I-type, K-rich calc-alkaline hornblende-biotite granitoids (HKCA) suites are typical of both pre- and post-collisional tectonic settings, and their petrogenesis is challenging, since they are dominated by rocks with intermediate composition (e.g., 60–68 wt% SiO2; 10–20 vol % mafic mineral) that are neither typical products of crust nor of mantle partial melting. These granitoids are particularly abundant as the main components of Ediacaran batholiths intruding the Neoproterozoic Ribeira Fold Belt in SE Brazil, e.g. constituting ∼70% of the exposition of the Agudos Grandes Batholith, where they were emplaced at 620-610 Ma, as indicated by U–Pb zircon dating from our previous work and new data presented here. Two types of coeval granitoids comprise an expanded suite of mafic to felsic titanite and allanite-rich biotite granitoids (Itapevi-type) and a single occurrence of biotite-muscovite leucogranite (Turvo-type). Elemental and isotope whole-rock data indicate that different sources were involved in the genesis of their parent magmas, with crust with old residence predominating for the HKCA and Itapevi types, as indicated by very negative (−16 to −20) εNd(t) and low initial 206Pb/204Pb (16.4–16.7), and younger metasedimentary rocks as the source of the Turvo leucogranite (εNd(t) (∼-14); initial 87Sr/86Sr = 0.717). This “syn-orogenic” magmatism was immediately followed, as demonstrated by field and geochronological data, by ∼605-600 Ma circumscribed plutons dominated by more fractionated (±muscovite)-biotite granites that derive from melts generated in the middle crust from sources that are slightly younger than those of the HKCA granitoids, as indicated by slightly less negative (εNd(t) (similar to those of the Turvo leucogranite) and higher 206Pb/204Pb (17.1–17.3). After a ∼15 Myr lull, A-type plutons were emplaced in a post-orogenic setting, probably from mantle and crust sources. The syn-to late-orogenic (620-600 Ma) magmatism of the Agudos Grandes was produced by a thermal anomaly, probably associated with a continental arc. A combination of elemental and isotope geochemistry and inherited zircon data suggests that crustal melting was probably induced by ascension of mantle-derived basic magmas, and initiated in hot zones at the lower portions of a previously thickened crust, formed by old Paleoproterozoic to Archean material (the basement of the Apiaí-São Roque Domain of the Ribeira Belt?), and progressed to the middle portions of the crust, where it provoked partial melting of a different crustal domain with younger age and crustal residence (Embu Complex orthogneisses and metasediments).
Read full abstract