African, southern Indian and South American cratons were not part of the Rodinia supercontinent: evidence from field relationships and geochronology

Abstract

We discuss the question whether the late Mesoproterozoic and early Neoproterozoic rocks of eastern, central and southern Africa, Madagascar, southern India, Sri Lanka and South America have played any role in the formation and dispersal of the supercontinent Rodinia, believed to have existed between about 1000 and 750 Ma ago. First, there is little evidence for the production of significant volumes of ~1.4–1.0 Ga (Kibaran or Grenvillian age) continental crust in the Mozambique belt (MB) of East Africa, except, perhaps, in parts of northern Mozambique. This is also valid for most terranes related to West Gondwana, which are made up of basement rocks older than Mesoproterozoic, reworked in the Brasiliano/Pan-African orogenic cycle. This crust cannot be conclusively related to either magmatic accretion processes on the active margin of Rodinia or continental collision leading to amalgamation of the supercontinent. So far, no 1.4–1.0 Ga rocks have been identified in Madagascar. Secondly, there is no conclusive evidence for a ~1.0 Ga high-grade metamorphic event in the MB, although such metamorphism has been recorded in the presumed continuation of the MB in East Antarctica. In South America, even the Sunsas mobile belt, which is correlated with the Grenville belt of North America, does not include high-grade metamorphic rocks. All terranes with Mesoproterozoic ages seem to have evolved within extensional, aulacogen-type structures, and their compressional deformation, where observed, is normally much younger and is related to amalgamation of Gondwana. This is also valid for the Trans-Saharan and West Congo belts of West Africa. Third, there is also no evidence for post-1000 Ma sedimentary sequences that were deposited on the passive margin(s) of Rodinia. In contrast, the MB of East Africa and Madagascar is characterized by extensive structural reworking and metamorphic overprinting of Archaean rocks, particularly in Tanzania and Madagascar, and these rocks either constitute marginal parts of cratonic domains or represent crustal blocks (terranes or microcontinents?) of unknown derivation. This is also the case for most terranes included in the Borborema/Trans-Saharan belt of northeastern Brazil and west-central Africa, as well as those of the Central Goı́ás Massif in central Brazil and the Mantiqueira province of eastern and southeastern Brazil. Furthermore, there is evidence for extensive granitoid magmatism in the period ~840 to <600 Ma whose predominant calc-alkaline chemistry suggests subduction-related active margin processes during the assembly of the supercontinent Gondwana. The location of the main Neoproterozoic magmatic arcs suggests that a large oceanic domain separated the core of Rodinia, namely Laurentia plus Amazonia, Baltica and West Africa, from several continental masses and fragments now in the southern hemisphere, such as the São Francisco/Congo, Kalahari and Rio de La Plata cratons, as well as the Borborema/Trans-Saharan, Central Goiás Massif and Paraná blocks. Moreover, many extensional tectonic events detected in the southern hemisphere continental masses, but also many radiometric ages of granitois that are already associated with the process of amalgamation of Gondwana, are comprised within the 800–1000 age interval. This seems incompatible with current views on the time of disintegration of Rodinia, assumed to have occurred at around 750 Ma.