Regional geochronological and geological criteria has lead to the subdivision of orogenic magmatism in the Araçuaí-Ribeira Belt into G1, G2, G3, G4 and G5 (Wiedemann et al., 1986; Wiedemann, 1993; Ludka et al., 1998; Pedrosa-Soares et al., 1999, 2001; Ludka & Wiedemann, 2000; Medeiros et al., 2000, 2001, 2003; Mendes et al. 1997, 1999, Wiedemann et al., 2002). The G1 and G2 suites (~ 580 Ma) are syn-collisional phases but have different proportions of source components. The G3 granitoids are less voluminous and resulted from the remelting of the previous ones, in a early post-collisional stage (~ 560-535 Ma). The G4 suite is restricted to the northwestern part of the belt and is not considered in this study any further. The G5 suite (~500 Ma) records the youngest post-collisional magmatic episode in the area.
The G1 suite consists of tectonically foliated tonalite, granodiorite and minor granite and diorite (unit 8 in Figure 2), usually showing deformed K-feldspar augen megacrysts in a foliated biotite-rich matrix. Mafic to intermediate microgranular enclaves up to gabbroic compositions, are commonly found stretched along the gneissic foliation (Lammerer, 1987; Fritzer, 1991; Geiger, 1993; Wiedemann et al., 1997, 2002). These granitoids are considered syn-collisional as they show the same deformation as the enclosing gneisses. U-Pb zircon ages (Söllner et al., 1991) constrain crystallization of G1 granitoids, in Espírito Santo, to between 586 and 575 Ma. Part of the syn-collisional G1 magmas crystallized under dry conditions and high CO2-pressures (Fritzer, 1991; Wiedemann et al., 1997; Mendes et al., 1997) originating magmatic charnockitoids with noritic cores. As an example the hypersthene-bearing orthogneiss from the Serra do Valentim, north of the Caparaó ridge, was formed under temperatures around 800 C and pressures of 8 to 9 kb (Fritzer, 1991; Seidensticker & Wiedemann, 1992).
Geochemical data from several G1 plutons point towards calc-alkaline, metaluminous to fairly peraluminous magmas, formed in a volcanic arc setting (Geiger, 1993; Campos-Neto & Figueiredo, 1995; Wiedemann et al., 1997; Pedrosa-Soares et al., 1999). Pre-collisional calc-alkaline granitoids, such as the Rio Negro granitoids described further south (Heilbron et al., 2000) have not yet been recognized in this area but zircons of similar age (631 19 Ma) are found in the Paraíba do Sul metasedimentary rocks (Noce et al., 2004). Based on geochemical and structural studies of G1 granitoids and enclosing rocks, Geiger (1993) proposed a geodynamic model for the area with a two-stage island arc evolution and two plate collisions. G1 granitoids have been then interpreted to represent the roots of a second subduction related magmatic arc, which intruded and crystallized short after the peak of metamorphic conditions. Further detailed studies are necessary in order to separate the different G1 granitoids along the entire mobile belt.
G2 is another set of intrusions, locally with gradational contacts with the Paraíba do Sul complexes, which crops out in this region (unit 9 in Figure 2). It consists mainly of batholitic bodies of tectonically foliated, S-type, subalkaline to calc-alkaline, mostly peraluminous garnet-biotite granites (Pedrosa Soares & Wiedemann-Leonardos, 2000). Ghost migmatitic structures and ubiquitous rafts of migmatites and paragneisses indicate a predominant autochthonous to para-autochtonous nature. A widespread process of in situ migmatization of the metasedimentary pile originated a "sea" of diatexites and foliated granites. G2 makes up the anatectic core of the Araçuaí-West-Congo orogen.
Around 560 Ma, after significant cooling, a new reheating of the gneissic/granulitic crust took place and resulted in retrograde metamorphism, under high-amphibolite/granulite conditions with restricted partial melting which generated the G3 suite (anatexis II). The retrograde metamorphism is well illustrated by the replacement of sillimanite by cordierite and of pyroxene by biotite (Sluitner, 1989; Kühn et al., 2004) and the development of symplectitic and coronitic textures in granulitic/charnockitic rocks (Wiedemann et al., 1997). Although very similar to the G2 granitoids, the G3 suite is clearly younger and non-foliated, save when it inherits ghost foliation from the source rocks.
At the end of the orogenic cycle, around 535-480 Ma, discordant G5 plutons intrude the area and caused a new phase of anatexis (anatexis III). During the last 25 to 30 years, this voluminous plutonism and its enclosing sequences were the focus of several geological, geochemical and geochronological studies by our group and collaborators.
The next section details the syn-collisional G2 and the late-collisional G3 suites (Wiedemann et al., 1986; Wiedemann, 1993; Ludka et al., 1998; Pedrosa-Soares et al., 1999, Ludka & Wiedemann, 2000; Medeiros et al., 2000, 2001, 2003; Mendes et al. 1997; 1999).This is then followed by a section describing examples of the late post-collisional G5 suite.