Journal of Petrology Advance Access originally published online on October 9, 2007
Journal of Petrology 2007 48(11):2149-2185; doi:10.1093/petrology/egm055
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Calc-Alkaline Magmatism at the Archean–Proterozoic Transition: the Caicó Complex Basement (NE Brazil)
1Pós-GraduaÇão Em Geodinâmica E Geofísica And Departamento De Geologia, Ccet-Ufrn, Caixa Postal 1502, Cep 59078-970, Natal/Rn, Brazil
2Laboratoire Magmas Et Volcans, Opgc, Cnrs, Ird, Université Blaise Pascal, 5, Rue Kessler, 63038, Clermont-Ferrand Cedex, France
3Géosciences Rennes, Cnrs, Université De Rennes 1, 35042, Rennes Cedex, France
RECEIVED JULY 26, 2006; ACCEPTED AUGUST 15, 2007
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTIONThe Paleoproterozoic metaplutonic rocks of the Caicó Complex Basement (Seridó region, NE Brazil) provide important and crucial insights into the petrogenetic processes governing crustal growth and may potentially be a proxy for understanding the Archean–Proterozoic transition. These rocks consist of high-K calc-alkaline diorite to granite, with Rb–Sr, U–Pb, Pb–Pb and Sm–Nd ages of c. 2·25–2·15 Ga. They are metaluminous, with high YbN, K2O/Na2O and Rb/Sr, low ISr ratios, and are large ion lithophile elements (LILE) enriched. Petrographic and geochemical data demonstrate that they belong to differentiated series that evolved by low-pressure fractionation, thus resulting in granodioritic liquids. We propose a model in which the petrogenesis of the Caicó Complex orthogneisses begins with partial melting of a metasomatically enriched spinel- to garnet-bearing lherzolite (with high-silica adakite melt as the metasomatic agent), generating a basic magma that subsequently evolved at depth through fractional crystallization of olivine, followed by low-pressure intracrustal fractionation. A subduction zone setting is proposed for this magmatism, to account for both negative anomalies in high field strength elements (HFSE) and LILE enrichment. Mantle-derived juvenile magmatism with the same age is also known in the São Francisco and West Africa cratons, as well as in French Guyana, and thus the Archean–Proterozoic transition marks a very important continental accretion event. It also represents a transition from slab-dominated (in the Archean) to wedge-dominated post-Archean magmatism.
KEY WORDS: calc-alkaline; magmatism; NE Brazil; Paleoproterozoic; petrogenesis
|
INTRODUCTION |
|---|
In Earth history, the Archean represents the most important period of continental crustal growth. It was characterized by much higher heat production than today and, as a consequence, higher geothermal gradients, which resulted in the genesis of unique lithologies such as komatiites and massive volumes of tonalite–trondhjemite–granodiorite (TTG) magmas (Condie, 1981
; Taylor & McLennan, 1985; Martin, 1986, 1987; Nisbet, 1987). TTGs have strongly fractionated rare earth element (REE) patterns, with low heavy REE (HREE) contents (YbN 
8) and are devoid of significant Eu anomalies. Their K2O/Na2O is low such that, in contrast to classical calc-alkaline basalt–andesite–dacite–rhyolite (BADR) suites, their differentiation results in a Na2O enrichment defining trondhjemitic differentiation trends.
Based on petrological and experimental studies, as well as on geochemical modelling, the genesis of Archean TTG has been explained by partial melting of an Archean tholeiite transformed into garnet-bearing amphibolite or eclogite (Barker & Arth, 1976; Martin, 1986, 1987, 1993, 1994; Rapp et al., 1991, 2003; Rapp & Watson, 1995; Martin et al., 1997, 2005; Foley et al., 2002; Martin & Moyen, 2002). Although there is consensus about the tholeiitic nature of the source of Archean TTGs, the tectonic setting in which they were generated is still a subject of controversy. It has been interpreted as either slab melting in a subduction zone (Condie, 1981; Tarney et al., 1982; Martin, 1986, 1987; Rapp et al., 1991, 2003; Rapp & Watson, 1995; Foley et al., 2002; Martin & Moyen, 2002; Martin et al., 2005) or hotspot-related melting of underplated basaltic crust (Atherton & Petford, 1993; Wolde et al., 1996).
Most Archean K2O-rich granites with the isotopic signature of a mantle-derived source were emplaced at the end of the Archean (2·8–2·5 Ga), and intruded both greenstone belts and TTGs. They are also referred to as ‘sanukitoids’ (Stern, 1989; Stern & Hanson, 1991; Smithies & Champion, 1999) or ‘Closepet-type’ granites (Moyen et al., 2001, 2003). However, they display geochemical characteristics intermediate between Archean TTG (strongly fractionated REE patterns and low YbN contents) and modern juvenile continental crust [K and more generally large ion lithophile element (LILE) enrichment] and their petrogenesis is still under debate. Nevertheless, they generally appear to have been derived through variable extents of interactions between mantle peridotite and TTG magmas (Jayananda et al., 1995; Moyen et al., 1997; Smithies & Champion, 1999, 2000; Martin et al., 2005).
When compared with TTGs, post-Archean granitoids are richer in K; their compositions range from granodiorite to granite, with high YbN (>10) and negative Eu anomalies. A number of them with trace element and isotopic compositions of mantle-derived magmas are considered as having been generated in a subduction zone environment by partial melting of a fluid metasomatized mantle wedge. The dehydration of the subducted oceanic crust produces LILE-enriched fluids that interact with the overlying mantle wedge and initiate its melting, resulting in potassic calc-alkaline magmatism (e.g. Wyllie, 1983; Tatsumi, 1989; Hawkesworth et al., 1993; Keppler, 1996; Kogiso et al., 1997; Bureau & Keppler, 1999; Kessel et al., 2005).
In modern subduction zones, Archean TTG-like magmas can be generated when high geothermal gradients are achieved along the Benioff plane; for instance, during subduction of an actively spreading mid-ocean ridge. These magmas, referred to as adakites, are richer in Mg, Ni and Cr than Paleoarchean and Mesoarchean (>3 Ga) TTG but are very similar to Neoarchean (<3·0 Ga) TTG (Martin, 1999; Smithies, 2000; Martin et al., 2005). These differences are explained by assuming that the adakitic magma, once generated by partial melting of the subducted oceanic crust, interacts with the overlying mantle wedge and/or lowermost arc crust (Defant & Drummond, 1990; Drummond & Defant, 1990; Sen & Dunn, 1994a, 1994b; Rapp & Watson, 1995; Schiano et al., 1995; Maury et al., 1996; Stern & Kilian, 1996; Sigmarsson et al., 1998; Martin, 1999; Martin et al., 2005).
The Archean–Proterozoic boundary is marked by changes from a generalized high geothermal gradient and subsequent production of about two-thirds to three-quarters of the continental crust by accretion of juvenile magmas in the Archean, to a regime of lower and more diversified geothermal gradients, with predominance of crustal recycling during the Proterozoic (Taylor & McLennan, 1985; Martin, 1986, 1993, 1994; Sylvester, 1994). Although at the world scale it is possible to find evidence for a progressive change in TTG composition throughout Archean times (Martin & Moyen, 2002), the transition at 2·5 Ga was not sharp. On the contrary, it was progressive, such that some TTG are still known in Early Proterozoic terrains. In this context the orthogneisses of the 2·2 Ga Caicó Complex in NE Brazil provide an attractive opportunity to study calc-alkaline magmatism at this period of important petrogenetic changes. In addition, the Early Proterozoic is also characterized by a very significant accretion event, leading to the production of huge volumes of new juvenile continental crust; for example, in the São Francisco (Conceição, 1997; Teixeira et al., 2000) and West Africa (Boher et al., 1992; Toteu et al., 2001) cratons, and in French Guyana (Gruau et al., 1985; Delor et al., 2003). It also marks the formation of voluminous juvenile crust after a period of 
300 Myr (2·5–2·2 Ga), characterized by negligible magmatic activity and even lack of magmatism in several areas (Martin, 1993). In this context, the purpose of this paper is: (1) to describe all magmatic components of this juvenile transitional crust from NE Brazil; (2) to constrain its petrogenesis; (3) to discuss, in the light of these data, the Archean–Proterozoic transition and the subsequent Paleoproterozoic evolution.
|
GEOLOGICAL SETTING |
|---|
Tectonic framework
Almeida et al. (1981
) defined the Borborema Province in northeastern Brazil (Fig. 1), which consists of tectonic units stabilized during the Brasiliano orogeny (0·60 ± 0·05 Ga). This province developed after the convergence of the West Africa–São Luís and São Francisco–Congo cratons during the assembly of Western Gondwana at c. 600 Ma. In a pre-drift reconstruction, it extends from central and SE Brazil (Brasília–Ribeira mobile belt) to West Africa through the Trans-Sahara belt composed of the Cameroon, Nigeria and Hoggar shields (Caby, 1989).
|
This area has been studied for many years and several contrasting geodynamic reconstructions have been proposed (Almeida et al., 1981
; Caby, 1989; Bertrand & Jardim de Sá, 1990; Caby et al., 1991; Jardim de Sá, 1994; Van Schmus et al., 1995, 2003). Briefly, the Borborema Province consists of several supracrustal sequences deposited over an Archean to Paleoproterozoic gneissic basement that has been intruded by large amounts of Brasiliano-age granitoids. Jardim de Sá (1994) interpreted it as being made up of a number of allochthonous terrains that amalgamated just before and/or during the Brasiliano orogeny, and Santos (1996) noted that tectonic collages occurred in both the Cariris Velho–Kibaran (1·1–0·95 Ga) and Brasiliano–Pan-African orogenies in the so-called Transversal Zone.
A notable feature of this province is the complex system of crustal-scale high-temperature shear zones (Corsini et al., 1991; Jardim de Sá, 1994) that separate domains of variably strained massifs and supracrustal sequences. These were developed (and/or activated) during and after the collision between the West Africa, Congo and São Francisco cratons, and are closely associated with the emplacement of the Brasiliano granitoids (Caby et al., 1981; Bertrand & Jardim de Sá, 1990; Archanjo & Bouchez, 1991; Corsini et al., 1991; Jardim de Sá, 1994). The Patos and Picuí–João Câmara dextral shear zones are believed to accommodate the displacement of the Rio Piranhas massif toward the São José de Campestre massif, which resulted in transpression of the Seridó belt.
In this context, the Seridó domain (Fig. 2), situated to the north of the Patos lineament, comprises: (1) the Caicó Complex Basement; (2) supracrustal sequences of indeterminate age belonging to the Seridó Group [late Paleoproterozoic according to Jardim de Sá (1994) and Jardim de Sá et al. (1995), or Neoproterozoic following Hackspacher & Dantas (1997) and Van Schmus et al. (2003)]; (3) granitoids of both late Paleoproterozoic (the so-called G2 orthogneisses) and late Neoproterozoic ages and interpreted as having been derived from melting of an enriched lithospheric mantle or the lower continental crust, with variable crustal contamination and mixing (Leterrier et al., 1990, 1994; Jardim de Sá, 1994; Hollanda et al., 2003).
|
The Caicó Complex Basement
Field relationships
In the regional literature, the Caicó Complex corresponds to the high-grade basement of the Seridó Group, which forms an area of
60% (35 000 km2) of the exposed Precambrian units in the region studied (Fig. 2). It consists mainly of Paleoproterozoic meta-plutonic rocks, intruded and/or interlayered with older and subordinate meta-supracrustal rocks (Jardim de Sá, 1984, 1994; Hackspacher et al., 1990; de Souza et al., 1993). This association occurs in both the Rio Piranhas and São José de Campestre massifs; in the latter, Archean protoliths have also been identified (Dantas et al., 2004). The present paper essentially deals with the Paleoproterozoic orthogneisses, which are hereafter simply referred to as the Caicó Complex.
The older tectonic fabric in these orthogneisses is a high-grade banding (D1) associated with isoclinal to intrafolial folds and strong transposition, followed by an event of tangential kinematics (D2). D1 and D2 are usually interpreted as temporally distinct events (e.g. Jardim de Sá, 1984, 1994); the deposition of the Seridó Group and intrusion of the G2 orthogneisses occurred between D1 and D2. The age of the D2 event is also controversial; the c. 1·8 Ga age proposed by Jardim de Sá (1994), Jardim de Sá et al. (1995) and others has been challenged by the younger (Neoproterozoic) U–Pb detrital zircon and Sm–Nd model dates of the Seridó belt supracrustal sequences (Van Schmus et al., 2003). Recently, Hollanda et al. (2007) reported precise U–Pb sensitive high-resolution ion microprobe (SHRIMP) zircon ages of 2·20 ± 0·03 Ga for G2 orthogneisses in the Seridó region, and thus constrained the timing of the D2 event. The last tectono-metamorphic event (D3) is marked by transcurrent to oblique shear zones and emplacement of the late Neoproterozoic (Brasiliano) granitoids. The associated metamorphism ranges from upper amphibolite to granulite facies near plutonic intrusions and crustal-scale shear zones to greenschist facies in other places.
Geochronology and geochemistry
Hackspacher et al. (1990) and Van Schmus et al. (1995) published U–Pb data on zircons for gneisses and metagabbros from the São Vicente–Florânia region (Fig. 2), which gave ages in the range 2·16–2·13 Ga. For granodiorites of the Caicó area, Legrand et al. (1991) reported a whole-rock Rb–Sr isochron of 2·12 ± 0·08 Ma and a U–Pb zircon age of 2·24 ± 0·01 Ma. Available Sm–Nd data for metagabbros indicate TDM values of 2·76–2·62 Ga (Hackspacher et al., 1990; Dantas, 1992; Van Schmus et al., 1995). Whole-rock Rb–Sr isochrons of granitic gneisses and porphyritic granodioritic gneisses in both the São Vicente–Florânia and Açu areas give ages in the range 2·2–2·0 Ga, and ISr of 0·7041–0·7028 (Macedo et al., 1984; Jardim de Sá et al., 1987; Legrand et al., 1991; Dantas, 1992).
U–Pb zircon data from the Caicó Complex in the Santa Cruz region, to the east of the Seridó belt, yield an age of 2·18 ± 0·02 Ga (Dantas, 1996). In the São José de Campestre massif, Paleoproterozoic terrains surrounding the Archean domains and correlated to the Caicó Complex orthogneisses yield the following conventional and SHRIMP U–Pb zircon and Nd model ages (Dantas, 1996; Dantas et al., 2004): (1) 3·5 Ga tonalitic gneiss with TDM of 4·0–3·8 Ga; (2) 3·3 Ga grey monzogranitic gneiss with TDM of 3·7–3·1 Ga; (3) 2·7 Ga alkaline clinopyroxene-bearing syenogranitic gneiss with TDM of 3·5–3·2 Ga. However, no Archean terrain has been recognized to the west in the Rio Piranhas massif.
Geochemical studies of the Caicó Complex led to two groups of genetic interpretation: (1) the orthogneisses consist of Archean-like TTG suites formed by several pulses of magmatism and associated processes of magma mixing and mingling (Dantas, 1992; Petta, 1995), and significant contamination by crustal material accounts for their negative 
Nd values (Dantas, 1996); (2) the parental magmas were derived by partial melting of an enriched mantle; these melts then evolved by fractional crystallization with little or no interaction with the continental crust (Martin et al., 1990; de Souza, 1991; de Souza et al., 1993).
|
ANALYTICAL PROCEDURES |
|---|
In this paper, the modal composition has been established from an average counting of 1300 points for each individual thin section. Microprobe analyses were carried out at the Universidade de Brasília with a Cameca SX50 electron microprobe, operating at 15 kV accelerating voltage, 25 nA beam current, and 10 s counting time, using synthetic and natural minerals as standards. The analytical errors are within ±0·5–2% for SiO2, Al2O3, Fe2O3, MgO, MnO, CaO and TiO2, and 4·5–5·6% for Na2O and K2O.
Concentrations of major and trace elements for 61 samples were determined by X-ray fluorescence (XRF) at the Université de Rennes I with a Philips PW 1404 spectrometer, and seven other samples were analysed for trace elements by inductively coupled plasma mass spectrometry (ICP-MS) at the Université de Lyon. Analytical precision for major elements is within 2%, but may reach 10% for elements of low abundance (MnO, P2O5). Total iron is reported as Fe2O3. For trace elements, precision is better than 5%, except for elements present at concentrations <30 ppm, where the uncertainties are within 10%. The REE contents of nine samples were determined by ICP-MS at the Université de Nancy (n = 3) and the Université Blaise Pascal (n = 6). Concentrations of REE, Ta, U, Th, Hf and Sc in nine samples were measured by instrumental neutron activation analysis (INAA) at the Pierre Sue laboratory (CEN, Saclay). Details of the analytical methods have been given elsewhere (Govindaraju et al., 1976; Martin, 1987). Chondrite normalization values used for the REE are from Sun & McDonough (1989).
Rb contents were measured by isotope dilution with a Cameca THN-206 mass spectrometer at the Université de Rennes I. A Finnigan Mat 262 multicollector mass spectrometer was used to determine Sr content as well as isotopic ratios. Total blanks were as follows: 0·1 ng for Rb, 1 ng for Sr, and measurements of NBS standard 987 gave an 87Sr/86Sr value of 0·71025 ± 0·00001. Uncertainties of 87Rb/86Sr are within 2%, and 87Sr/86Sr ratios are quoted at 2
. Sr and Nd isotopic compositions measured at Clermont-Ferrand were determined by mass spectrometry with a Cameca THN-206 [analytical methods have been described by Pin & Paquette (1997)]. 87Sr/86Sr ratios were normalized to 86Sr/88Sr = 0·1194 (Faure, 1986), and 143Nd/144Nd ratios were normalized to 146Nd/144Nd = 0·7219. Single zircon analyses were performed at the Université de Rennes I using a Cameca THN-206 mass spectrometer and steps at 2·6, 2·8 and 3·2 A, following the procedure of Köber (1986). Decay constants and isotopic abundance ratios for all methods are those of Steiger & Jäger (1977). The ages, MSWD and errors were calculated using the Excel-based version 3 of Isoplot (Ludwig, 2003). All isotopic ratios and age calculations in this paper, as well as previously published data, were (re)calculated to a 2 error.
|
STRATIGRAPHY AND STRUCTURAL PATTERNS |
|---|
The Caicó Complex is composed of two units, a meta-volcano-sedimentary unit and a volumetrically dominant, meta-plutonic one. In the region investigated, the supracrustal sequences represent <6% of outcropping area; they mainly consist of garnet-bearing paragneisses and fine-grained amphibolites (meta-basalts and meta-andesites) together with intermediate to felsic gneisses (meta-rhyolites and meta-greywackes). Subordinate amounts of banded iron formations (BIFs), quartzites, marbles and calc-silicate gneisses are also found. The meta-supracrustal rocks may form 20–150 cm xenoliths included in the intrusive meta-plutonic rocks. The meta-plutonic rocks consist of (an estimation of the exposed area is indicated as a percentage of the total area of basement rocks): (1) quartz diorites and subordinate meta-gabbro and meta-ultramafic (hornblendites, serpentinites, steatites) bodies (
3%); (2) fine- to medium-grained tonalitic (28%) and granitic (11%) gneisses; (3) medium- to coarse-grained porphyritic granodioritic and granitic gneisses (52%). Basic to intermediate rocks, which are volumetrically subordinate, may form 100–500 m diameter stocks or, more commonly, occur as 10–200 cm enclaves within the granitoids, and as 1–5 m thick sheets in the meta-supracrustal rocks. Quartz diorites, which are volumetrically more abundant than gabbros, diorites and meta-ultramafic rocks, may contain small elliptical dioritic microgranular enclaves, and euhedral to rounded millimetre-sized plagioclase phenocrysts. Field relationships indicate that the tonalitic gneisses are intruded by augen gneisses, which are in turn intruded by granitic gneisses. In low-strain regions, dioritic, quartz dioritic, granodioritic, granitic and tonalitic gneisses have gradational, interlobate, or wedge-shaped contacts, the first two lithotypes corresponding to the less differentiated petrographic facies. All features and intrusive relationships described above indicate that the meta-plutonic rocks of the Caicó Complex are coeval intrusions, spatially related to each other and probably with a common, less evolved, basic to intermediate parental magma.
The most penetrative fabric (D2) is a metamorphic banding that overprints earlier magmatic fabrics (contacts between enclaves and more differentiated granitoid hosts; alignment of feldspar and amphibole phenocrysts). The D2 fabric is also marked on G2 granitoid sheets intruded into the interface between the Caicó basement and supracrustal rocks of the Seridó Group. The metamorphism associated with D2 is generally in upper amphibolite facies and of low to medium pressure, as indicated by paragenesis including cordierite ± sillimanite ± kyanite ± rutile in garnet-bearing paragneisses. Jardim de Sá (1994) and Jardim de Sá et al. (1995) ascribed the D2 event to a late Paleoproterozoic stage, based on the assumption of a syntectonic (syn-D2) emplacement of the G2 orthogneisses and meta-pegmatites, dated at 1· 9–1· 8 Ga according to U–Pb zircon and Rb–Sr isochron ages (Jardim de Sá et al., 1995); a U–Pb titanite age of 1· 97 ± 0·02 Ga from a Caicó Complex orthogneiss (Hackspacher et al., 1995) may be an indication of basement overprint during the D2 thermotectonic event. The D2 tangential fabrics are overprinted by NE–SW Brasiliano-age transcurrent (in the Rio Piranhas massif) and extensional (in the São José de Campestre massif) shear zones (D3). Near and inside the shear zones, amphibole, biotite and feldspar are dynamically retrogressed into epidote, carbonate, chlorite, actinolite and titanite.
|
PETROGRAPHY AND TEXTURES |
|---|
General characteristics
Table 1 shows the average modal compositions of 128 meta-plutonic rocks of the Caicó Complex from both the Rio Piranhas and São José de Campestre massifs. All samples were plotted in the Q–A–P (quartz–alkali feldspar–plagioclase) triangle (Fig. 3; Lameyre & Bowden, 1982
). The modal compositions of basic to intermediate rocks are gabbro and quartz diorite, respectively, which all follow a tholeiitic differentiation trend. Tonalitic gneisses plot along a low-K calc-alkaline (trondhjemitic) trend akin to the most evolved members of the Paleoproterozoic low-K gabbro–diorite–tonalite–trondhjemite series of SW Finland (Arth et al., 1978). Augen gneisses vary from granodiorite to syenogranite, with a few samples having monzodioritic and monzonitic compositions. In fact, both augen and granitic gneisses do not define real trends but rather plot on the medium-K to high-K calc-alkaline trends. Consequently, they are clearly different from typical Archean TTG, which have low-K affinity (Martin, 1987, 1994). On the other hand, they are very similar to Neoproterozoic K2O-enriched calc-alkaline granitoids as exemplified by rocks of the Armorican Massif (Graviou & Auvray, 1985; Graviou et al., 1988).
|
|
One outstanding feature of the Caicó meta-plutonic rocks is the abundance of ferromagnesian minerals, mainly clino-amphibole and biotite, which distinguishes them from the amphibole-poor typical Archean TTG (Martin, 1987
, 1994). In tonalitic, augen and granitic gneisses the less evolved facies are richer in amphibole and poorer in biotite than the more differentiated members; this feature emphasizes the role played by the fractionation of these phases at the beginning of differentiation. The regular variation of mafic and felsic minerals together with preserved igneous textures (clinopyroxene, amphibole, feldspar, titanite and apatite phenocrysts), absence of metasomatic replacement of K-feldspar by Na-plagioclase (Drummond et al., 1986) and conservation of magmatic geochemical trends (see below) all suggest that the mineral assemblage observed at present is the same as in the magmatic protoliths.
Basic to intermediate rocks (BIR)
According to their degree of recrystallization the basic to intermediate rocks of the Caicó Complex display granoblastic, nematoblastic, pokilitic and laminated textures. Based on modal composition, three main petrographic facies can be distinguished: (1) hornblende > biotite, the most widespread facies; (2) biotite > hornblende; (3) clinopyroxene + hornblende and rare biotite, subordinate to (1) and (2).
Clinopyroxene is a colourless or pale green diopside up to 2–5 mm long that is sometimes transformed into green amphibole or brown biotite. Amphibole often occurs as euhedral to subhedral polygonal aggregates; it is strongly pleochroic (X pale yellow, Z deep green to blue) and its length ranges from 0·5 to 4 mm. Its optical properties are those of common green hornblende, but chemical variation from Mg-hornblende to actinolite and XMg of 0·7–0·4 were reported by Petta (1995) in the São Vicente–Florânia region. Plagioclase (An25–40) appears as millimetre-sized phenocrysts with sharp contacts or rounded margins, commonly forming recrystallized polygonal mosaics.
Accessory minerals are: (1) grey to brown lozenge-shaped titanite phenocrysts (with quartz, amphibole, apatite, and biotite inclusions), intergranular crystals or small grains also following the cleavage of biotite, amphibole or clinopyroxene; (2) small light yellow prisms and irregular crystals of epidote, with frequent metamictic allanite core; (3) opaque minerals that occur as lamellae and quadratic or poikilitic grains associated to biotite and titanite; (4) apatite and zircon inclusions in clinopyroxene, amphibole and plagioclase. Plagioclase and biotite alteration occasionally and locally gives rise to carbonate and chlorite, respectively.
Augen gneisses (AG)
The augen gneisses are derived from porphyritic plutonic protoliths and have granoblastic to granonematoblastic textures. The most important feature is millimetre- to centimetre-sized (0·1–20 mm) augens of perthitic K-feldspar (microcline Or93Ab7, Table 2) and slightly zoned plagioclase (An22–30; Table 2). K-feldspar augens often contain inclusions of plagioclase, biotite, amphibole, titanite and zircon. Both types of augen can be deformed, recrystallized, and wrapped by quartz ribbons and new feldspar grains. Myrmekite and replacement of plagioclase by microcline is common between recrystallized aggregates as well as in pressure shadows near feldspar augens. Accessory minerals are: titanite (poikilitic or interstitial grains); pistacite-rich epidote (Pss = 37; Table 2) forming anhedral rims around metamictic allanite core, anhedral grains in reaction contacts with biotite and amphibole, or associated with saussuritization of plagioclase; oxides (usually bordered by epidote and titanite); and apatite and zircon included in other mineral phases.
|
Amphibole and biotite mark the main planar fabric (S2). The former occurs as anhedral to subhedral prismatic and strongly pleochroic grains (x yellowish green, z deep green), 0·1–2 mm in size. It consists of Ca-rich amphibole (Table 2) with (Na + K)A = 0·7, (Ca + Na)B = 1·8, Ti = 0·1 p.f.u., Fe3 + > AlVI and XMg = 0·4, and can be classified as magnesian–hastingsitic hornblende according to the classification of Leake (1978
). Some crystals contain inclusions of plagioclase, quartz, titanite, biotite and apatite. Biotite appears as isolated flakes, locally as inclusions or in reaction contact with amphibole, in this case associated with epidote and titanite. It has variable size (0·1–4 mm), strong pleochroism (X light yellow, Z deep yellow), with low Ti contents and XMg of 0·5 (Table 2), and can be classified as Fe-biotite.
Granitic gneisses (GR)
Granitic gneisses are mineralogically similar to augen gneisses, except that they contain smaller amounts of dark minerals (Table 1). Texturally, they are equigranular (1–2 mm) or slightly inequigranular, and microporphyritic. Plagioclase (An20–25) is slightly zoned or optically homogeneous. Biotite is brown and relatively rare, and colourless clinopyroxene (diopside) has also been scarcely observed in the less differentiated samples. Amphibole is green to blue with optical properties of common green hornblende. Epidote, opaque minerals, apatite and zircon are frequent accessory phases.
Tonalitic gneiss (TON)
Tonalitic gneisses are compositionally and texturally similar to granitic and augen gneisses except that they have little or no K-feldspar and are richer in mafic minerals. Plagioclase (An12–18) is slightly less calcic than in the augen and granitic gneisses (Table 2). Amphibole is strongly pleochroic ranging from brown to deep green, with other optical properties similar to amphiboles of the augen gneisses. Chemically, they have (Na + K)A = 0·7, (Ca + Na)B = 1·7, Ti = 0·1 p.f.u., Fe3 + > AlVI and XMg = 0·3, they are slightly Si- and Mg-impoverished when compared with amphibole from the augen gneisses, and they can be classified as hastingsitic hornblende (Leake, 1978). Biotite is slightly Ti-richer and Mg-poorer than in the augen gneisses.
P–T conditions of both emplacement and recrystallization
The Caicó Complex has been variably deformed and recrystallized under amphibolite-facies conditions. Despite this, mineral shapes and inclusion relationships allow us to distinguish between relicts of igneous textures and metamorphic features. The former are represented by plagioclase and K-feldspar, as well as amphibole and titanite phenocrysts. Generally, plagioclase, K-feldspar and amphibole are texturally strongly similar to feldspar and amphibole phenocrysts described in quartz diorite and granodiorites from well-preserved calc-alkaline granitoids (Graviou & Auvray, 1985; Graviou et al., 1988). Taking into account these points, we selected the less deformed and/or metamorphically recrystallized samples for microprobe study.
The Al-in amphibole geobarometer (Schmidt, 1992) and the plagioclase–hornblende geothermometer (Blundy & Holland, 1990) were used to constrain the P–T conditions of re-equilibration of amphibole (data from Table 2). Based on the experimental errors of the method (±0·6 kbar and ±75°C), the calculated P–T values are in the range 7·4–6·8 kbar and 732–705°C; they are the same for amphibole of both tonalitic and augen gneisses. This corresponds to the transition between the upper amphibolite to granulite facies, in the field of partial melting of water-saturated granitic systems. These values are consistent with both recrystallization of feldspar phenocrysts in meta-plutonic rocks and migmatization of the meta-pelitic components of the Caicó Complex. On the other hand, as coexisting amphibole and biotite have different XMg, overall chemical equilibrium was not achieved (Vynhal et al., 1991). It is proposed that the syntectonic emplacement and cooling of the meta-plutonic rocks occurred between 7·4 and 6·8 kbar and 732 and 705°C, which is consistent with all other field data, textural observations, and the mineralogical sequence described above. This corresponds to an average geothermal gradient of 30°C/km near the pluton contacts.
|
GEOCHRONOLOGY AND ISOTOPIC DATA |
|---|
Five samples of tonalitic gneisses from Caicó and Açu were analysed for Rb–Sr isotopic composition (Table 3). They yielded an age of 2229 ± 64 Ma with MSWD = 1·9 and initial 87Sr/86Sr (ISr) of 0·7023 ± 0·0005 (Fig. 4a). Single zircon from sample EV10A gave a 207Pb/206Pb age of 2181 ± 10 Ma (Table 4, Fig. 4a), which is within the error limits of the whole-rock Rb–Sr age. The zircon grains are idiomorphic, dark (metamictic?) to light brown, and may contain minute inclusions of apatite and fluid.
|
|
|
Seven samples of augen gneisses from Açu were analysed for the Rb–Sr whole-rock composition (Table 3). They define an isochron with an age of 2195 ± 62 Ma and ISr of 0·7027 ± 0·0009, with MSWD = 5·1 (Fig. 4b). Three-step heating of single zircon from sample EV12C gives similar results (Table 4, Fig. 4b), with an average 207Pb/206Pb age of 2179 ± 17 Ma, which is similar to the Rb–Sr age. The zircon grains are idiomorphic, usually concentrically zoned, colourless or light brown, with many apatite inclusions.
In the Santa Cruz region (Fig. 2), in the São José de Campestre massif, seven samples of the Caicó Complex were analysed for Sr and Nd isotopes (Table 5). For these samples, an Rb–Sr isochron yielded an age of 2144 ± 70 Ma, with ISr of 0·7025 ± 0·0005 and MSWD of 24 (Fig. 5a). The whole-rock Sm–Nd isochron with all points resulted in an extremely elevated error on age and MSWD (2253 ± 450 Ma and 189, respectively). The best fit is produced when samples ES56A, ES145 and ES196 are discarded. In this case, the Sm–Nd whole-rock isochron gave an age of 2216 ± 97 Ma, with INd of 0·50928 ± 0·00006, MSWD of 4·3 and Nd of –0·7 (Fig. 5b). The TDM ages vary from 2·69 to 2·53 Ga, and the Nd(t = 2·2 Ga) ranges from –1·87 to +0·02 (Table 5). Samples ES56A and ES145 have the highest titanite (2·2–3·1%) and apatite (1·0–1·2%) modal contents; sample ES196 has very low titanite (0·1%) and the highest zircon (0·6%). The reason for dispersion of the samples on the Sm–Nd isochron could be the cumulative nature of titanite, apatite and zircon. Indeed, these minerals have high distribution coefficients (>5) for Sm and Nd (e.g. Rollinson, 1993); consequently, addition of small amounts of them into the magma would significantly modify the initial Sm/Nd ratio, resulting in an erroneous estimation of TDM and Nd. Another reason for the dispersion of samples ES56A, ES145 and ES196 could be a slight difference in age and/or source.
|
|
Within the range of analytical errors, the whole-rock Rb–Sr isochron and U/Pb ages, and ISr ratios of the orthogneisses studied are similar. This reveals a comparable isotopic history, with parental magmas possibly derived from a common source. This conclusion is in agreement with the presence of rounded and elliptical enclaves of diorite within tonalitic gneisses, as well as the rounded or interlobate contacts between tonalitic, granitic and augen gneisses. These features typically indicate the prevalence of low viscosity contrast between enclaves and host magma and lead to the conclusion that they were comagmatic at the time of intrusion. The ages are then interpreted as emplacement ages at about 2·2 Ga for the plutonic protoliths of these orthogneisses.
|
PETROGENESIS |
|---|
Sixty-nine samples were analysed: nine basic to intermediate rocks (BIR), 13 tonalites (TON), 16 granites (GR) and 31 augen gneisses (AG), as well as one basic dyke (now transformed into orthoamphibolite) and two meta-volcanic rocks (one meta-andesite and one meta-basalt). The complete whole-rock analysis dataset is given as an Electronic Appendix (which may be downloaded from http://www.petrology.oxfordjournals.org), and summarized in Table 6. They are displayed according to increasing SiO2 content, the iron being expressed as Fe2O3t. Because of the low contents of water, all analyses were recalculated to a volatile-free basis, the loss on ignition being reported.
|
Geochemical characteristics
The main geochemical features of the rocks studied are presented in Fig. 6. In the A–F–M diagram (Fig. 6a), all groups plot within the calc-alkaline field defined by Kuno (1968), and scatter around the reference trondhjemitic trend delineated by Paleoproterozoic granitoids from SW Finland (Barker & Arth, 1976
). The calc-alkaline affinity is also shown in the (Na2O + K2O) vs SiO2 diagram (Fig. 6b); there all samples plot in the upper part of the sub-alkaline field (Rickwood, 1989), which also corresponds to the calc-alkaline field (MacDonald & Katsura; 1964). However, four augen gneisses (samples EV12E, EV12F, CA8, ES145) are alkali-enriched, such that they plot in the alkaline field. The K–Na–Ca triangle (Fig. 6c) discriminates the behaviour of Na2O and K2O. In this diagram, all samples define a trend that evolves from Ca-rich magmas towards the K apex. This classical calc-alkaline evolution is in strong contrast to typical Archean TTG, which evolves towards the Na apex following a trondhjemitic trend (Martin, 1993, 1994). This conclusion is corroborated by the normative An–Ab–Or triangle, where the whole series evolve toward the orthoclase (Or) apex, whereas Archean TTG is Na-rich and generally plots in the trondhjemitic and tonalitic fields (Fig. 6d). In the same figure, most GR and AG samples overlap the field of late Archean calc-alkaline granites (Sylvester, 1994; Jayananda et al. 1995; Moyen et al., 2003).
|
When all the meta-plutonic rocks of the Caicó Complex are plotted together in Harker diagrams for both major and trace elements (Fig. 7a and b), they show gentle differentiation trends, where most major (Al2O3, Fe2O3t, MgO, CaO, TiO2, and P2O5) and trace (Sr, Co, V, and Ni) elements are negatively correlated with SiO2; only K2O and Rb, despite some scatter, are positively correlated with SiO2. Some elements (Na2O, Zr and possibly Ba) define broken lines where positive correlation for SiO2 <
65% turns into negative correlation for SiO2 > 65%. Such broken lines are not consistent with mixing processes; these trends in the Harker diagrams cannot result from mixing between two magmas or from assimilation of older rocks by the magma, but are produced by magmatic differentiation (partial melting or fractional crystallization; Wilson, 1989). In this case, the break is due to changes in the fractionating mineral assemblage in the course of differentiation; for instance, a change from the fractionation of Al- and Na-poor phases (e.g. pyroxenes) towards Al- and Ca-rich phases (Ca-plagioclase, hornblende).
|
The basic to intermediate rocks (BIR) define two subsets with contrasting compositions. The subset I (VS1A, VS2A, VS1B, CA9, ES12) plots on the less differentiated portions of the general trend. The subset II (EV6D, VC52B, CA7, VC51D) deviates from the general trend by lower Al2O3, Na2O and Sr contents and higher MgO, TiO2, V, Co, and Cr contents; this deviation is not yet clearly understood.
As already pointed out in Fig. 6b, a group of augen gneisses clearly plot out of the general trend; they are characterized by higher contents of K2O, TiO2, P2O5, Nb and Ba and lower contents of MgO, CaO and Co. As their modal composition indicates that, compared with other augen gneisses, they are richer in alkali feldspar, titanite, apatite, and magnetite, it can be tentatively proposed that this enrichment results from the accumulation of minerals during magma differentiation. In this case these rocks would not represent pure magmatic liquids but rather magmatic liquid together with imperfectly extracted cumulate.
Figure 8a shows the REE patterns of diorites EV6D and ES12, as well as the associated meta-andesite EV6E and meta-basalt EV9C. All samples are light REE (LREE)-enriched (LaN = 62–143) with YbN of 8–14; this results in moderately fractionated patterns [(La/Yb)N = 9–18] with no significant Eu anomaly (Eu/Eu* = 1·1–0·9). Because of their high LREE contents, BIRs are more fractionated than the average of Enriched Archean Tholeiite [EAT; (La/Yb)N = 4·2; Condie, 1981].
|
Compared with the BIR, tonalitic gneisses are slightly LREE-richer (LaN = 71–199). However, because of generally lower Yb contents (YbN = 3·7–14), this results in more fractionated patterns [(La/Yb)N = 7·5–40]. In addition, tonalitic gneisses systematically display a slightly negative Eu anomaly (Eu/Eu* = 0·9) and a concave-shaped HREE end.
Granitic gneisses are LREE-rich (LaN = 116–380), with moderately high HREE (YbN = 8·2–9·6), with a systematic important negative Eu anomaly (Eu/Eu* = 0·4). These REE patterns are intermediate between those of Late Archean and modern juvenile granites (Fig. 8c).
Among augen gneisses, four samples (EV12C, EV12F, EV13E, ES145) have both high LREE contents (LaN = 168–464) and high HREE contents (YbN = 9·1–14·4); consequently, the general shape of the REE patterns is similar to that of granitic gneisses, but with a systematic negative Eu anomaly. One sample (EV13B) differs by its low Yb content (YbN = 4·7), resulting in (La/Yb)N = 46·7, similar to the average Late Archean calc-alkaline granites (Fig. 8d).
The REE overall patterns of the Caicó gneisses are intermediate between those of average Archean TTG and modern calc-alkaline granitoids (Martin, 1994); their average composition is very similar to that of late Archean granites (Fig. 8b; Sylvester, 1994). Sample ES104A has very low HREE contents (YbN = 1·1), which are lower than those of average TTG but are similar to those of HREE-depleted TTG (Martin, 1987). This could indicate the contribution of an older Archean crustal component in its genesis.
Mechanism of differentiation
Procedures
All the meta-plutonic rocks of the Caicó Complex have several similarities; the same geographical occurrence, similar ages of emplacement, analogous petrographic, geochemical and isotopic compositions, and, more particularly, the same Nd model ages. Consequently, they can be assumed to be contemporaneous and cogenetic; therefore, the main trend defined in Harker plots will be considered as being due to differentiation from a generally similar source protolith. As discussed above (Fig. 7a and b), some elements such as Na2O, Zr and possibly Ba define broken or curved trends, a characteristic that allows us to discard their derivation by mixing–mingling mechanisms, and instead indicates that they result from magmatic differentiation (partial melting or fractional crystallization), with a change of composition of the solid cumulate or residue with time.
As fractional crystallization, contrarily to partial melting, is a very powerful process to impoverish magmatic liquid in compatible elements, the discrimination between the two mechanisms will be based on the behaviour of these elements. Indeed, in a log (compatible element) vs log (incompatible element) plot, differentiated liquids produced by partial melting will show a sub-horizontal trend whereas fractional crystallization will give rise to a sub-vertical trend (Cocherie, 1986; Martin, 1987, 1994). Figure 7b shows that Sr, V, Co and Ni contents in magma decrease in the course of differentiation (with increasing SiO2), thus demonstrating their compatible behaviour, whereas positive correlations point to the incompatible behaviour of Rb and Ba. Figure 9 shows log (compatible element) vs log (incompatible element) diagrams (Ni vs Rb, Ni vs Ba, Co vs Rb, V vs Rb, and Co vs Rb), where, despite the small scatter of incompatible elements, the trends shown by the meta-plutonic rocks of the Caicó Complex are always vertical without any affinity to the sub-horizontal trend of partial melting. Consequently, it can be concluded that the main mechanism of differentiation is fractional crystallization.
|
The first step in quantification of fractional crystallization was based on major elements and used a classical mass-balance equation system that was solved using the algorithm of Störmer & Nicholls (1978
). The theoretical modelling was computed assuming the differentiation of a parental magma (CO) into a differentiated liquid (CL). The accuracy of the adjustment of the theoretical model to the data is expressed by 
r2 [=(mi – ci)2, where mi is the measured concentration and ci is the calculated concentration of oxide i]. The mineral compositions used in the calculations were those analysed in this study (biotite, amphibole, plagioclase, K-feldspar); the other phase compositions were taken from Deer et al. (1983).
The second step consisted of reintroducing the computed modal compositions (Xi) of the cumulate and the degree of crystallization in trace element modelling. The equation chosen for fractional crystallization is that of Rayleigh (1896): CL = COF(D–1), where CL is the concentration of a trace element in the differentiated liquid, CO is the concentration of a trace element in the parental magma, F = (1 – FC) (FC is the degree of crystallization, with 0 < FC < 1) and D is the bulk distribution coefficient. The partition coefficients (Kdi) used for D calculation [D = (Xi.Kdi)], were those compiled by Martin (1985, 1987), Rollinson (1993) and Nielsen (2007).
Quantification of fractional crystallization
Table 7 shows the results for both major and trace element modelling. To model the behaviour of the subset II of basic to intermediate rock (BIR), sample VC52B was chosen as CO and VC51D as CL, whereas for subset I CO and CL were VS1A and CA9, respectively. The best fit of computed model to analytical data is obtained for the crystallization of a mineral assemblage of hornblende and clinopyroxene for BIR subset I, and of hornblende, plagioclase and magnetite for BIR subset II; the degree of fractional crystallization (FC) is 80% and 30%, respectively. In BIR subset II the behaviour of trace elements and especially of Zr is accounted for only when small amounts (0·015%) of zircon are added to the cumulate.
|
Tonalitic gneiss crystallization was modelled assuming CO = ES56B and CL = EV10B (Table 7). The best statistical result (
r2 = 0·6) was obtained for 45% fractional crystallization of a cumulate composed of hornblende, plagioclase (An40), magnetite, and traces of zircon. A good agreement between model and analytical data is observed for all other trace elements, except Ba and Cr. The addition of 0·025% zircon to the cumulate is needed to account for Zr behaviour. In Fig. 7a and b, augen gneisses display broken or curved trends for Na2O, Zr and Ba; this indicates that the composition of the cumulate changed over the course of differentiation, and consequently the evolution of augen gneisses has been divided into two stages: stage (1) considers differentiation from SiO2 57% to 67%; in this case CO = VS1E and CL = CA3; stage (2) models liquid behaviour from SiO2 67% to 77%; from CO = CA3 to CL = EV13D. The results of modelling are given in Table 7. For both stages the computed cumulate is made up of the same major minerals (hornblende + plagioclase + magnetite). These cumulates differ only in their relative modal proportions, with more hornblende and less plagioclase in stage (1); the degree of fractional crystallization is also different: FC = 55% for stage (1) and 40% for stage (2). In stage (2), less than 0·4% of apatite and 0·07% of zircon are required to account for the behaviour of P2O5 and Zr, respectively. In both cases, the calculated liquid compositions fit the analytical data, except for Cr and Rb in stage (2).
The granitic gneisses have the same behaviour as the augen gneisses but as they do not have SiO2 contents as low as those of the less differentiated augen gneisses, the broken or curved trends are not so well marked. Their behaviour resembles the second stage of crystallization of augen gneisses. Samples ES56A and CA4 were chosen as representative of CO and CL. The modelling (Table 7) shows that granitic gneisses evolved by extraction of a cumulate similar to that of both tonalitic gneisses and augen gneisses (hornblende + plagioclase + magnetite) but with less hornblende and more plagioclase and magnetite. Here too, fractionation of 0·04% zircon is needed to account for Zr behaviour. All the calculated element concentrations fit the analytical data well, except Rb.
Role of assimilation and fractional crystallization (AFC)
Figure 10a (Rb/Sr vs Sr) and 10 (Sr/Y vs Y) shows the results of fractional crystallization modelling for the subgroups presented above. When some granitic and augen gneisses are excluded the analysed rocks perfectly fall on the computed fractional crystallization curves, which corroborates the results discussed above and presented in Table 7. However, many granitic and augen gneisses with SiO2 > 72% deviate from the calculated trends (Fig. 10a) which indicates that other processes took place in addition to ‘pure’ fractional crystallization. Indeed, some of these rocks have slightly negative Nd(t = 2·2 Ga) values that could reflect some kind of contamination of the parental magma with older crustal components. Indeed, Hildreth & Moorbath (1988) considered that melting of host rock, assimilation, storage, and homogenization (MASH) are expected in the lower crust or at the mantle–crust transition beneath a large magmatic centre. In this region, the basic magmas that ascent from the mantle wedge become neutrally buoyant, induce local partial melting of surrounding rocks, assimilate and mix extensively, and either crystallize completely or fractionate to the degree necessary to re-establish buoyant ascent, and then constitute starting points for subsequent fractionation and contamination. Crustal assimilation and concurrent fractional crystallization (AFC) is now widely considered as an important mechanism of evolution of mantle-derived magmas interacting with the lower and upper crust (DePaolo, 1981; Huppert & Sparks, 1985; Wilson, 1989; Stern & Kilian, 1996; Moyen et al., 1997, 2001). Below, we describe the modelling of AFC for the Caicó orthogneisses.
|
Mixing equations for trace elements and isotopic ratios were originally presented by Langmuir et al. (1978
) and subsequently by DePaolo & Wasserburg (1979) and DePaolo (1981), with reviews by Faure (1986) and Wilson (1989). For any trace element CL = CL°f + [r/(r – 1 + D)]C*(1 – f), where CL° is the concentration of the trace element in the original magma, CL is the concentration of the trace element in the contaminated magma, C* is the concentration of the trace element in the contaminant, r is the ratio of the rate of assimilation to the rate of fractional crystallization, D is the bulk distribution coefficient for the fractionating assemblage, f = F–(r–1+D)/(r–1), and F is the fraction of magma remaining. For any radiogenic isotope L = L° + (* – L°)[1 – (CL°/CL)f], where L, L° and * are isotopic ratios whose subscripts are defined above. AFC has been modelled using the same CL° as for perfect fractional crystallization calculations (BIR subset I: CL° = VS1A; BIR subset II: CL° = VC52B; TON: CL° = ES56B; GR: CL° = ES56A; AG (1): CL° = VS1E; AG (2): CL° = CA3; Table 7). The lower continental crust has been assumed to be the potential contaminant, the composition proposed by Rudnick & Fountain (1995) was taken for trace elements and that of Faure (1986) for Sr and Nd isotopic ratios, whereas the partition coefficients used for D calculation are those compiled by Martin (1987), Rollinson (1993) and Nielsen (2007). The computed models (Fig. 10) clearly indicate that some of the silica-rich granitic and augen gneisses compositions can be accounted for by assimilation of lower continental crust concomitant with fractional crystallization of mainly hornblende + plagioclase. This is well exemplified in the Rb/Sr vs Sr plot (Fig. 10a) where about 10 granitic and augen gneisses samples plot above the curves of ‘pure’ fractional crystallization. The effect of continental crust assimilation results in an efficient increase of the Rb/Sr ratio of magma, which accounts for the ‘deviant’ behaviour.
Source
Possible sources
Didier et al. (1982) proposed a classification of granitoids based on their source: M granitoids originate from a mantle source whereas C granitoids are continental crust derived. Following the earlier S- and I-type classification of Chappell & White (1974), Didier et al. subdivided the C-type into CI (crustal igneous source) and CS (crustal sedimentary source). In fact, as reviewed by Pearce (1996), the source of granitoids is a combination between two extreme end-members: the mantle and the continental crust. The mantle may be either asthenospheric or lithospheric, whereas the continental crustal sources may consist of igneous or sedimentary protoliths. In addition, for each source, the degree, temperature and depth of partial melting, as well as diverse kinds of interaction between mantle and crust, are highly variable, thus accounting for the great chemical variability of most granitoid magmas.
The mineralogical and chemical compositions of the Caicó Complex orthogneisses show that they all belong to the M-type granitoids: (1) their composition varies gradually from basic (gabbro or diorite, quartz diorite) to acidic facies (leucotonalites, granites); (2) hornblende is common, with sometimes relicts of clinopyroxene; (3) muscovite and aluminous silicates (cordierite, garnet, sillimanite) are totally absent; (4) microgranular mafic (hornblende-rich) enclaves are abundant; (5) normative corundum is <1·1%; (6) they mostly are metaluminous, with Shand's A/NCK ratios <1·1 and A/NK ratios >1·2; (7) they contain normative diopside or <1% of normative corundum; (8) they have low, mantle-like initial 87Sr/86Sr (0·7022–0·7027).
The geochemical characteristics outlined above are consistent with island arc and continental arc granitoid magmas (Maniar & Piccoli, 1989); they are similar to those of the classical calc-alkaline basalt–andesite–dacite–rhyolite (BADR) suites. In a multi-element diagram (Fig. 11a and b), it appears that although the compositions of tonalitic and augen gneisses show roughly parallel patterns, they are poorer in almost all elements when compared with the average Andean continental margin granitoids of Pearce et al. (1984). The dioritic gneiss EV6D and a meta-andesite EV6E also have patterns parallel to the Andean continental margin (ACM) granitoids although they are slightly LILE-poorer than the other gneisses (Fig. 11c). The Caicó gneisses are very distinct with respect to the average of within-plate granites, which are richer in all elements from Th to Yb (Fig. 11a–c) When plotted together, the Caicó Complex rocks show strong similarities, with parallel patterns (Fig. 11d), the dioritic gneiss and the meta-andesite being the LILE-poorer and the augen gneisses the LILE-richer.
|
In conclusion, and by analogy with Andean continental margin granitoids, a subduction-related tectonic setting could be proposed for the Caicó Complex meta-plutonic rocks. These orthogneisses are regarded as synorogenic intrusions and a magmatic arc setting is proposed for their generation and emplacement. Both experimental and theoretical arguments have led to consideration of the genesis of juvenile calc-alkaline magmas in modern subduction zones as a function of heat distribution between the subducted oceanic lithosphere and the overlying mantle wedge (Wyllie, 1979
, 1983; Martin, 1986, 1993, 1994, 1999; Peacock, 1990, 1993; Maury et al., 1996). The place where calc-alkaline magmas are generated is controlled by the interplay between dehydration and partial melting processes in the subducted slab, which in turn depends on its age and on the geothermal gradients. High geothermal gradients along Benioff planes (assumed to be the common Archean situation) would favour the partial melting of the subducted lithosphere at comparatively shallower depths (Stern & Futa, 1982; Martin, 1986, 1999; Defant & Drummond, 1990; Drummond & Defant, 1990; Rapp et al., 1991, 1999; Peacock et al., 1994; Morris, 1995; Maury et al., 1996; Prouteau et al., 1996; Stern & Kilian, 1996; Sigmarsson et al., 1998; Bourdon et al., 2002; Samaniego et al., 2002; Martin et al., 2005; Samsonov et al., 2005); whereas low geothermal gradients (as today) favour the partial melting of the mantle wedge metasomatized by fluids released by the dehydration of the subducted lithosphere (Wyllie & Sekine, 1982; Tatsumi, 1989; Schmidt & Poli, 1998; Bureau & Keppler, 1999; Manning, 2004; Schmidt et al., 2004; Bindeman et al., 2005; Kessel et al., 2005). To try to account for the mineralogy and geochemistry of the Caicó Complex orthogneisses, these two possible sources (oceanic crust basalt and mantle lherzolite) will be discussed.
Basalt (oceanic crust) melting
In the last 30 years, many basalt and amphibolite melting experiments have been performed (Helz, 1976; Beard & Lofgren, 1989, 1991; Rapp et al., 1991, 1995, 2003; Rushmer, 1991; Winther & Newton, 1991; Sen & Dunn, 1994a, 1994b; Wolf & Wyllie, 1994; Patiño Douce & Beard, 1995; Rapp & Watson, 1995; Zamora, 2000). Partial melting of low-K tholeiite under both water-saturated and water-undersaturated (dehydration melting) conditions leaves a residue made up of amphibole ± plagioclase ± pyroxenes ± magnetite ± ilmenite for pressures lower than 8 kbar, with garnet appearing at pressures greater than 10 kbar, and amphibole disappearing above 16 kbar (Beard & Lofgren, 1991; Rapp et al., 1991; Peacock et al., 1994; Sen & Dunn, 1994a; Rapp & Watson, 1995). In all these experiments, 825–1000°C is the common temperature range for 10–60% partial melting. The liquids formed are peraluminous (corundum >1·3, 1 <A/CNK <1·3) and vary from diorite to tonalite–trondhjemite and granodiorite. Dacitic or rhyolitic liquids coexist with amphibole, clinopyroxene, plagioclase and magnetite in the temperature range of 800–900°C, whereas andesitic to dacitic liquids coexist with amphibole, clinopyroxene and magnetite up to the thermal stability limit of amphibole at 1000–1050°C (Rapp et al., 1991; Rapp & Watson, 1995).
The hypothesis of genesis of the parental magma of the Caicó Complex orthogneisses by partial melting of tholeiites has been tested. The approach is the same as for crystallization modelling: first major element behaviour has been modelled using mass-balance equations and the Störmer & Nicholls (1978) algorithm, whereas the equilibrium melting equation CL = CO/[D + F(1 – D)] of Shaw (1970) has been used for trace elements. As shown above, the trends in Harker diagrams result from the differentiation of a parental magma by fractional crystallization and AFC processes. Consequently, the melting modelling will not attempt to account for differentiation trends but only for the composition of the less differentiated parental magmas (i.e. with <63 wt % SiO2). The melting of two different sources has been computed: (1) a low-K EAT (SiO2 = 50·2 wt %, Mg-number = 53, K2O/Na2O = 0·2, LaN = 55, YbN = 13, reported by Condie, 1981); (2) the enriched tholeiite sample EV9C (SiO2 = 51·0 wt %, Mg-number = 56, K2O/Na2O = 0·4, LaN = 191, YbN = 11). The modelling leads to residues composed of hornblende ± clinopyroxene ± garnet ± magnetite and to degrees of partial melting ranging from 40 to 55%. However, augen gneiss sample VS1E requires a higher degree of partial melting (65%) and a different residue (clinopyroxene + orthopyroxene + magnetite). Because, in andesitic to dacitic liquids, 
and 
, the magma must be impoverished in Y and Yb with respect to the solid source, which also results in too high (La/Yb)N and Sr/Y in magma. All the computed models, with or without residual garnet, predict Yb and Y impoverishment in magma whereas Yb and Y enrichment is required for the Caicó Complex orthogneisses (Figs 12a and b), and consequently, unlike Archean TTGs and modern adakites, melting of a hydrous tholeiite does not appear to be a realistic source for the Caicó Complex magmas. In addition, the high Yb and Y contents preclude garnet as a significant residual phase. It must also be noted that the less evolved Caicó Complex samples may not be exactly the parental magmas but that they could also have undergone small degrees of fractional crystallization. As shown above (Fig. 10), the crystallization of an assemblage made up of hornblende and plagioclase would result in a decrease of the Y and Yb content in the magma. Consequently, the parental magmas were probably Y- and Yb-richer than the less evolved samples of the Caicó Complex, thus making even more unrealistic their origin by melting of a basalt tholeiitic source.
|
Lherzolite (mantle) melting
Earlier experimental melting of lherzolite generated liquids that varied in composition from basalt to dacite. Some researchers considered that silicic liquids would be primary magmas (Kushiro et al., 1972
; Kushiro, 1974; Mysen & Boettcher, 1975; Tatsumi, 1981), whereas others (Nicholls & Ringwood, 1972; Green, 1973) believed that andesitic and dacitic magmas could not be produced by direct melting of mantle peridotite, the most likely explanation being that they formed from a parental basic magma that evolved by fractional crystallization of olivine at depth. Experimental silicic liquids were generated under hydrous conditions for 1025–1150°C and 10–26 kbar with about 20–30% partial melting (Kushiro et al., 1972; Nicholls & Ringwood, 1972; Green, 1973; Kushiro, 1974; Mysen & Boettcher, 1975) or under water-undersaturated conditions for a similar P–T range (Tatsumi, 1982).
In the last 15 years, improvement in experimental techniques has allowed researchers to circumvent quenching problems and analyse liquids formed by small-degree (<5%) of melting (Baker & Stolper, 1994; Baker et al., 1995; Hirose, 1997; Robinson et al., 1998; Wasylenki et al., 2003). Experimental melting (Baker et al., 1995) of fertile peridotite at low pressure (<15 kbar) gives silica-rich (>55 wt % SiO2) near-solidus melts that are also alkali-rich. Experimental melting of the fertile peridotite KLB-1 (Hirose, 1997) for both water-undersaturated and water-saturated conditions generated high-silica (54–60 wt % SiO2) and high-magnesian (MgO = 5·6–6·8 wt %) liquids for temperatures of 1000–1050°C. For T > 1100°C, the liquids formed are basaltic in composition. On the other hand, melting of depleted peridotite generated low-silica and low-alkali basaltic liquids (Robinson et al., 1998; Wasylenki et al., 2003).
The genesis of the parental magma of the Caicó Complex orthogneisses by partial melting of the mantle has been modelled and the results are presented in a (La/Yb)N vs YbN plot (Fig. 13). Three mantle compositions were tested: (1) DM (depleted mantle), with (La/Yb)N = 1 and YbN = 2 (Martin, 1985); (2) EM [slightly enriched (fluid metasomatized) mantle], with (La/Yb)N = 6·6 and YbN = 3·4 (Martin, 1985; Graviou et al., 1988; Graviou & Auvray, 1990); (3) RS1, a phlogopite- and pargasite-bearing lherzolite representing the lithospheric mantle, with (La/Yb)N = 12·2 and YbN = 2·4 [sample RS1 of Menzies et al. (1987)]. In each group of gneisses, the sample analysed for REE and having the lowest SiO2 and the highest MgO contents has been chosen as representative of the parental liquid EV6D (diorite), ES56B (tonalitic gneiss), ES145 (augen gneiss), and ES56A (granitic gneiss). One meta-basalt (EV9C) and one meta-andesite (EV6E) of the supracrustal component of the Caicó basement were also plotted.
|
Figure 13 shows the curves for partial melting of spinel-bearing lherzolite (5% spinel) and garnet-bearing lherzolite (2, 3 and 6% garnet). It appears that partial melting of a depleted mantle, whatever the residual mineral assemblage, cannot generate magmas with La/Yb as high as in the Caicó Complex; consequently, the more likely source seems to be an enriched mantle. The genesis of the parental magma of diorite (EV6D), tonalite (ES56B), granite (ES56A) and meta-basalt (EV9C) is achieved for
10% partial melting of the enriched lherzolite EM leaving 2–3% garnet as residual phase; the augen gneiss (ES145) and the meta-andesite (EV6E) would require 8 and 20% partial melting, respectively. However, primary magmas derived directly from partial melting of lherzolite are believed to be basaltic, having high Mg-number (>70), Ni (>400 ppm) and Cr (>1000 ppm) with SiO2 < 50 wt % (Wilson, 1989), or Mg-andesites (references above), which is not the case of the Caicó orthogneisses. Consequently, we admit that a basic magma, once formed by melting of the enriched mantle, evolves by fractionation of olivine at mantle depth or during ascent to the lower continental crust to form the parental magmas of the Caicó orthogneisses. Fractionation of olivine does not modify the La/Yb ratio but would impoverish the liquid in magnesium, thus providing a better fit of the model to the analysed samples. The parental magmas of augen (ES145) and granitic (ES56A) gneisses, as well as the meta-basalt (EV9C), have higher La/Yb, which would require either more garnet in the residue or a lower degree of melting of the same source, or a more enriched mantle source with chemical characteristics similar to those of RS1.
Assuming present-day 87Sr/86Sr and Rb/Sr ratios of 0·7045 and 0·031 for the Bulk Silicate Earth (BSE; Workman & Hart, 2005), the 87Sr/86Sr ratio would be 0·702 at 2·2 Ga. The ISr values of the gneisses of the Caicó Complex range from 0·702 to 0·703, which are very close to the mantle value at 2·2 Ga. This clearly precludes an origin of the parental magma of the Caicó Complex by direct recycling of an older (Archean) continental crust, and consequently this also precludes large-scale crustal contamination. This is corroborated by the Nd vs 87Sr/86Sr at t = 2·2 Ga plot (Fig. 14). In this diagram, we calculated mixing curves between a mantle (MORB-like) derived magma and the upper continental crust (UCC), the lower continental crust (LCC), and silica-rich 2·7 Ga (Arch1) and 3·3 Ga (Arch2) gneisses of the São José de Campestre massif [U/Pb zircon data after Dantas et al. (2004)]. For modelling we used the mixing equations of Langmuir et al. (1978) and DePaolo (1981), with MORB, UCC and LCC trace element and isotopic compositions from Faure (1986) and Rollinson (1993); for Arch1 and Arch2 contaminants we used our unpublished data. It is not possibile that the UCC could be a contaminant. If mixing or contamination occurred it would involve less than 3% of LCC (see detail in Fig. 14b). In addition, the Nd (t = 2·2 Ga) values of +0·3 to –1·9 of the Caicó Complex are significantly different from and greater than those of the Archean gneisses of the São José de Campestre massif [in the range –10 to –17 at 2·2 Ga; data from Dantas et al. (2004)], thus corroborating that older continental crust did not play any significant role in the genesis of the Caicó Complex. Moreover, 3·3–2·7 Ga gneisses of the São José de Campestre massif are (our unpublished data) silica-rich (70–75 wt % SiO2) and Mg-poor (MgO = 0·1–0·2 wt %) and obviously could not be the source of diorites, quartz diorites, tonalites and granodiorites of the Caicó Complex.
|
Discarding any significant contribution of an older Archean oceanic or continental crust leads to the conclusion that the LILE- and LREE-rich nature of the Caicó Complex orthogneisses is a characteristic of the mantle source. To accommodate their arc signature, the LILE and LREE enrichment, variable
Nd (in the range +0·3 to –1·9 at 2·2 Ga; Table 5), and an enriched mantle source (Fig. 13) we considered the possibility of slab-modified peridotite as the source of the Caicó magmas.
This mechanism has already been proposed to account for the genesis of Archean sanukitoids (e.g. Rapp et al., 1999; Martin et al., 2005). Although admitted as generated by direct partial melting of an LILE- and LREE-enriched peridotite because of their trace element contents, the Nd isotope composition of sanukitoids requires a depleted mantle source (Stern et al., 1989; Stern & Hanson, 1991; Stevenson et al., 1999; see also reviews by Martin et al., 2005; Rollinson, 2006). In an experimental study at 4 GPa, Rapp et al. (1999) allowed the infiltration of an adakitic melt into an overlying peridotite layer, simulating melt–rock interaction at the subducted slab–mantle interface. The hybridization of slab-derived melts by reaction with mantle peridotite produced high-Mg adakitic liquids. Figure 15 shows Mg-number vs SiO2 and Sr/Y vs Mg-number diagrams in which the experimental results of Rapp et al. (1999) are reported. In addition, the composition of liquids produced by experimental melting of both depleted and enriched mantle peridotite (Takahashi et al., 1993; Baker & Stolper, 1994; Hirose, 1997; Hirschmann et al., 1998; Robinson et al., 1998; Wasylenki et al., 2003) as well as the average of Archean sanukitoids (Martin et al., 2005) are also shown. Obviously, the hybridized melts have Mg-number and Sr/Y ratio significantly higher than in the gneisses of the Caicó Complex. As discussed above (e.g. see Fig. 12), the field of experimental slab melts does not fit the composition of the gneisses of the Caicó Complex, especially when the Sr/Y ratio is considered.
|
Taking into account the discussion above, we calculated the composition of a depleted mantle (Workman & Hart, 2005
) metasomatized by slab melts having the composition of the high-silica adakite (HSA) of Martin et al. (2005). The best results are obtained for a metasomatized mantle (MM) formed by mixing 93% DM and 7% HSA (0·93DM:0·07HSA). Figure 16 shows that all of the Caicó Complex samples have patterns generally parallel to MM, thus providing an additional argument in favour of this common source. Modelling has been performed, assuming a two-stage evolution: (1) partial melting of a MM; (2) fractional crystallization of olivine in the generated magma. The input partition coefficients are those compiled by Rollinson (1993) and Nielsen (2007); the source is assumed to have phlogopite (2%), amphibole (2%), garnet (0–5%) and spinel (5–0%) as accessory phases. The best fit is obtained for 10% melting of peridotite with up to 3% garnet, followed by 50–80% of olivine fractionation. The predicted curves are well adjusted for tonalitic (Fig. 16a), granitic (Fig. 16c) and quartz dioritic gneisses (Fig. 16d). Only augen gneisses show slightly different patterns, which have Rb to Ce and Zr to Sm normalized values greater than the expected ones (Fig. 16b). If we assume that the distribution coefficients used are correct, this could reflect either a lesser degree of melting of MM or a greater amount of olivine fractionation.
|
Figure 17 summarizes the petrogenetic model for the Caicó Complex. Four stages are considered: in the first stage, a depleted mantle lherzolite is metasomatized by a slab-derived melt with high-silica adakite chemistry and possibly generated during an earlier (Late Archean?) subduction episode, giving rise to an enriched mantle (MM); in the second stage, 10–15% partial melting of this MM generates a basic magma that, in the third stage, after 40–80% fractional crystallization of olivine at depth produces the less evolved samples of the Caicó Complex, which, in the fourth stage, evolve by low-pressure intracrustal fractionation of variable proportions of hornblende, plagioclase and magnetite, with eventual AFC for some silica-rich augen and granitic gneisses samples.
|
In conclusion, the geochemical modelling shows that the parental magmas of the Caicó Complex orthogneisses could have been generated by partial melting of LREE- and LILE-enriched lherzolite with minor amounts of, or no, residual garnet, followed by olivine fractionation at depth. This petrogenetic model is very similar to that proposed for late Archean sanukitoids and Closepet-type granites: remelting of a peridotite previously metasomatized by reaction with slab melts (Shirey & Hanson, 1984
; Stern, 1989; Stern & Hanson, 1991; Rapp et al., 1999; Smithies & Champion, 1999, 2000; Moyen et al., 2001, 2003; Halla, 2005; Lobach-Zuchenko et al., 2005; Martin et al., 2005). In this case, melting of LREE-enriched peridotite is assumed to generate diorites, monzodiorites and syenodiorites with high Mg-number, Ni, Cr, Sr, Ba, P2O5 and LREE (Stern et al., 1989; Stern & Hanson, 1991). Subsequent differentiation of these melts would yield granodiorite with the following characteristics (at 65 wt % SiO2): (1) abundant hornblende, titanite and apatite; (2) Mg-number50, MgO > 3 wt %; (3) Sr and Ba 1000–2000 ppm, Cr 130–50 ppm, Ni 70–30 ppm; (4) Rb/Sr <0·1; (5) fractionated REE patterns with only minor Eu anomaly. These are characteristics shared by most of the Caicó Complex orthogneisses, except for their slightly lower Mg-number, MgO, Ni and Sr, and higher Rb/Sr and Cr. One important point to emphasize is that not only do the parental magmas of felsic orthogneisses appear to have had an enriched mantle source, but so also do the associated less silicic samples of unambiguous mantle origin (Table 6), such as (1) coarse-grained younger amphibolite SV3 (Mg-number = 69, Ni = 284 ppm, Cr = 783 ppm), and (2) meta-basalt (fine-grained amphibolite) EV9C (Mg-number = 56, Ni = 104 ppm, Cr = 406 ppm). Amphibolite SV3 is a dyke crosscutting granodioritic augen gneisses in the São Vicente–Florânia region and it seems to be affected by the same deformational history as the other Caicó Complex units. Meta-basalt EV9C forms metre-thick intercalations within meta-andesites, meta-rhyolites and garnet-bearing paragneisses in the Açu region (Fig. 2). It follows that the production of LILE- and LREE-enriched mantle-derived magmas was a recurrent phenomenon during Paleoproterozoic times, as found in earlier meta-basalt EV9C and late amphibolite SV3, which pre- and post-date the emplacement of the meta-plutonic rocks.
|
DISCUSSION |
|---|
The processes and timing of formation of continental crust have been controversial, and a number of mechanisms have been proposed, such as addition of new material from the mantle, re-addition of crustal material that has been cycled through the mantle, and redistribution of crustal rocks as a result of sedimentary and tectonic processes (see reviews by Condie, 1989
; Rudnick, 1995; Kemp & Hawkesworth, 2003). It appears that collision of arcs and aggregation of microcontinents are the major mechanisms by which continents have grown (Condie, 1989; Drummond & Defant, 1990; Davidson & Arculus, 2006). However, alternative models, such as delamination of continental lithospheric mantle (Rudnick, 1995), underplating of basaltic magma at the base of the continental crust (McCulloch, 1987; Rudnick & Fountain, 1995), intralithospheric differentiation (Taylor & McLennan, 1985; Neves et al., 2000; McLennan et al., 2006), and mantle plumes (Abbott, 1996; Condie, 2001), have also been proposed.
Much debate also concerns the steady or episodic nature of the continental growth. The episodic growth of juvenile crust has been recognized during the last 15 years, with major events of rapid crustal growth at 3·6, 2·7 and 1·8 Ga according to McCulloch & Bennett (1994), or at 2·7, 1·9 and 1·2 Ga according to Condie (1998, 2000). The episodic pattern of continent formation led Albarède (1998) to postulate mantle plume periodicity in addition to continuum of subduction zone activity. The close temporal links between mafic volcanic rocks, supposed to represent products of mantle plumes, pre-dating silica-rich syn-tectonic plutons, in Paleoproterozoic terrains of French Guyana (Vanderhaeghe et al., 1998; Delor et al., 2003) and West Africa (Abouchami et al., 1990; Boher et al., 1992; Béziat et al., 2000), has led researchers to admit mantle plume activity associated with subduction processes. In both regions, the major event of juvenile crust formation was completed in less than 50 Myr. Regardless of the crustal growth process, there is a consensus that the Archean–Proterozoic boundary corresponds to a major change in terrestrial geodynamic conditions (rapid crustal growth, which may or may not be related to falling geotherms in the Late Archean) that also resulted in changes in continental petrogenesis (Taylor & McLennan, 1985; Condie, 1989; McLennan et al., 2006).
In the model presented here, the parental magma of the Caicó Complex orthogneisses is interpreted as subduction-related. Major and trace element, and Nd isotope contents all agree with a metasomatized mantle as the source. The metasomatic agent was modelled as high-silica (TTG-like) slab-derived melt that hybridized with the depleted mantle. Of course, adakitic melt requires a previous episode of subduction (Moyen et al., 2001; Martin et al., 2005). The timing of the subduction should be somewhere between 2·7–2·5 Ga (TDM values for the Caicó Complex) and the emplacement age of plutonic rocks at 2·2 Ga (U–Pb and Pb–Pb zircon, and whole-rock Rb–Sr ages). However, assuming the enriched nature of the Caicó source, it is possible to estimate model age using, instead of depleted mantle, a chondritic mantle (e.g. CHUR, 147Sm/144Nd = 0·1967; Jacobsen & Wasserburg, 1980). The calculated TCHUR ranges from 2·4 to 2·2 Ga, thus indicating that the subduction-related enrichment of the mantle peridotite took place 100–200 Myr before the emplacement of the parental magmas of the Caicó Complex. When tentatively correlated with the West Africa craton (Abouchami et al., 1990; Boher et al., 1992) and French Guyana shield (Delor et al., 2003), this subduction would have followed an earlier episode of plume-related oceanic plateau magmatism (interpreted for juvenile mafic magmatism in both regions). Nevertheless, until today, no evidence of this plume event has been found in northeastern Brazil.
Table 8 summarizes the general features of the Caicó Complex orthogneisses compared with Archean TTG, calc-alkaline granites, adakites and sanukitoids, as well as with modern juvenile granitoids. Most petrographic and chemical characteristics of the plutonic series of the Caicó Complex are clearly distinct from Archean TTG, particularly in their cogenetic association with basic and intermediate rocks, their wide compositional range in SiO2, higher YbN, Rb/Sr, Cr/Ni and K2O/Na2O, and lower Mg-number (but basic to intermediate rocks), A/CNK, (La/Yb)N and Zr/Sc. The sources envisaged are also distinct: the Archean TTGs were derived by garnet-bearing amphibolite or eclogite melting, whereas the Caicó orthogneisses were derived from metasomatized lherzolite with little or no residual garnet (<5%). The Caicó orthogneisses are different from typical Archean sanukitoids by having higher K2O/Na2O, Rb/Sr and Cr/Ni ratios and less fractionated REE patterns. On the other hand, Archean calc-alkaline granites have lower YbN and Cr/Ni and higher (La/Yb)N.
|
Major and trace element modelling points to a four-stage evolution. After a first stage of assimilation–reaction of a depleted mantle with a slab-derived adakitic melt, this hybridized spinel- or garnet-bearing source is melted, generating a basic magma (second stage), which subsequently evolves by fractional crystallization of olivine to form the parental magmas to the Caicó Complex (third stage). Subsequently, fractional crystallization at low pressure (lower crust) of different proportions of amphibole + plagioclase + magnetite ± clinopyroxene gives rise to the differentiated Caicó Complex suites. According to this model, melting should have taken place at the spinel lherzolite–garnet lherzolite transition at pressure <20–25 kbar or equivalent depths of
66–83 km (Takahashi & Kushiro, 1983; Green & Falloon, 1998). All these characteristics are widespread in magmas generated from partial melting of enriched shallow mantle in continental arc settings and involving the sub-continental lithosphere (Pearce & Parkinson, 1993).
As indicated by our modelling, the lherzolitic source of the Caicó Complex was already LILE-enriched and had Ta–Nb, Sc, Ti and Yb negative anomalies. Ta, Nb and Ti anomalies are generally considered as typical features of magmas generated in subduction-like tectonic setting (see reviews by Pearce, 1982; Wilson, 1989). Several explanations have been proposed to account for these anomalies: (1) interaction between a fertile arc derived fluid and a depleted peridotite (Kelemen et al., 1990; Schiano et al., 1995); (2) infiltration of a rutile-saturated, slab-derived melt or vapour through a depleted peridotite produced by a previous episode of MORB extraction (Ryerson & Watson, 1987; Thirlwall et al., 1994); (3) presence of residual Ti-bearing minerals with high-
high field strength elements (HFSE) such as titanite, rutile, ilmenite, amphibole or garnet in the source (Green & Pearson, 1986; Ryerson & Watson, 1987; Hoffman, 1988; Drummond & Defant, 1990), that retain the HFSE, producing HFSE-impoverished melts.
Based on Nd of –2·5 to –3·7 at 2·2 Ga, Hackspacher et al. (1990) and Van Schmus et al. (1995) considered that a crustal component played an important role in the genesis of the Caicó Complex magmas. This interpretation, based only on Nd values, is clearly in contrast to the trace element signatures discussed here, which suggest instead an enriched mantle source with very little or no crustal contamination. Consequently, the Nd(t = 2·2 Ga) of +0·3 to –1·9 (see Table 5) should reflect the enriched nature of the source rather than contamination with older continental crust. Geochemical characteristics (metaluminous rocks, wide SiO2 range, a very low proportion of garnet, or no garnet, in the source), geochronological data (U–Pb, Pb–Pb and whole-rock Rb–Sr and Sm–Nd isochrons with similar ages; no inherited zircon) and comparison with experimental results all show that the Caicó Complex orthogneisses mainly represent juvenile magmatism, with no, or very subordinate, crustal contribution.
Paleoproterozoic gneisses form c. 38% (155 760 km2) of the exposed surface of the Precambrian rocks in NE Brazil. However, as the Neoproterozoic plutons, which make up about 34 800 km2 exposure, have Nd isotope signatures indicating a major contribution by 2·4–1·9 Ga sources (Neves, 2003), and 2·2 Ga detrital zircon in Meso- to Neoproterozoic supracrustal belts (Van Schmus et al., 2003), the reconstituted Paleoproterozoic crust should represent more than 46% (190 560 km2) of the exposed Precambrian units. The continental crust in NE Brazil has been modelled by gravity and isostasy studies by Castro et al. (1997a, 1997b), who concluded that it is 30 km thick. Seismic refraction data also indicate a somewhat similar crust thickness (34 km) in West Africa (Dorbath et al., 1986). This estimated thickness should be considered a minimum value, as at least 23 km (considering emplacement at about 7 kbar for the Caicó orthogneisses; see Table 2) have been eroded and incorporated into younger supracrustal belts (e.g. Seridó) and siliciclastic components of Phanerozoic cover. This indicates that a significant volume (10 x 106 km3 for a 53 km thick crust) of magma formed at 2·2 Ga, corresponding to the Caicó Complex.
Juvenile magmatic rocks of about 2·18–2·1 Ga age and a less voluminous extent of 2·35–2·2 Ga age cover huge areas extending for thousands of square kilometres. They are also known in the northeastern São Francisco (Conceição, 1997; Teixeira et al., 2000), São Luís (Klein et al., 2005b) and West African (Doumbia et al., 1988; Abouchami et al., 1990; Boher et al., 1992; Toteu et al., 2001; Gasquet et al., 2002; Feybesse et al., 2006) cratons, French Guyana (Gruau et al., 1985; Vanderhaeghe et al., 1998; Delor et al., 2003; Ledru et al., 2003; McReath & Faraco, 2006) and NE Brazil (Fetter et al., 2000; Martins & Oliveira, 2003; Neves, 2003; Klein et al., 2005a; Neves et al., 2006). Based on geological and geochronological correlations, Neves (2003) interpreted that the cratons (São Francisco, Congo, West African, Amazonian) and the neighbouring Brasiliano–Pan-African belts (Borborema, Brasília–Ribeira, Nigerian) as part of the Atlantica supercontinent, which accreted at the end of the Eburnean cycle (2·0 Ga). In the Birimian terrains of West Africa, Boher et al. (1992) concluded that juvenile crust formation spanned <50 Myr, a conclusion based on the similarity between U–Pb and Rb–Sr (2·19–2·16 Ga) ages, Sm–Nd ages (TDM = 2·34–2·14 Ga in magmatic rocks) and synchronous metamorphism (isochron with 2·2 Ga in garnet-bearing pelite). In these areas, granite–greenstone-like associations were formed, and all 2·2–2·1 Ga magmatic rocks have been derived from a depleted mantle source, with Nd(t = 2·2 Ga) in the range +0·4 to +6·8, which drastically differs from our conclusions for the Caicó Complex.
Consequently, it can be proposed that this specificity could reflect local mantle heterogeneities, an enriched mantle source being located under NE Brazil. This assumption is strongly supported by the fact that at the Proterozoic–Paleozoic boundary (Brasiliano orogeny), all magmas produced from mantle melting also show these peculiar geochemical signatures (e.g. Sial et al., 1989; Hollanda et al., 2003). It is worth noting that this enriched mantle is also proposed as the source for Mesoproterozoic and Neoproterozoic plutonic as well as Cretaceous and Cenozoic volcanic rocks in NE Brazil (Sial, 1976; Bellieni et al., 1992; Fodor et al., 1998; Neves et al., 2000; Mariano et al., 2001; Hollanda et al., 2006). It is, thus, suggested that the mantle enrichment process in NE Brazil is an ancient feature, probably dating back to at least late Archean times or shortly before the onset of Paleoproterozoic crust-forming events. A viable way to metasomatize the mantle is by hybridization of the depleted mantle through mixing with a slab-derived high-silica (TTG-like) adakite melt. Successive episodes of oceanic subduction during the Eburnean and Brasiliano orogenies enhanced this enrichment so that all magmas generated in this region show the LILE and HFSE characteristics described here.
|
CONCLUSION |
|---|
Our main results can be summarized as follows.
- Field, petrographic, geochemical and isotopic data show that the magmatic rocks of the Caicó Complex were generated by the same petrogenetic mechanisms.
- They are metaluminous, high-K calc-alkaline LILE- and LREE-enriched magmas emplaced at about 2·2 Ga.
- They have geochemical and isotopic characteristics of juvenile magmatism emplaced in a subduction-like tectonic setting, the most probable source being an enriched spinel- or garnet (<5% garnet)-bearing lherzolite.
- This tectonic setting favoured the hybridization of the depleted mantle source by slab-derived high-silica adakite melt, resulting in a metasomatized peridotite that generated by partial melting the parental magmas of the Caicó gneisses.
- The petrogenetic model involves two stages: first, partial melting (10–20%) of an enriched lherzolite gave rise to a basic magma that subsequently evolved by high-pressure fractionation of olivine, thus resulting in the parental magmas of the Caicó Complex orthogneisses; second, each parental magma evolved by fractional crystallization at crustal pressures (5–8 kbar) of a combination of amphibole + plagioclase + magnetite ± pyroxenes, thus giving rise to the plutonic suite.
- This juvenile magmatism extended throughout northeastern Brazil and has age and lithostratigraphic equivalents in French Guyana and in the West Africa and São Francisco cratons. Consequently, the Paleoproterozoic (2·2 Ga) juvenile magmatism represents a major continental accretion event far from the influence of older continental basement, and thus limiting contamination from it.
The data allow us to assign four specific features for the juvenile magmatism at the Archean–Proterozoic transition: (1) most of the geochemical and petrographic parameters are akin to those of modern granitoids; (2) granitoid magmas are mantle-derived, and recycling of continental crust is limited or absent; (3) the mantle can be either depleted (as in the West Africa, São Luís and northeastern São Francisco cratons, and French Guyana) or metasomatically enriched (as in the case studied here); (4) the metasomatic agent is believed to be a high-silica adakite (TTG-like) melt that hybridized with the depleted mantle. Finally, it should be stressed that the prevalence of wedge-dominated lithospheric mantle as the source for the granitoids of the Caicó Complex is comparable with processes responsible for the generation of modern juvenile granitoids, although the volume of magma generated resembles slab-dominated Archean continental crust-forming events.
|
SUPPLEMENTARY DATA |
|---|
Supplementary data for this paper are available at Journal of Petrology online.
|
ACKNOWLEDGEMENTS |
|---|
Z.S.S. thanks CAPES (Brazil) for providing scholarships for research activities at the universities of Rennes I (grant 3878/90-11) and Blaise Pascal (grant 3070/95-11). The authors thank J. Cornichet, M. Le Coz-Bouhnic (XFR) and S. Blais (neutron activation analysis) of the Institute of Geoscience of the Université de Rennes I, F. Vidal (Sr and Nd isotopes) of the Université Blaise Pascal (Clermont-Ferrand) and J. C. Gaspar (microprobe) of the Universidade de Brasília for analytical support, and V. P. Fonseca for great help during fieldwork. This research was financed by FINEP/PADCT and co-operation programmes between the Brazilian (CAPES) and French (COFECUB) governments (grants 97/89 and 177/95). We thank reviewers Robert Rapp, Hugh Rollinson and David Champion, and Editor Marjorie Wilson for their fruitful comments, which greatly improved the manuscript. Special thanks go to J.-W. Li, J. Fossa and E. Souza.
*Corresponding author. Telephone: 55-84-32153831. Fax: 55-84- 32153831. E-mail: zorano{at}geologia.ufrn.br
|
REFERENCES |
|---|
Abbott DH. Plumes and hot spots as sources of greenstone belts. Lithos (1996) 37:113–127.[CrossRef][Web of Science]
Abouchami WM, Boher M, Michard A, Albarède F. A major 2·1 Ga old event of mafic magmatism in West Africa: an early stage of crustal accretion. Journal of Geophysical Research (1990) 95:17605–17629.[CrossRef]
Albarède F. The growth of continental crust. Tectonophysics (1998) 296:1–14.[CrossRef][Web of Science]
Almeida FFM, Brito Neves BB, Fuck R. Brazilian structural provinces: an introduction. Earth-Science Reviews (1981) 17:1–29.
Archanjo CJ, Bouchez JL. Le Seridó, une chaîne transpressive dextre au Protérozoïque supérieur du Nord-Est du Brésil. Bulletin de la Société Géologique de France (1991) 162:637–647.[Abstract]
Arth JG, Barker F, Peterman ZE, Friedman I. Geochemistry of the gabbro–diorite–tonalite–trondhjemite suite of southwest Finland and its implications for the origin of tonalitic and trondhjemitic magmas. Journal of Petrology (1978) 19:289–316.
Atherton MP, Petford N. Generation of sodium-rich magmas from newly underplated basaltic crust. Nature (1993) 362:144–146.[CrossRef]
Baker MB, Stolper EM. Determining the composition of high-pressure mantle melts using diamond aggregates. Geochimica et Cosmochimica Acta (1994) 58:2811–2827.[CrossRef][Web of Science]
Baker MB, Hirschmann MM, Ghiorso MS, Stolper EM. Compositions of near-solidus peridotite melts from experimental and thermodynamic calculations. Nature (1995) 375:308–311.[CrossRef]
Barker F. Trondhjemites: definition, environment and hypotheses of origin. In: Trondhjemites, Dacites and Related Rocks—Barker F, ed. (1979) Amsterdam: Elsevier. 1–12.
Barker F, Arth JG. Generation of trondhjemite–tonalite liquids and Archean bimodal trondhjemite–basalt suites. Geology (1976) 4:596–600.
Beard JS, Lofgren GE. Effect of water on the composition of partial melts of greenstones and amphibolites. Science (1989) 144:195–197.
Beard JS, Lofgren GE. Dehydration melting and water-saturated melting of basaltic and andesitic greenstones and amphibolites at 1, 3, and 6·9 kb. Journal of Petrology (1991) 32:365–401.
Bellieni G, Macedo MHF, Petrini R, Piccirillo EM, Cavazzini G, Comin-Chiaramont P, Ernesto M, Macedo JWP, Martins G, Melfi AJ, Pacca IG, de Min A. Evidence of magmatic activity related to Middle Jurassic and Lower Cretaceous rifting from northeastern Brazil (Ceará–Mirim): K/Ar age, palaeomagnetism, petrology and Sr–Nd isotope characteristics. Chemical Geology (1992) 97:9–32.[CrossRef][Web of Science]
Bertrand JM, Jardim de Sá EF. Where are the Eburnean–Transmazonian collisional belts? Canadian Journal of Earth Sciences (1990) 27:1382–1393.
Béziat D, Bourges F, Debat P, Lompo M, Martin F, Tollon F. A Paleoproterozoic ultramafic–mafic assemblage and associated volcanic rocks of the Boromo greenstone belt: fractionates originating from island-arc volcanic activitiy in the West African craton. Precambrian Research (2000) 101:25–47.[CrossRef][Web of Science]
Bindeman IN, Eiler JM, Yogodzinski GM, Tatsumi Y, Stern CR, Grove TL, Portnyagin M, Hoernle K, Danyushevsky LV. Oxygen isotope evidence for slab melting in modern and ancient subduction zones. Earth and Planetary Science Letters (2005) 235:480–496.[CrossRef][Web of Science]
Blundy JD, Holland TJB. Calcic amphibole equilibria and a new amphibole–plagioclase geothermometer. Contributions to Mineralogy and Petrology (1990) 104:208–224.[CrossRef][Web of Science]
Boher M, Abouchami W, Michard A, Albarède F, Arndt N. Crustal growth in west Africa at 2·1 Ga. Journal of Geophysical Research (1992) 97:345–369.
Bourdon E, Eissen JP, Monzier M, Robin C, Martin H, Cotten J, Hall ML. Adakite-like lavas from Antisana Volcano (Ecuador): Evidence for slab melt metasomatism beneath Andean Northern Volcanic Zone. Journal of Petrology (2002) 43:199–217.
Bureau H, Keppler H. Complete miscibility between silicate melts and hydrous fluids in the upper mantle: experimental evidence and geochemical implications. Earth and Planetary Science Letters (1999) 165:187–196.[CrossRef][Web of Science]
Caby R. Precambrian terranes of Benin–Nigeria and northeast Brazil and the late Proterozoic South Atlantic fit. Terranes in the circum-Atlantic Paleozoic orogens. Geological Society of America, Special Papers—Dallmeyer RD, ed. (1989) 230:145–158.
Caby R, Bertrand JM, Black R. Pan-African ocean closure and continental collision in the Hoggar–Iforas Segment, Central Sahara. In: Precambrian Plate Tectonics—Kröner A, ed. (1981) Amsterdam: Elsevier. 407–431.
Caby R, Sial AN, Arthaud MH, Vauchez A. Crustal evolution and the Brasiliano orogeny in northeast Brazil. In: The West African Orogens and Circum-Pacific–Atlantic Correlatives—Dallmeyer RD, Lécorché JP, eds. (1991) Berlin: Springer. 373–397.
Castro DL, Barbosa CCF, Silva JB, Medeiros WE. Relevo da interface crosta–manto no Nordeste setentrional do Brasil: comparação entre vínculos de isostasia e suavidade. (1997a) In: Proceedings of the 5th International Congress of the SBGf, São Paulo Abstracts, Volume 2. São Paulo: Sociedade Brasileira de Geofísica. 682–685.
Castro DL, Medeiros WE, Moreira JA, Jardim de Sá EF. Mapa gravimétrico do Nordeste setentrional do Brasil e margem continental adjacente. (1997b) In: Proceedings of the 5th International Congress of the SBGf, São Paulo, Abstracts, Volume 2. São Paulo: Sociedade Brasileira de Geofísica. 678–681.
Chappell BW, White AJR. Two contrasting granite types. Pacific Geology (1974) 8:173–184.
Cocherie A. Systematic use of trace element distribution patterns in log–log diagrams for plutonic suites. Geochimica et Cosmochimica Acta (1986) 50:2517–2522.[CrossRef][Web of Science]
Conceição H. Magmatismo sienítico alcalino potássico no leste da Bahia: evolução química, fonte e implicações geodinâmicas para a estrutura do manto paleoproterozóico no nordeste do Brasil. (1997) In: Proceedings of the 6th Brazilian Geochemistry Congress, Salvador, Volume 2. Salvador: Sociedade Brasileira de Geoquímica. 808–810.
Condie KC. Archean Greenstone Belts (1981) Amsterdam: Elsevier. 434.
Condie KC. Plate Tectonics and Crustal Evolution (1989) Oxford: Pergamon Press. 476.
Condie KC. Episodic continental growth and supercontinents: a mantle avalanche connection? Earth and Planetary Science Letters (1998) 163:97–108.[CrossRef][Web of Science]
Condie KC. Episodic continental growth models: afterthoughts and extensions. Tectonophysics (2000) 322:153–162.[CrossRef][Web of Science]
Condie KC. Mantle Plumes and their Record in Earth History (2001) Cambridge: Cambridge University Press. 306.
Corsini M, Vauchez A, Archanjo C, Jardim de Sá EF. Strain transfer at continental scale from a transcurrent shear zone to a transpressional fold belt: the Patos–Seridó system, northeastern Brazil. Geology (1991) 19:586–589.
Dantas EL. Evolução tectono-magmática do maciço polidiapírico São Vicente/Florânia–RN. (1992) Rio Claro: Universidade Estadual Paulista (Unesp). 272. M.Sc. dissertation.
Dantas EL. Geocronologia U–Pb e Sm–Nd de terrenos arqueanos e paleoproterozóicos do Maciço Caldas Brandão, NE do Brasil. (1996) Rio Claro: Universidade Estadual Paulista (Unesp). 206. Ph.D. thesis.
Dantas EL, Van Schmus WR, Hackspacher PC, Fetter AH, Brito Neves BB, Cordani UG, Nutman AP, Williams IS. The 3·4–3·5 Ga São José do Campestre massif, NE Brazil: remnants of the oldest crust in South America. Precambrian Research (2004) 130:113–137.[CrossRef][Web of Science]
Davidson JP, Arculus RJ. The significance of Phanerozoic arc magmatism in generating continental crust. In: Evolution and Differentiation of the Ccontinental Crust—Brown M, Rushmer T, eds. (2006) Cambridge: Cambridge University Press. 134–172.
Deer WA, Howie RA, Zussman J. An Introduction to the Rock-forming Minerals (1983) Hong Kong: Longman. 528.
Defant MJ, Drummond MS. Derivation of some modern arc by melting of young subducted lithosphere. Nature (1990) 347:662–665.[CrossRef]
Defant MJ, Richerson PM, de Boer JZ, Stewart RH, Maury RC, Bellon H, Drummond MS, Feigenson MD, Jackson TE. Dacite genesis via both slab melting and differentiation: petrogenesis of La Yeguada Volcanic Complex, Panama. Journal of Petrology (1991) 32:1101–1142.
Delor C, Lahondère D, Egal E, Lafon JM, Cocheril A, Guerrot C, Rossi P, Truffert C, Théveniaut H, Phillips D, Avelar VG. Transamazonian crustal growth and reworking as revealed by the 1:500,000-scale geological map of French Guiana (2nd edition). Géologie de la France (2003) 2–3–4:5–57.
DePaolo DJ. Trace element and isotopic effects of combined wall rock assimilation and fractional crystallization. Earth and Planetary Science Letters (1981) 53:189–202.[CrossRef][Web of Science]
DePaolo DJ. Neodymium Isotope Geochemistry—an Introduction (1988) Berlin: Springer. 187.
DePaolo DJ, Wasserburg GJ. Petrogenetic mixing models and Sr–Nd isotopic patterns. Geochimica et Cosmochimica Acta (1979) 43:615–627.[CrossRef][Web of Science]
de Souza ZS. Petrogenèse des métagranitoïdes du Complexe de Caicó, Province Borborema (État du Rio Grande do Norte; Brésil). (1991) Rennes: CAESS, Université de Rennes I. 90. DESS Dissertation.
de Souza ZS, Martin H, Macedo MH, Peucat JJ, Jardim de Sá EF. Un segment de croûte continentale juvénile d’âge protérozoïque inférieur: le Complexe de Caicó (Rio Grande do Norte, NE-Brésil). Comptes Rendus de l’Académie de Sciences (1993) 316:201–208.
Didier J, Duthou JL, Lameyre J. Mantle and crustal granites: genetic classification of orogenic granites and the nature of their enclaves. Journal of Volcanology and Geothermal Research (1982) 14:125–132.[CrossRef][Web of Science]
Dorbath C, Dorbath L, Fairhead JD, Stuart GW. A teleseismic delay time study across the Central African Shear Zone in the Adamawa region of Cameroon, East Africa. Geophysical Journal of the Royal Astronomical Society (1986) 86:752–766.
Doumbia S, Pouclet A, Kouamelan A, Peucat JJ, Vidal M, Delor C. Petrogenesis of juvenile-type Birimain (Paleoproterozoic) granitoids in Central Côte d’Ivoire, West Africa: geochemistry and geochronology. Precambrian Research (1988) 87:33–63.[CrossRef]
Drummond MS, Defant MJ. A model for trondhjemite–tonalite–dacite genesis and crustal growth via slab melting: Archean to modern comparisons. Journal of Geophysical Research (1990) 95:21503–21521.[CrossRef]
Drummond MS, Ragland PC, Wesolowki D. An example of trondhjemite genesis by means of alkali metasomatism: Rockford Granite, Alabama Appalachians. Contributions to Mineralogy and Petrology (1986) 93:98–113.[CrossRef][Web of Science]
Faure G. edn. In: Principles of Isotope Geology (1986) 2nd. New York: John Wiley. 589.
Fetter AH, Van Schmus WR, Santos TS, Nogueira Neto JA, Arthaud MH. U–Pb and Sm–Nd geochronological constraints on the crustal evolution and basement architecture of Ceará state, NW Borborema Province, NE Brazil: implications for the existence of the Paleoproterozoic supercontinent ‘Atlantica’. Revista Brasileira de Geociências (2000) 30:102–106.
Feybesse JL, Billa M, Guerrot C, Duguey E, Lescuyer JL, Milesi JP, Bouchot V. The Paleoproterozoic Ghanian province: geodynamic model and ore control, including regional stress modelling. Precambrian Research (2006) 149:149–196.[CrossRef][Web of Science]
Fodor RV, Mukasa SB, Sial AN. Isotopic and trace-element indications of lithospheric and asthenospheric components in Tertiary alkalic basalts, northeastern Brazil. Lithos (1998) 43:197–217.[CrossRef][Web of Science]
Foley SF, Tiepolo M, Vannucci R. Growth of early continental crust controlled by melting of amphibolite in subduction zones. Nature (2002) 417:637–640.
Gasquet D, Barbey, P. Adou M, Paquette JL. Structure, Sr–Nd isotope geochemistry and zircon U–Pb geochronology of the granitoids of the Dabakala area (Côte d’Ivoire): evidence for a 2·3 Ga crustal growth event in the Palaeoproterozoic of West Africa? Precambrian Research (2002) 127:329–354.[CrossRef]
Govindaraju K, Mévelle G, Chouard C. Automated optical emission spectrochemical bulk analyses of silicate rocks with microwave plasma excitation. Analytical Chemica USA (1976) 48:1325–1331.
Graviou P, Auvray B. Caractérisation pétrographique et géochimique des granitoïdes cadomiens du domaine nord-armoricain. Comptes Rendus de l’Académie des Sciences (1985) 301:315–318.
Graviou P, Auvray B. Late Precambrian M-type granitoid genesis in the Cadomian belt of NW France. In: D’Lemos, R. S. Strachan, R. A. & Topley, C. G. (eds). The Cadomian Orogeny. Geological Society, London, Special Publications (1990) 51:231–244.[CrossRef]
Graviou P, Peucat JJ, Auvray B, Vidal P. The Cadomian Orogeny in the northern Armorican Massif: petrological and geochronological constraints on a geodynamic model. Hercynica (1988) 1:1–13.
Green DH. Experimental melting studies on a model upper mantle composition at high pressure under water-saturated and water-undersaturated conditions. Earth and Planetary Science Letters (1973) 19:37–53.[CrossRef][Web of Science]
Green DH, Falloon TJ. Pyrolite: A Ringwood concept and its current expression. In: The Earth's Mantle—Composition, Structure, and Evolution—Jackson I, ed. (1998) Cambridge: Cambridge University Press. 311–378.
Green TH, Pearson NJ. Ti-rich accessory phase saturation in hydrous mafic–felsic compositions at high P, T. Chemical Geology (1986) 54:185–201.[CrossRef][Web of Science]
Gruau G, Martin H, Leveque B, Capdevila R. Rb–Sr and Sm–Nd geochronology of Lower Proterozoic granite–greenstone terrains in French Guiana, South America. Precambrian Research (1985) 30:63–80.[CrossRef][Web of Science]
Hackspacher PC, Dantas EL. Northwestern overthrusting and related lateral escape during the Brasiliano orogeny north of the Patos Lineament, Borborema Province, Northeast Brazil. International Geology Review (1997) 39:609–620.
Hackspacher PC, Dantas EL, Van Schmus WR. Datação do metamorfismo associado à tectônica colisional Transamazônica—U/Pb em titanita e zircão. (1995) In: Proceedings of the 5th National Symposium on Tectonic Studies, Gramado. Porto Alegre: Só Cópias Encadernações Ltda. 379–381.
Hackspacher PC, Van Schmus WR, Dantas EL. Um embasamento transamazônico na Província Borborema. (1990) In: Proceedings of the 36th Brazilian Geological Congress, Natal, Volume 6. Natal: Sociedade Brasileira de Geologia. 2683–2696.
Halla J. Late Archean high-Mg granitoids (sanukitoids) in the southern Karelian domain, eastern Finland: Pb and Nd isotopic constraints on crust–mantle interactions. Lithos (2005) 79:161–178.[CrossRef][Web of Science]
Hammarstrom JM, Zen E. Aluminium in hornblende: an empirical igneous geobarometer. American Mineralogist (1986) 89:317–329.
Hawkesworth CJ, Gallagher K, Hergt J, McDermott F. Mantle and slab contribution in arc magmas. Annual Review of Earth and Planetary Sciences (1993) 21:175–204.[Web of Science]
Helz RT. Phase relations in basalt in their melting range at P(H2O) = 5 kb. Part II. Melt compositions. Journal of Petrology (1976) 17:139–193.
Hildreth W, Moorbath S. Crustal contributions to arc magmatism in the Andes of Central Chile. Contributions to Mineralogy and Petrology (1988) 98:455–489.[CrossRef][Web of Science]
Hirose K. Melting experiments on lherzolite KLB-1 under hydrous conditions and generation of high-magnesian andesitic melts. Geology (1997) 25:42–44.
Hirschmann MM, Baker MB, Stolper EM. The effect of alkalis on the silica content of mantle-derived melts. Geochimica et Cosmochimica Acta (1998) 62:883–902.[CrossRef][Web of Science]
Hoffman AW. Chemical differentiation of the Earth: the relationship between mantle, continental crust, and oceanic crust. Earth and Planetary Science Letters (1988) 90:297–314.[CrossRef][Web of Science]
Hollanda MHBM, Pimentel MM, Jardim de Sá EF. Paleoproterozoic subduction-related metasomatic signatures in the lithospheric mantle beneath NE Brazil: inferences from trace element and Sr–Nd-Pb isotopic compositions of Neoproterozoic high-K igneous rocks. Journal of South American Earth Sciences (2003) 15:885–900.[CrossRef][Web of Science]
Hollanda MHBM, Pimentel MM, Oliveira DC, Jardim de Sá EF. Lithosphere–asthenosphere interaction and the origin of Cretaceous tholeiitic magmatism in Northeastern Brazil: Sr–Nd–Pb isotopic evidence. Lithos (2006) 86:34–49.[CrossRef][Web of Science]
Hollanda MHBM, Archanjo CJ, Souza LC, Armstrong R. U/Pb geochronology of the augen gneisses from the Seridó belt (Borborema Province): Geodynamic significance. (2007) In: Proceedings of the 5th International Symposium on Tectonics of the SBG, Natal, Abstracts. Natal: Sociedade Brasileira de Geologia. 347–347.
Huppert HE, Sparks RSJ. Cooling and contamination of mafic and ultramafic magmas during ascent through the continental crust. Earth and Planetary Science Letters (1985) 74:371–386.[Web of Science]
Irvine TN, Baragar WRA. A guide to the chemical classification of the common volcanic rocks. Canadian Journal of Earth Sciences (1971) 8:523–548.
Jacobsen SB, Wasserburg GJ. Sm–Nd isotopic evolution of chondrites. Earth and Planetary Science Letters (1980) 50:139–155.[CrossRef][Web of Science]
Jardim de Sá EF. A evolução Proterozóica da Província Borborema. (1984) In: Proceedings of the 11th Geological Symposium of NE Brazil, Natal. Natal: Sociedade Brasileira de Geologia. 297–316.
Jardim de Sá EF. A Faixa Seridó (Província Borborema, NE do Brasil) e o seu significado geodinâmico na cadeia Brasiliana/Pan-Africana. (1994) Brasília: Universidade de Brasília (UnB). 803. Ph.D. thesis.
Jardim de Sá EF, Macedo MHF, Legrand JM, McReath I, Galindo AC, Sá JM. Proterozoic granitoids in a polycyclic setting: the Seridó region, NE Brazil. (1987) In: Proceedings of the International Symposium on Granites and Associated Mineralizations, Salvador, Extended Abstracts. Salvador: Superintendência de Geologia e Recursos Minerais. 103–110.
Jardim de Sá EF, Fuck RA, Macedo MHF, Peucat JJ, Kawashita K, de Souza ZS, Bertrand JM. Pre-Brasiliano orogenic evolution in the Seridó Belt, NE Brazil: conflicting geochronological and structural data. Revista Brasileira de Geociências (1995) 25:307–314.
Jayananda M, Martin H, Peucat JJ, Mahabaleswar B. Late Archean crust–mantle interactions: geochemistry of LREE-enriched mantle derived magmas. Example of the Closepet batholith, southern India. Contributions to Mineralogy and Petrology (1995) 119:314–329.[Web of Science]
Kelemen PB, Johnson KTM, Kinzler RJ, Irving AJ. High-field-strength element depletion in arc basalts due to mantle–magma interaction. Nature (1990) 345:521–524.[CrossRef]
Kemp AIS, Hawkesworth CJ. Granitic perspectives on the generation and secular evolution of the continental crust. In: Treatise on Geochemistry, Volume 3—Holland HD, Turekian KK, eds. (2003) Amsterdam: Elsevier/Pergamon. 349–410.
Keppler H. Constraints from partitioning experiments on the composition of subduction zone fluids. Nature (1996) 380:237–240.[CrossRef]
Kessel R, Schmidt MW, Ulmer P, Pettke T. Trace element signature of subduction-zone fluids, melts and supercritical liquids at 120–180 km depth. Nature (2005) 437:724–727.[CrossRef][Medline]
Klein EL, Moura CAV, Krymsky RS, Griffin WL. The Gurupi Belt, northern Brazil: lithostratigraphy, geochronology, and geodynamic evolution. Precambrian Research (2005a) 141:83–105.[CrossRef][Web of Science]
Klein EL, Moura CAV, Pinheiro BLS. Paleoproterozoic crustal evolution of the São Luís Craton, Brazil: evidence from zircon geochronology and Sm–Nd isotopes. Gondwana Research (2005b) 8:177–186.[CrossRef][Web of Science]
Köber B. Whole grain evaporation for 207Pb/206Pb-age investigations on single zircon using a double-filament thermal ion source. Contributions to Mineralogy and Petrology (1986) 93:482–490.[CrossRef][Web of Science]
Kogiso T, Tatsumi Y, Nakano S. Trace element transport during dehydration processes in the subducted oceanic crust: 1. Experiments and implications for the origin of ocean island basalts. Earth and Planetary Science Letters (1997) 148:193–205.[CrossRef][Web of Science]
Kuno H. Lateral variation of basaltic magma types across continental margins and island arcs. Bulletin of Volcanology (1966) 29:195–222.
Kuno H. Differentiation of basalt magmas. In: The Poldervaart Treatise on Rocks of Basaltic Composition, Volume 2—Hess HH, Poldervaart A, eds. (1968) New York: Interscience. 623–688.
Kushiro I. Melting of hydrous upper mantle and possible generation of andesitic magmas. Earth and Planetary Sciences Letters (1974) 22:294–299.[CrossRef]
Kushiro I, Shimizu N, Nakamura Y. Compositions of coexisting liquid and solid phases formed upon melting of natural garnet and spinel lherzolites at high pressures: a preliminary report. Earth and Planetary Science Letters (1972) 14:19–25.[CrossRef][Web of Science]
Lameyre J, Bowden P. Plutonic rock type series: discrimination of various granitoid series and related rocks. Journal of Volcanology and Geothermal Research (1982) 14:169–186.[CrossRef][Web of Science]
Langmuir CH, Vocke RD, Jr, Hanson GN, Hart SR. A general mixing equation with applications to Icelandic basalts. Earth and Planetary Science Letters (1978) 37:380–392.[CrossRef][Web of Science]
Leake BE. Nomenclature of amphiboles. American Mineralogist (1978) 63:1023–1052.[Abstract]
Ledru P, Johan V, Milési JP, Tegvey M. Markers of the final stages of the Paleoproterozoic collision: evidence for a 2 Ga continent involving circum-South Atlantic provinces. Precambrian Research (2003) 69:139–165.
Legrand JM, Liegois JP, Deutsch S. Datação U/Pb e Rb/Sr de rochas precambrianas da região de Caicó. Reavaliação da definição de um embasamento Arqueano. (1991) In: Proceedings of the 14th Geological Symposium of NE Brazil, Recife. Recife: Sociedade Brasileira de Geologia. 276–279.
Leterrier J, Jardim de Sá EF, Macedo MHF. Magmatic and geodynamic signature of the Brasiliano cycle plutonism in the Seridó belt, NE Brazil. (1990) In: Proceedings of the 36th Brazilian Geological Congress, Natal, Volume 4. Natal: Sociedade Brasileira de Geologia. 1640–1655.
Leterrier J, Jardim de Sá EF, Bertrand JM, Pin C. Ages U–Pb sur zircon de granitoïdes ‘brasilianos’ de la ceinture du Seridó (Province Borborema, NE Brésil). Comptes Rendus de l’Académie de Sciences (1994) 318:1505–1511.
Lobach-Zhuchenko SB, Rollinson HR, Chekulaev VP, Arestova NA, Kovalenko AV, Ivanikov VV, Guseva NS, Sergeev SA, Matukov DI, Jarvis KE. The Archean sanukitoid series of the Baltic Shield: geological setting, geochemical characteristics and implications for their origin. Lithos (2005) 79:107–128.[CrossRef][Web of Science]
Ludwig KR. Isoplot 3.00. (2003) 70. Berkeley Geochronology Center Special Publication No. 4.
MacDonald GA, Katsura T. Chemical composition of Hawaiian lavas. Journal of Petrology (1964) 5:82–133.
Macedo MHF, Jardim de Sá EF, Sá JM. Datações Rb–Sr em ortognaisses e a idade do Grupo Seridó. (1984) In: Proceedings of the 11th Geological Symposium of NE Brazil, Natal. Natal: Sociedade Brasileira de Geologia. 253–262.
Maniar PD, Piccoli PM. Tectonic discrimination of granitoids. Geological Society of America Bulletin (1989) 101:635–643.
Manning CE. The chemistry of subduction-zone fluids. Earth and Planetary Science Letters (2004) 223:1–16.[CrossRef][Web of Science]
Mariano G, Neves SP, Silva Filho AF, Guimarães IP. Diorites of the high-K calc-alkalic association: geochemistry and Sm–Nd data and implications for the evolution of the Borborema Province, northeast Brazil. International Geology Review (2001) 43:921–929.[Web of Science]
Martin H. Nature, origine et évolution d’un segment de croûte continentale archéenne: contraintes chimiques et isotopiques. In: Exemple de la Finlande Orientale. (1985) Rennes: CAESS, Université de Rennes I. 392. Ph.D. thesis.
Martin H. Effect of steeper Archean geothermal gradient on geochemistry of subduction-zone magma. Geology (1986) 14:753–756.
Martin H. Petrogenesis of Archean trondhjemites, tonalites and granodiorites from eastern Finland: major and trace element geochemistry. Journal of Petrology (1987) 28:921–953.
Martin H. The mechanisms of petrogenesis of the Archean continental crust—Comparison with modern processes. Lithos (1993) 30:373–388.[CrossRef][Web of Science]
Martin H. The Archean grey gneiss and the genesis of continental crust. In: The Archean Crustal Evolution—Condie KC, ed. (1994) Amsterdam: Elsevier. 205–259.
Martin H. The adakitic magmas: modern analogs of Archean granitoids. Lithos (1999) 46:411–429.[CrossRef][Web of Science]
Martin H, Moyen JF. Secular changes in TTG composition as markers of the progressive cooling of the Earth. Geology (2002) 30:319–322.
Martin H, Peucat JJ, Sabaté P, Cunha JC. Crustal evolution in the early Archean of South America: example of the Sete Voltas Massif, Bahia State, Brazil. Precambrian Research (1997) 82:35–62.[CrossRef][Web of Science]
Martin H, de Souza ZS, Fonseca VP, Jardim de Sá EF. Geochemistry of high-grade metaplutonics: the Caicó Complex, NE Brazil. (1990) In: Proceedings of the 36th Brazilian Geological Congress, Abstracts. Natal: Sociedade Brasileira de Geologia. 171–171.
Martin H, Smithies RH, Rapp R, Moyen J.-F, Champion D. An overview of adakite, tonalite–trondhjemite–granodiorite (TTG), and sanukitoid: relationships and some implications for crustal evolution. Lithos (2005) 79:1–24.[CrossRef][Web of Science]
Martins G, Oliveira EP. The Algodões amphibolites, central Ceará domain: geochronologic and trace element evidence for accretion of Palaeoproterozoic oceanic plateau basalts in the Borborema Province, Brazil. (2003) In: Companhia Baiana de Pesquisa Mineral (CBPM) and Institut de Recherche pour le Dévelopment (IRD) (eds) Proceedings of the 4th South American Symposium on Isotope Geology, Salvador. Salvador: CBPM and IRD. 211–214.
Maury R, Sajona FG, Pubellier M, Bella H, Defant M. Fusion de la croûte océanique dans les zones de subduction/collision récent: l’exemple de Mindanao (Philippines). Bulletin de la Société Géologique de France (1996) 167:579–595.[Abstract]
McCulloch MT. Sm–Nd isotopic constraints on the evolution of Precambrian crust in the Australian continent. Proterozoic Lithosphere Evolution. American Geophysical Union, Geodynamic Series—Kröner A, ed. (1987) 17:115–130.
McCulloch MT, Bennett VC. Progressive growth of the Earth's continental crust and depleted mantle: geochemical constraints. Geochimica et Cosmochimica Acta (1994) 58:4717–4738.[CrossRef][Web of Science]
McLennan SM, Taylor SR, Hemming SR. Composition, differentiation, and evolution of continental crust: constraints from sedimentary rocks and heat flow. In: Evolution and Differentiation of the Continental Crust—Brown M, Rushmer T, eds. (2006) Cambridge: Cambridge University Press. 92–134.
McReath I, Faraco MTL. Paleoproterozoic greenstone–granite belts in Northern Brazil and the former Guyana shield–West African craton province. Revista do Instituto de Geociências/USP (2006) 5:49–63.
Menzies M, Rogers N, Tindle A, Hawkesworth C. Metasomatic and enrichment processes in lithospheric peridotites, an effect of asthenosphere–lithosphere interaction. In: Mantle Metasomatism—Menzies MK, Hawkesworth CJ, eds. (1987) London: Academic Press. 313–361.
Morris PA. Slab melting as an explanation of Quaternary volcanism and aseismicity in southwestern Japan. Geology (1995) 23:395–398.
Moyen JF, Martin H, Jayananda M. Origine du granite fini-Archéen de Closepet (Inde du Sud): apports de la modélisation géochimique du comportement des éléments en trace. Comptes Rendus de l’Académie de Sciences (1997) 325:659–664.
Moyen JF, Martin H, Jayananda M. Multi-element geochemical modelling of crust–mantle interactions during late-Archean crustal growth: the Closepet granite (South India). Precambrian Research (2001) 112:87–105.[CrossRef][Web of Science]
Moyen JF, Martin H, Jayananda M, Auvray B. Late Archean granites: a typology based on the Dharwar Craton (India). Precambrian Research (2003) 127:103–123.[CrossRef][Web of Science]
Mysen SJB, Boettcher AL. Melting of hydrous mantle: II. Geochemistry of crystals and liquids formed by anatexis of mantle peridotite at high pressures and high temperatures as a function of controlled activities of water, hydrogen, and carbon dioxide. Journal of Petrology (1975) 16:549–593.
Neves SP. Proterozoic history of the Borborema Province (NE Brazil): correlations with neighboring cratons and Pan-African belts and implications for the evolution of western Gondwana. Tectonics (2003) 22:1031. doi:10.1029/2001TC001352.[CrossRef]
Neves SP, Mariano G, Guimarães IP, Silva Filho AF, Melo SC. Intralithospheric differentiation and crustal growth: evidence from the Borborema Province, northeastern Brazil. Geology (2000) 28:519–522.
Neves SP, Bruguier O, Vauchez A, Bosch D, Silva JMR, Mariano G. Timing of crust formation, deposition of supracrustal sequences, and Transamazonian and Brasiliano metamorphism in the East Pernambuco belt (Borborema Provinca, NE Brazil): implications for western Gondwana assembly. Precambrian Research (2006) 149:197–216.[CrossRef][Web of Science]
Nicholls IA, Ringwood AE. Production of silica-saturated tholeiitic magmas in island arcs. Earth and Planetary Science Letters (1972) 17:243–246.[CrossRef][Web of Science]
Nielsen R. Geochemical Earth Reference Model (GERM) partition coefficient (Kd) database. (2007) (accessed 9 April 2007). Available at http://www.geo.oregonstate.edu/people/faculty/nielsenr.htm.
Nisbet EG. The Young Earth: an Introduction to Archean Geology (1987) Boston, MA: Allen and Unwin. 402.
Nockolds SR, Allen R. The geochemistry of some igneous rock series. Geochimica et Cosmochimica Acta (1953) 4:105–142.[CrossRef][Web of Science]
O’Connor JT. A classification for quartz-rich igneous rocks based on feldspar ratios. US Geological Survey, Professional Papers (1965) 525B:79–84.
Patiño Douce AEP, Beard JS. Dehydration-melting of biotite gneiss and quartz amphibolite from 3 to 15 kbar. Journal of Petrology (1995) 36:707–738.
Peacock SM. Fluid processes in subduction zones. Science (1990) 248:329–337.
Peacock SM. Large-scale hydration of the lithosphere above subduction slabs. Chemical Geology (1993) 108:43–59.
Peacock SM, Rushmer T, Thompson AB. Partial melting of subducted oceanic crust. Earth and Planetary Science Letters (1994) 121:227–244.[CrossRef][Web of Science]
Pearce JA. Trace element characteristics of lavas from destructive plate boundaries. In: Andesites: Orogenic Andesites and Related Rocks—Thorpe RS, ed. (1982) Chichester: John Wiley. 525–548.
Pearce J. Sources and settings of granitic rocks. Episodes (1996) 19:120–125.[Web of Science]
Pearce JA, Parkinson IJ. Trace element models for mantle melting: application to volcanic arc petrogenesis. Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications—Prichard HM, Alabaster T, Harris NBW, Neary CR, eds. (1993) 76:373–403.
Pearce JA, Harris NB, Tindle AG. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology (1984) 25:956–983.
Petta RA. Estudo geoquímico e relações petrogenéticas do batólito múltiplo composto São Vicente/Caicó, RN–Brasil. (1995) Rio Claro: Universidade Estadual Paulista (Unesp). 430. Ph.D. thesis.
Pin C, Paquette JL. A mantle-derived bimodal suite in the Hercynian belt: Nd isotope and trace element evidence for a subduction-related rift origin of the Late Devonian Brévenne metavolcanics, Massif Central (France). Contributions to Mineralogy and Petrology (1997) 129:222–238.[CrossRef][Web of Science]
Prouteau G, Maury R, Rangin C, Suparka E, Bellon H, Pubellier M, Cotten J. Les adakites miocènes du NW de Bornéo, témoins de la fermeture de la proto-mer de Chine. Comptes Rendus de l’Académie des Sciences (1996) 323:925–932.
Rapp RP, Watson EB. Dehydration melting of metabasalt at 8–32 kbar: implications for continental growth and crust–mantle recycling. Journal of Petrology (1995) 36:891–931.
Rapp RP, Watson EB, Miller CF. Partial melting of amphibolite/eclogite and the origin of Archean trondhjemites and tonalites. Precambrian Research (1991) 51:1–25.[CrossRef][Web of Science]
Rapp RP, Shimizu N, Norman MD, Applegate GS. Reaction between slab-derived melts and peridotite in the mantle wedge: experimental constraints at 3·8 GPa. Chemical Geology (1999) 160:335–356.[CrossRef][Web of Science]
Rapp RP, Shimizu N, Norman MD. Growth of early continental crust by partial melting of eclogite. Nature (2003) 425:605–609.[CrossRef][Medline]
Rayleigh JWS. Theoretical considerations respecting the separation of gases by diffusion and similar processes. Philosophical Magazine (1896) 42:77–107.
Rickwood PC. Boundary lines within petrologic diagrams which use oxides of major and minor elements. Lithos (1989) 22:247–263.[CrossRef][Web of Science]
Robinson JAC, Wood BJ, Blundy JD. The beginning of melting of fertile and depleted peridotite at 1·5 GPa. Earth and Planetary Science Letters (1998) 155:97–111.[CrossRef][Web of Science]
Rollinson HR. Using Geochemical Data: Evaluation, Presentation, Interpretation (1993) New York: John Wiley. 352.
Rollinson H. Crustal generation in the Archean. In: Evolution and Differentiation of the Continental Crust—Brown M, Rushmer T, eds. (2006) Cambridge: Cambridge University Press. 173–230.
Rudnick RL. Making continental crust. Nature (1995) 378:571–578.[CrossRef]
Rudnick RL, Fountain DM. Nature and composition of the continental crust: a lower crustal perspective. Reviews in Geophysics (1995) 33:267–309.[CrossRef]
Rushmer T. Partial melting of two amphibolites: contrasting experimental results under fluid-absent conditions. Contributions to Mineralogy and Petrology (1991) 107:41–59.[CrossRef][Web of Science]
Ryerson FJ, Watson EB. Rutile saturation in magmas: implications for Ti–Nb–Ta depletion in island-arc basalts. Earth and Planetary Sciences Letters (1987) 86:225–239.[CrossRef]
Samaniego P, Martin H, Robin C, Monzier M. Transition from calc-alkalic to adakitic magmatism at Cayambe volcano, Ecuador: Insights into slab melts and mantle wedge interactions. Geology (2002) 30:967–970.
Samsonov AV, Bogina MM, Bibikova EV, Petrova AY, Shchipansky AA. The relationship between adakitic, calc-alkaline volcanic rocks and TTGs: implications for the tectonic setting of the Karelian greenstone belts, Baltic Shield. Lithos (2005) 79:83–106.[CrossRef][Web of Science]
Santos EJ. Ensaio preliminar sobre terrenos e tectônica acrescionária na Província Borborema. (1996) In: Proceedings of the 39th Brazilian Geological Congress, Salvador, Volume 6. Salvador: Sociedade Brasileira de Geologia. 47–50.
Schiano P, Clocchiatti R, Shimizu N, Maury RC, Jochum KP, Hofmann AW. Hydrous, silica-rich melts in the sub-arc mantle and their relationships with erupted arc lavas. Nature (1995) 377:595–600.[CrossRef]
Schmidt MW. Amphibole composition in tonalite as a function of pressure: an experimental calibration of the Al-in hornblende barometer. Contributions to Mineralogy and Petrology (1992) 110:304–310.[CrossRef][Web of Science]
Schmidt MW, Poli S. Experimentally based water budgets for dehydrating slabs and consequences for arc magma generation. Earth and Planetary Science Letters (1998) 163:361–379.[CrossRef][Web of Science]
Schmidt MW, Vielzeuf D, Auzanneau E. Melting and dissolution of subducting crust at high pressures: the key role of white mica. Earth and Planetary Science Letters (2004) 228:65–84.[CrossRef][Web of Science]
Sen C, Dunn T. Dehydration melting of a basaltic composition amphibolite at 1·5 and 2·0 GPa: implications for the origin of adakites. Contributions to Mineralogy and Petrology (1994a) 117:394–409.[CrossRef][Web of Science]
Sen C, Dunn T. Experimental modal metasomatism of a spinel lherzolite and the production of amphibole-bearing peridotite. Contributions to Mineralogy and Petrology (1994b) 119:422–432.[CrossRef][Web of Science]
Shaw DM. Trace element fractionation during anatexis. Geochimica et Cosmochimica Acta (1970) 34:237–243.[CrossRef][Web of Science]
Shirey SB, Hanson GN. Mantle derived Archean monzodiorites and trachyandesites. Nature (1984) 310:222–224.[CrossRef]
Sial AN. The post-Paleozoic volcanism of north-east Brazil and its tectonic significance. Anais da Academia Brasileira de Ciências (1976) 48:299–311.[Web of Science]
Sial AN, Ferreira VP, Whitney JA, Wenner DB, Sadaki A, Long LE. LILE 18O- and 34S-enriched mantle beneath northern Brazil: evidences from shoshonitic to ultrapotassic plutonic rocks. In: International Union of Geological Sciences (IUGS). In: 28th International Geological Congress, Washington, Abstract, Volume 3 (1989) Washington: International Union of Geological Sciences (IUGS). 106–107.
Sigmarsson O, Martin H, Knowles J. Melting of a subducting oceanic crust from U–Th disequilibria in austral Andean lavas. Nature (1998) 394:566–569.[CrossRef]
Smithies RH. The Archaean tonalite–trondhjemite–granodiorite (TTG) series is not an analogue of Cenozoic adakite. Earth and Planetary Science Letters (2000) 182:115–125.[CrossRef][Web of Science]
Smithies RH, Champion DC. High-Mg diorite from the Archean Pilbara Craton; anorogenic magmas derived from a subduction-modified mantle. Geological Survey of Western Australia, Annual Review (1999) 1998–1999:45–59.
Smithies RH, Champion DC. The Archean high-Mg diorite suite: links to tonalite–trondhjemite–granodiorite magmatism and implications for Early Archean crustal growth. Journal of Petrology (2000) 41:1653–1671.
Steiger RH, Jäger E. Subcommission on geochronology: convention on the use of decay constants in geo- and cosmochronology. Earth and Planetary Sciences Letters (1977) 36:359–362.[CrossRef]
Stern CR, Futa K. An Andean andesite derived directly from subducted MORB or from LIL depleted subcontinental mantle. Transactions of the American Geophysical Union (1982) 63:456.
Stern CR, Kilian R. Role of subducted slab, mantle wedge and continental crust in the generation of adakites from the Andean Austral Volcanic Zone. Contributions to Mineralogy and Petrology (1996) 123:263–281.[CrossRef][Web of Science]
Stern R. Petrogenesis of the Archaean sanukitoid suite. (1989) State University of New York at Stony Brook. 275. Ph.D. thesis.
Stern RA, Hanson GN. Archean high-Mg granodiorite: a derivative of light rare earth element-enriched monzodiorite of mantle origin. Journal of Petrology (1991) 32:201–238.
Stern RA, Hanson GN, Shirey SB. Petrogenesis of mantle-derived, LILE-enriched Archean monzodiorites and trachyandesites (sanukitoids) in southwestern Superior Province. Canadian Journal of Earth Sciences (1989) 26:1688–1712.
Stevenson RK, Henry P, Gariepy C. Assimilation–fractional crystallisation origin of Archean sanukitoid suites: Western Superior Province, Canada. Precambrian Research (1999) 96:83–99.[CrossRef][Web of Science]
Störmer J. C. Jr, Nicholls J. XLFRAC: a program for interactive testing of magmatic differentiation models. Computers and Geosciences (1978) 4:143–159.[CrossRef]
Streckeisen AL. To each plutonic rocks its proper name. Earth-Science Reviews (1976) 12:1–33.
Sun SS, McDonough WF. Chemical and isotopic systematic of ocean basalts: implications for mantle composition and processes. Magmatism in the Ocean Basins. Geological Society, London, Special Publications—Saunders AD, Norry MJ, eds. (1989) 42:313–345.
Sylvester PJ. Archean granite plutons. In: The Archean Crustal Evolution—Condie KC, ed. (1994) Amsterdam: Elsevier. 261–314.
Takahashi E, Kushiro I. Melting of a dry peridotite at high pressures and basalt magma genesis. American Mineralogist (1983) 68:859–879.[Abstract]
Takahashi E, Shimazaki T, Tsuzaki Y, Yoshida H. Melting study of a peridotite KLB-1 to 6·5 GPa and the origin of basaltic magmas. Philosophical Transactions of the Rotal Society of London (1993) 342:105–120.[CrossRef]
Tarney J, Weaver BL, Winley BF. Geological and geochemical evolution of the Archean continental crust. Revista Brasileira de Geociências (1982) 12:53–59.
Tatsumi Y. Melting experiments on a high-magnesian andesite. Earth and Planetary Science Letters (1981) 54:357–365.[CrossRef][Web of Science]
Tatsumi Y. Origin of high-magnesian andesites in the Setouchi volcanic belt, southwest Japan, II. Melting phase relations at high pressures. Earth and Planetary Science Letters (1982) 60:305–317.[CrossRef][Web of Science]
Tatsumi Y. Migration of fluid phase and genesis of basalt magma in subduction zones. Journal of Geophysical Research (1989) 94:4697–4707.
Taylor SR, McLennan SM. The Continental Crust: its Composition and Evolution (1985) Oxford: Blackwell. 312.
Teixeira W, Sabaté P, Barbosa J, Noce CM, Carneiro MA. Archean and Paleoproterozoic tectonic evolution of the São Francisco Craton, Brazil. In: Tectonic Evolution of South America. 31st International Geological Congress, Rio de Janeiro: In-fólio Produção Editorial Gráfica e Programação. (2000) 101–147.
Thirlwall MF, Smith TE, Graham AM, Theodorou N, Hollings P, Davidson JP, Arculus RJ. High field strength element anomalies in arc lavas: source or process? Journal of Petrology (1994) 35:819–838.
Toteu SF, Van Schmus WR, Penaye J, Michard A. New U–Pb and Sm–Nd data from north–central Cameroon and its bearing on the pre-Pan African history of Central Africa. Precambrian Research (2001) 108:45–73.[CrossRef][Web of Science]
Vanderhaeghe O, Ledru P, Thiéblemont D, Egal E, Cocherie A, Tegyey M, Milési JP. Contrasting mechanism of crustal growth. Geodynamic evolution of the Paleproterozoic granite–greenstone belts of French Guiana. Precambrian Research (1998) 92:165–193.[CrossRef][Web of Science]
Van Schmus WR, Brito Neves BB, Hackspacher PC, Babinski M. U/Pb and Sm/Nd geochronological studies of the Eastern Borborema Province, Northeastern Brazil, initial conclusions. Journal of South American Earth Sciences (1995) 8:267–288.[CrossRef][Web of Science]
Van Schmus WR, Brito Neves BB, Williams IS, Hackspacher PC, Fetter AH, Dantas EL, Babinski M. The Seridó Group of NE Brazil, a late Neoproterozoic pre- to syn-collisional basin in West Gondwana: insights from SHRIMP U–Pb detrital zircon ages and Sm–Nd crustal residence (TDM) ages. Precambrian Research (2003) 127:287–327.[CrossRef][Web of Science]
Vynhal CR, McSween HY, Jr, Speer JA. Hornblende chemistry in southern Appalachian granitoids: implications for aluminium hornblende thermobarometry and magmatic epidote stability. American Mineralogist (1991) 76:176–188.[Abstract]
Wasylenki LE, Baker MB, Kent AJR, Stolper EM. Near-solidus melting of the shallow upper mantle: partial melting experiments on depleted peridotite. Journal of Petrology (2003) 44:1163–1191.
Wilson M. Igneous Petrogenesis—a Global Tectonic Approach (1989) London: Chapman & Hall. 466.
Winther TK, Newton RC. Experimental melting of an hydrous low-K tholeiite: evidence on the origin of Archean cratons. Bulletin of the Geological Society of Denmark (1991) 39:213–228.[Web of Science]
Wolde B, Abebe B, Ayalew T, Megerssa B, Teklay M, Abraham A, Berhe K, Demissie M, Gitchile S, Meshesha S, Shagi T, Shiferaw A, Teferra M, Tesfaye T, Moore JM, Morgan J. Tonalite–trondhjemite–granite genesis by partial melting of newly underplated basaltic crust: example from the Neoproterozoic Birbir magmatic arc, western Ethiopia. Precambrian Research (1996) 76:3–14.[CrossRef][Web of Science]
Wolf MB, Wyllie PJ. Dehydration-melting of amphibolite at 10 kbar: the effects of temperature and time. Contributions to Mineralogy and Petrology (1994) 115:369–383.[CrossRef][Web of Science]
Workman RK, Hart SR. Major and trace element composition of the depleted MORB mantle (DMM). Earth and Planetary Science Letters (2005) 231:53–72.[CrossRef][Web of Science]
Wyllie PJ. Magmas and volatile components. American Mineralogist (1979) 64:469–500.[Abstract]
Wyllie PJ. Experimental and thermal constraints on the deep-seated parentage of some granitoid magmas in subduction zones. In: Migmatites, Melting and Metamorphism—Atherton MP, Gribble CD, eds. (1983) Nantwich: Shiva. 37–51.
Wyllie PJ, Sekine T. The formation of mantle phlogopite in subduction zone hybridization. Contributions to Mineralogy and Petrology (1982) 79:375–380.[CrossRef][Web of Science]
Zamora D. Fusion de la croûte océanique subductée: approche expérimentale et géochimique. (2000) Clermont-Ferrand: Université Blaise Pascal. 314. Ph.D. thesis.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||