tectonic settings and is considered to remain unaffected by remobilization during weathering, as well as medium-grade metamorphic processes (Bhatia and Crook 1986). Therefore, these elements have a great significance to decipher the source rock composition and tectonic setting. The ternary plot of La-Th-Sc (Bhatia and crook 1986; Fig. 9)can depict four types of tectonic settings, namely(a)oceanic arcs,(b)continental arcs,(c)active continental margin, and (d) passive continental margin. The results on this diagram indicate a continental arc provenance for the MGB samples and a continental passive and active margin provenance for the SGB samples (Fig. 9).

Fig. 8 a Sandstone classification of the Sonakhan and Mahakoshal samples using log (SiO2/Al2O3) versus log(Na2O/K2O) (Pettijohn et al.1973). b A plot of SiO2 versus Al2O3 to understand the effect of weathering and sedimentary sorting

Fig. 9 Th–La–Sc trace element tectonic setting discrimination plot for the Sonakhan and Mahakoshal samples (after Bhatia and Crook 1986)

Fig. 10 Al2O3–(CaO* + Na2O)–K2O (A–CN–K) ternary plot, after Nesbitt and Young (1982) (CaO* = CaO in silicate phase) shows the weathering trends of Sonakhan and Mahakoshal samples and their probable source rocks. Numbers 1–4 denote compositional trends of initial weathering profiles of different rocks. 1—gabbro; 2—tonalite;3—granodiorite; 4—granite

Fig. 11 Bivariate diagrams showing correlation between (a) Al2O3 and K2O (b) Al2O3 and Na2O and (c) Al2O3 and CaO

5.4 Paleoweathering and paleoweathering trends

Fig.12 Bivariate plots showing(a)SiO2 versus Al2O3 concentrations and (b) Al2O3 versus FeO concentrations. The Sonakhan and Mahakoshal samples have been plotted relative to the idealized composition of the observed minerals(Cullers and Podkovyrov 2000)

Fig.13 Weathering trends of the Sonakhan and Mahakoshal samples on the Index of Compositional Variability (ICV) versus Chemical Index of Alteration (CIA) plot (Potter et al. 2005)

It is clear from the UCC normalized diagrams that, almost for the same values of SiO2, Al2O3, and Rb, the MGB samples have higher values of immobile elements like TiO2,and lower values of Na2O,CaO,and Sr than the SGB samples. This suggests that the MGB samples underwent an intenser degree of chemical weathering than the SGB samples (Table 3). Nesbitt and Young (1982) defined a chemical index of alteration (CIA) to quantitatively measure the degree of weathering (in molecular proportions):CIA = [Al2O3/(Al2O3+ CaO*+ Na2O + K2O)] × 100;where CaO*represents the CaO in the silicate fraction only(Nesbitt and Young 1982). The average CIA values of the SGB samples vary from 54.22 to 71.75 with an average of 59.8 indicating a low to moderate chemical weathering.The CIA values of the MGB samples have a wide range of CIA values and range from 55.83 to 90.71 with an average of 74.50 suggesting moderate to intense chemical weathering. Moderate to intense weathering of the source area for the MGB samples and low to moderate chemical weathering for the SGB samples is further indicated by plagioclase index of alteration(PIA)values(Table 3)given by the following equation: PIA = [(Al2O3–K2O)/(Al2O3-+ CaO + Na2O–K2O)] × 100, proposed by Fedo et al.(1995). The MGB samples have PIA values of 58.9–97.1 with an average of 89.4 indicating near-complete alteration of plagioclase. However, the SGB samples have a PIA value in the range of 54.93–71.75 with an average of 64.34 suggesting very low weathering of plagioclases.

Another approach to understanding the weathering trend of the original source rock is to plot molar ratios of Al2O3–(CaO* + Na2O)–K2O in A–CN–K compositional space to potentially separate compositional changes of sediments related to chemical weathering, transportation, diagenesis,and metamorphism (Fedo et al. 1995, 1997; Fig. 10). The A-CN-K diagram is useful for evaluating fresh rock compositions and examining their weathering trends because the upper crust is dominated by plagioclase and K-feldspar–rich rocks and their weathering products, the clay minerals (Nesbitt and Young 1984, 1989). Normally sediments show weathering trends that plot parallel to the A–CN boundary,and they extract back to a plagioclase–alkali feldspar horizontal line of the source composition unless post-depositional K-metasomatism affected the rocks(Fig. 10). The K-metasomatism of kaolinite weathered rocks can produce illite, which plots at right angles to the A–K side of the diagram (Fig. 10). Chemical weathering reactions involving the destruction of plagioclase drive these trends to more aluminous or potassic compositions along the A–K join (Nesbitt and Young 1989). The predicted weathering trends of the SGB and the MGB samples in relation to probable source rocks of granite,granodiorite,diorite, tonalite, and basalt compositions are shown in Fig. 8. The CIA values of the SGB samples vary from 54.22 to 71.75 with an average of 69.68 indicating a low to moderate chemical weathering and pointing to the availability of a weakly weathered profile available for erosion in the source area. The SGB samples plot on a trend reflecting a predominantly granitic to a granodioritic source. This trend reflects alteration of K-feldspar to illite,and samples that plot to the right of the granodiorite and granite trend indicate that some minor addition of K has occurred.

In comparison to SGB,the MGB samples that plot in the A–CN–K diagram produce no clear trend back to the source composition. It is clear from the diagram that most of the MGB samples neither plot parallel to the A-CN boundary nor perpendicular to the A-K side of the diagram,but show an advanced weathering trend in which the samples plot along the A–K line suggestive of intense chemical weathering of the source rocks ranging in composition from tonalite to basalt(Fig. 10).However,the samples show a wide range of CIA values from 55.83 to 90.71(average = 74.50)and the anomalously low contents of Na2O that range from 0.01 to 0.87 wt.% (average = 0.18), whereas K2O is high and ranges from 0.22 to 7.69 wt.% (average = 4.11) suggesting that post-depositional K-metasomatism has affected these rocks (Table 3).This is further supported by a positive correlation coefficient value between Al2O3and K2O (r = 0.85) and, negative correlation between Al2O3and Na2O(r = - 0.27)and Al2O3and CaO (- 0.14) (Fig. 11). Potassium metasomatism is particularly common in the Precambrian rocks,and typically involves the conversion of kaolin (residual weathering product) to illite by reaction with K-bearing pore waters during diagenesis (Fedo et al. 1995). Such a trend is most likely produced by intense weathering of mafic rocks,which were later affected by post-depositional K-metasomatism. This interpretation is consistent with other major and trace element characteristics, which indicate the mafic source rocks for the MGB samples as already discussed. An estimate of premetasomatized or‘corrected’CIA values can be made by drawing a line from the K2O apex through individual data points; the intersection point of this line with the ‘predicted weathering path’gives the pre-metasomatized CIA value,which can be directly read off the scale on the left. The difference between the pre-metasomatized and the current CIA values allows quantitative estimation of K enrichment in a rock.The present CIA values (uncorrected) range from 55.83 to 90.71 with an average value of 74.50. Since these sediments have undergone K metasomatism, the pre-metasomatised CIA values must have been quite higher. The corrected pre-metasomatised CIA values of the MGB samples have been calculated and fall in the range of 66–97.5, with an average of 90.1 indicating near-complete destruction of plagioclases.This is consistent with the PIA values of these samples. Therefore, CIA and PIA values suggest moderate to intense chemical weathering of the source area.

In another attempt to characterize weathering of the source rocks,Cox et al.(1995)have observed that the K2O/Al2O3ratios are useful in depicting how much alkali feldspar vs. plagioclase and clay minerals may have been present in the original sediments. The K2O/Al2O3ratios of different K and Al-bearing minerals vary significantly, for example, K2O/Al2O3ratios of alkali feldspar = ～0.4–1,illite = ～0.3, and for other clay minerals = ～0. Shales with K2O/Al2O3ratios greater than 0.5 suggest a significant quantity of alkali feldspar relative to other minerals in the original sediments. The K2O/Al2O3ratios of the MGB samples vary from 0.01 to 0.29 with an average of 0.26,suggesting minimal alkali feldspar relative to other minerals.The K2O/Al2O3ratios of the SGB samples vary from 0.18 to 0.29 with an average of 0.23. This in turn implies that alkali feldspar should have been minimal in the source rocks. This is further revealed on the SiO2versus Al2O3concentrations and Al2O3versus FeO concentrations of samples which have been plotted relative to the idealized composition of the observed minerals (Fig. 12). It is clear from the SiO2versus Al2O3plot that most of the samples plot between quartz and illite composition. However, four samples two from each the MGB and SGB are skewed towards hematite and opaques. On the Al2O3versus FeO plot the samples plot between quartz and Mg-rich chlorite and between Mg-chlorite and illite indicating the presence of both illite and Mg-chlorites in the original sediments.However, a few samples are skewed towards Fe-rich chlorite. The occurrence of illite is very well in agreement with the A-CN-K diagram in which the samples plot near the illite field.

Another method of extracting information on weathering trends of the source rocks from major element data of clastic rocks that show the influence of chemical weathering is to pair the CIA value with the ICV (index of chemical variability) (Gaschnig et al. 2016; Fig. 13). ICV values are high for mafic source lithologies and when plotted against CIA values (Fig. 13), straight lines connecting a completely weathered end member to average igneous compositions reflect weathering trends (Potter et al. 2005), on this plot the MGB samples cluster on the weathering trend of granite,while the majority of the MGB samples plot on a broad weathering trend between andesite and basalt-komatiite with an average basaltic protolith,which is consistent with other geochemical signatures indicating contrasting source rocks available for SGB during the Late Archean to Early Paleoproterozoic and for the MGB during the Late Paleoproterozoic in the Central Indian Shield.

6 Implications

6.1 Implications for greenstone belt development

Based on the summary of geochemistry from the SGB and MGB of the Central Indian Shield as discussed above,it is reasonable to envisage that the greenstone belt formation during the Late Archean and Late Paleoproterozoic in the central Indian shield involved two different mechanisms.The Late Archean to Early Paleoproterozoc SGB reveals that the basin received most of its sediments from highly differentiated rocks (granitic) like that of a craton with a passive-active tectonic setting and therefore, appears to have been developed as an autochthnous greenstone belt.The study of Mondal and Raza (2009) using geochemical data of volcanic rocks of the SGB reveals that theSonakhan greenstone belt of the Bastar craton evolved by mantle-plume upwelling.Several pieces of evidence show,that mantle plumes may have played a role in Archean magma production, crustal underplating, and development of Archean greenstone belts(Tomlinson and Condie 2001).The relative abundance of plume-related greenstones in the Archean may reflect the hotter, less easily subducted Archean oceanic lithosphere as compared to the Proterozoic(Tomlinson and Condie 2001).In contrast to the SGB,the Late Paleoproterozoic MGB shows a robust geochemical signature of bimodal source rocks with subduction signature and continental arc setting and therefore,appears to have been developed as near continental arc systems as an allochthonous greenstone belt,similar to those observed in younger orogens. Arc rocks in greenstone belts occur mainly in the bimodal geochemical association in both the basaltic basal zones of these sequences and in the isolated felsic eruptive centers(Ayer et al.2005).Since the MGB is present in the Central Indian Tectonic Zone, which has very well-developed fold and thrust belts, metamorphic core complexes, blue schists, and, therefore, should be considered as an allochthonous greenstone belt. Therefore,SGB and MGB record both vertical(mantle plume related)and lateral accretionary tectonics (plate subduction-related), building continents in the early Earth in India as seen in other parts of the world (VanKranendonk et al.2004, Santosh 2013). Furthermore, the geochemical signatures of the sedimentary record of the SGB and MGB are not consistent with an Archaean upper crust dominated by the bimodal basic-felsic suite with both units derived by melting at mantle depths and the Proterozoic upper continental crust dominated by K-rich granites, derived by intracrustal melting. Therefore, revealing a reverse of the typical age-composition relations,that is,the occurrence of bimodal basic-felsic suite in the Late Paleoproterozoic MGB sediments and felsic in the Late Archean to Early Paleoproterozoic SGB sediments. Such compositional trend can be interpreted in terms of tectonic differences between the SGB and MGB as revealed in the present study. This does not pose any problem to the general understanding of the worldwide secular change in crustal composition from bimodal in the Archean to felsic in the Proterozoic. This contrast between the typical age-composition relations has been identified in many places of the world and has been attributed to the tectonic difference by McLennan and Taylor (1988).

Fig. 14 Figures showing the position of India in the configuration of 2.45 Ga Columbia supercontinent and 1.78 Ga Rodina Supercontinent(Pesonen et al. 2003)

6.2 Implications for paleogeography

Studying weathering regimes of the Precambrian rocks is important for a better understanding of the influence of climate on weathering, erosion, and runoff in the geological past. The high chemical weathering inferred from the Late Paleoproterozoic MGB can be attributed to weathering under moderate to severe environmental conditions such as low pH, high temperature and humidity, and extreme leaching typical of tropical climatic conditions.However, the CIA values of the Late Archean to Early Paleoproterozoic SGB indicate chemical weathering under cool conditions. The change in CIA values from the Late Archean to the Late Paleoproterozoic in the Central Indian Shield can be interpreted either by a change in the global climate from the Late Archean to the Late Paleoproterozoic or due to the migration of the Central Indian Shield through various climatic zones.The latter interpretation seems to be more accurate since the paleomagnetic studies of the Late Archean rocks indicate that the Indian plate was set at higher latitudes during the Late Archean to Early Paleoproterozoic (at 2.45 Ga) while in low latitudes during the later part of the Paleoproterozoic (1.78 Ga). The Paleomagnetic data used for reconstruction at 2.45 Ga reveals that the Ur fragments Yilgarn from Australia and especially Dharwar from India, was placed at high, almost polar(south)latitudes(Pesonen et al.2003).These Late Archean to Early Paleoproterozoic supracrustal strata have been considered similar to Neoproterozoic strata that also have glaciogenic sequences and paleoweathering zones (e.g.,Evans 2000). It is interesting to note that the Jonk Conglomerate of the SGB also suggests its fluvio-glacial origin(Das et al 1990). Good quality palaeomagnetic data at 1.78 Ga from Amazonia, Australia, Baltica, India, Kalahari, Laurentia, and North China (Fig. 14) reveals that the 1.78 Ga‘‘Early Columbia’’is made of two landmasses:the one consists of an elongate large continental area of Amazonia, Baltica, Laurentia, and North China, and another includes Australia, India, and Kalahari. These continents were positioned at low to intermediate latitudes during 1.88–1.77 Ga (Pesonen et al. 2003). Thus, the variation in the CIA values from the Late Archean to the Late Paleoproterozoic in the Central Indian Shield is the result of the migration of the Indian plate from southern high to lower latitudes,and the results of this study are very much consistent with the paleomagnetic data and the paleographic positions that were placed for India in the configurations for supercontinents (Fig. 14).

7 Conclusions

The origin of greenstone belts remains unclear and there is an intense debate regarding their origin. The Indian Peninsula posses a number of greenstone belts providing a great opportunity to understand their development. Based on the geochemical study of metasedimenatry sequences from the SGB and MGB of the Central Indian Shield, it is concluded that those greenstone formations involved two entirely different mechanisms that evolved from Late Archean to Late Paleoproterozoic and recorded both vertical (plume related rift) and lateral accretionary tectonics(subduction-related). The Late Archean to Early Paleoproterozoc SGB developed as an autochthonous greenstone belt and received most of its sediments from highly differentiated continental rocks in a passive-active tectonic setting like that of a continental rift. The Late Paleoproterozoic MGB developed as an allochthonous greenstone belt, similar to those in younger orogens, and received sediments from continental arc-related bimodal sources.The variation in the CIA values from the Late Archean to the Late Paleoproterozoic is due to the migration of the Indian plate from southern high latitudes to lower latitudes.This is very much consistent with the paleomagnetic data and paleographic positions of India in the supercontinent configurations.

AcknowledgementsWe wish to express sincere thanks to the Director of National Geophysical Research Institute (NGRI),Hyderabad, and the Director of Wadia Institute of Himalayan Geology, Dehra Dun for providing laboratory facilities during chemical analysis.We are thankful to the Chairperson,Department of Geology,A.M.U., India for providing the necessary facilities to carry out this investigation.H.Wani thankfully acknowledges the financial support of UGC, Govt. of India in the form of a research project F. No.40-304/2011 (SR). We also thank three anonymous reviewers for their suggestions and critical reviews.

Author contributions Data acquisition: HW, MEAM. Conception:MEAM, IA. Writing/Drafting: HW, IA. Data interpretation: HW,MEAM,IA.Revision:MEAM,IA.All authors read and approved the final manuscript.

FundingHW acknowledges the financial support from UGC, Govt.of India in the form of a research project F. No. 40-304/2011(SR).MEAM is thankful to the Department of Science and Technology,Govt. of India for financial support in the form of a Research Project(SR/S4/ES-180/2005).

Data availability All data generated and analyzed during this study are included in this published article and its supplementary information files.

Declarations

Conflict of interest The authors declare that they have no competing interests.