Discussion and Conclusion

The Neoproterozoic MHM granitoids (881 - 479 Ma) are peraluminous and show high-K calc-alkaline and I-type characteristics. They exhibit fractionated LILE/HFSE and LREE/HREE patterns and plot within the volcanic arc granite (VAG) filed in the Rb vs. Y+Nb diagram (Fig. 11A) and common syn-collision granite (Syn-COLG) and VAG field in the Nb vs. Y tectonic discrimination diagram (Fig. 11B).

Shillong plateau experienced Pan-African tectono-metamorphic events during Neoproterozoic resulting in granulite-facies rocks in the basement. It is observed that several Pan-African terrains experienced granulite-facies metamorphism and granitoid magmatism during Neoproterozoic and the later is formed from anatectic melts produced during granulite-facies metamorphism. For example in the southern Granulite Terrain of India the Madurai Block and Kerala Khondalite belt experienced a Pan-African tectono-metamorphic evolution in Neoproterozoic times that culminated in granulite-facies metamorphism at 650–550 Ma and extensive granitoids magmatism at 590–525 Ma (Braun and Kriegsmann, 2003). In the Prydz Bay region of East Antarctica the emplacement of syntectonic granites and prograde amphibolite-granulites facies metamorphism is found coeval during 500–540 Ma (Mikhalsky et al., 2001; Liu et al., 2006). It has been observed that the in general the granulites of lower continental crust are characterized by low concentration of LILE (Rb, Th, U and K) and a number of different mechanisms have been proposed to explain the observed LILE depletion in granulites. An adequate mechanism to explain the observed LILE depletion is partitioning of LILE into an anatectic melt fraction that took place during granulite-facies metamorphism within the lower crust (Fyfe, 1973; Tarney and Weaver, 1987; Cartwright, 1995). A cross-check was thus initiated whether the granitoids under consideration are anatectic magmas produced during granulite-facies metamorphism comparing the trace and rare earth elemental signatures of the granitoids and the granulites in Fig.s 9, 10. The granitoids show contrasting geochemical characteristics with the basement granulites. The granitoids show negative Nb and Ti and positive Zr anomalies in the spidergram, whereas the granulites show positive Nb, Ti and negative Zr. Moreover, the LILE abundances of both the granulites and the granitoids are almost similar suggesting that the granitoids are not anatectic melts produced during the granulite-facies metamorphism. In the event of an anatectic melt the granulites should have been depleted in the LILEs. The granulite S08-42 shows moderately fractionated REE patterns with no negative anomaly and the enrichment level of the elements over the entire spectrum is 30 - 40 folds lesser compared to the granitoids while the other granulite S08-27 shows strongly fractionated patterns with LREE enrichment level similar to the granitoids and HREE is significantly less (20 – 30 folds) (Fig. 10) compared to the granitoids along with a strong negative Eu anomaly. Based on the geochemical observation, it can be inferred that the granitoids are not anatectic magmas derived from partial melting of the basement granulitic rocks. The geochemical signatures of the granitoids point towards magmatism in a convergent margin tectonic setting within Shillong plateau and the emplacement of these granitoids during Neoproterozoic times, thus have significant implications for the construction of the Gondwana supercontinent.

The supercontinent Rodinia formed around a core of Precambrian North America (Laurentia) during the Grenvillian event (~1300–1000 Ma ago) and is supposed to have occurred by accretion of all existing continental fragments of that time, including those that now make up the cratonic components of the Gondwana continents. The process of accretion and amalgamation of continental fragments was followed by a period of supercontinent disintegration and dispersal, considered to have started around 750–800 Ma ago (Hoffman, 1991, 1999; Pisarevsky et al., 2003) and by 550 Ma, most of the dispersed terranes had reassembled, probably in a different configuration, in the supercontinent Gondwana (Meert and Van der Voo, 1997). This occurred after closure of several oceans of different sizes. For instance, the Mozambique Ocean between East and West Gondwana, conventionally interpreted to have been subducted and resulted in collision of these two large continental masses (Burke and Dewey, 1972; Kröner, 2002). Likewise, the South American and African continental blocks amalgamated into West Gondwana in several stages of ocean closure and collision between 900 and ~550 Ma (see, e.g., Campos-Neto, 2000; Alkmim et al., 2001, Cordani et al., 2003).

The high-grade metamorphic basement of Sri Lanka consists of ~895–1100 Ma old calc-alkaline granitoids rocks that are exposed in the Wanni (western Sri Lanka) and Vijayan (eastern Sri Lanka) Complexes (Kehelpannala, 1997; Kröner, 1991; Kröner et al., 2003), which are interpreted to have been produced due to subduction/accretion event at around 1000 Ma. Similarly, the Goia´s Magmatic Arc of Brazil includes a series of 900–600 Ma granitoids rocks, formed during successive episodes of intraoceanic subduction. This large amount of juvenile crustal material, located close to the Neoproterozoic megasuture of the Trans-Brasiliano lineament in central Brazil, is the main evidence for the existence of the Goia´s Ocean, which must have had a considerable size (Cordani and Sato, 1999; Pimentel et al., 2000).

The geochemical signatures, the age of the granitoids and the position of Shillong plateau at the leading margin of India during Neo-proterozoic the geochemical signatures thus point to the generation of the melts for the granitoids by partial melting of metasomatized mantle and /or overlying crust and subsequent mingling of both the magma and their emplacement in a convergent tectonic setting during re-assembling of the continents in Gondwana subsequent to disintegration and dispersal of Rodinia. Similar explanations are drawn for the 895–1100 Ma old calc-alkaline granitoids of Srilanka; 900–600 Ma granitoids rocks of the Goia´s Magmatic Arc of Brazil. However, the full extent and knowledge of these rocks within Shillong plateau requires integrated geological, geochemical, isotopic, geochronological and palaeomagnetic studies.