The occurrence of S-type granitic rocks associated with thermal gneiss migmatitic domes in orogenic context is demonstrated by several authors for example at the Appalachians (Maine, USA) (Solar & Brown, 2001), Armorican Massif (St. Malo, France) (Milorde et al., 2001) and Bohemian Massif (Austria) (Vanderhaeghe, 2009). The existence of these thermal domes has a relation of cause-effect with syn-orogenic magmatism and metamorphism, The association of isochronous lateral thermal gradients more significantly by advective rather than by conductive heat transport is proved in this kind of orogenic context (Lux et al., 1986; Lancaster et al., 2008). This advection could be related with ascension of fluids, melts or magmas.

In the Iberian Massif some works have been developed about this theme (Viruete et al., 2000; Alcock et al., 2009; Díez Fernández et al., 2012). Since the 80's, multiple publications put in evidence the association of anatetic granites with gneiss-migmatite rocks at the Portuguese Variscan Orogeny (Holtz & Barbey, 1991; Moita et al., 2009; Bento dos Santos et al., 2010; Valle Aguado et al., 2010; Ribeiro et al., 2011; Areias et al., 2012; Ferreira, 2013; Ferreira et al., 2013).

Regarding the NW sector of Portugal, during the 1950/60's decades Carlos Teixeira referred the existence of migmatitic rocks associated to the “Oporto Granite” which is a two-mica syntectonic granite (Carríngton da Costa & Teixeira, 1957). In the last decades the granites outcropping in the eastern border of Porto Massif were described and dated taking in account their relation with variscan deformation phases, namely D₂ and D₃ (Pinto et al., 1987; Pereira et al., 1992; Mendes, 1967/1968; Martins et al., 2001).

The objective of this study is the geological, mineralogical and geochemical characterization of the Pedregal granite and its relation with surrounding magmatic and metamorphic rocks. The aim is to contribute to a better understanding of the relationship between orogenic granites, anatexis and metamorphic grade in orogenic context.

The studied granite is mapped on the Geological Map of Portugal scale 1/50.000 (9C-Porto) (Car-rington da Costa and Teixeira, 1957) and it has been classified as an alkaline two-mica granite. The Pedregal granite is a NW-SE elongated small granite body (ca. 3 km²) that is exposed to the north margin of the Douro river (Fig. 1). This pluton is intruded in staurolitic micaschists that belong to a metapelitic sequence of Edicarian-Cambrian age known as the “Complexo Xisto-Grauváquico” (CXG). The Pedregal granite body, with a SHRIMP U–Pb zircon age of 311 ± 5 Ma (Ferreira et al., 2014) is spatially associated with other synorogenic granites: Porto granite, a medium-grained alkaline two-mica granite with a zircon/monazite age of 318 ± 14 Ma (Martins et al., 2001, Gondomar granite, a coarse-grained alkaline tourmaline-rich granite) and Fânzeres granite, a foliated garnet-rich granite with a Rb-Sr age of 320 Ma (Pinto, 1987) (Fig. 1). The country rocks are also represented by gneiss-migmatite rocks. Locally, the contact is marked by igneous breccias with clasts of gneiss-migmatite and the Pedregal granite as matrix.

The Pedregal granite is a fine to medium-grained two mica granite, with small biotitic nodules (1 to 2 cm) that present an internal foliation whose direction ranges between NE-SW and E-W. This trend is opposite to regional NW-SE metamorphic structure and to the elongation of the pluton. It is a homogeneous rock, with an isogranular texture and locally the xenoliths have a preferential orientation.

The petrography analysis was done through the observation of seven samples of the Pedregal granite using a transmission microscope. Some samples were analyzed by SEM (Scanning Electron Microscope), specifically the EDS (X-ray microanalysis) method, for identification and qualitative analysis of accessory and opaque minerals. These methodologies were carried out in the laboratories of the “Departamento de Geociências, Ambiente e Ordenamento do Território” (DGAOT) and “Centro de Materiais da Universidade do Porto” (CEMUP).

The whole-rock geochemistry analyses for lithogeochemistry purposes were done in the Activation Lab in Ontario, Canada, to obtain the major and trace element composition. These analyses were performed in five samples of the Pedregal granite (same samples of the petrographic study). The lithogeochemistry analyses required a lithium metaborato/tetraborate fusion with subsequent ICP (Inductively Coupled Plasma) and ICP-MS (Inductively Coupled Plasma-Mass Spectrometer) (FUS-ICP-MS). Fused sample is diluted and analyzed by Perkin Elmer Sciex ELAN 6000, 6100 or 9000 ICP/MS. Three blanks and five controls (three before sample group and two after) are analyzed per group of samples. Duplicates are fused and analyzed every 15 samples. Instrument is recalibrated every 40 samples (http://www.actlabs.com/).

The Pedregal granite has a holocristallyne and heterogranular fine to medium-grained texture. The mineral assemblage consists of quartz + plagioclase + K-feldspar + biotite + muscovite + zircon + apatite + monazite + rutile ± sillimanite ± allanite ± Zn-rich hercynite.

The biotite crystals are elongated and corroded. The biotite-quartz and biotite-plagioclase intergranular boundaries (Fig. 2A) point to a textural disequilibrium suggesting that biotite could be a residual phase. The subhedral biotite crystals are often chloritized and contain small zircon and rutile needle-like inclusions. Small grains of no-altered biotite are included in subhedral muscovite. The elongated biotite crystals show preferential alignment and homogeneous distribution. In the small nodules the biotite is subhedral and less corroded than the elongated crystals. These nodules present compositional banding marked by biotite and secondary muscovite.

The quartz is subhedral to anhedral and when included in others minerals (muscovite and plagioclase) assumes a globular shape. Frequently, the quartz crystals show inclusions of sillimanite and rutile needles (Fig. 2B).

The albite-oligoclase plagioclase is more abundant than K-feldspar, although the later presents crystals with great dimensions (ca. 600 µm). The plagioclase crystals are subhedral and frequently altered.

The muscovite has a secondary character, distinguishing two types: a subhedral muscovite showing irregular borders and quartz droplet inclusions (Fig. 2C) and an euhedral muscovite, with anomalous birefringence (Fig. 2D). The zircon, sillimanite and hercynite are included in both types of muscovite.

The Pedregal granite is peraluminous (A/CNK parameter ranges from 1.18 to 1.62 and ASI ranges from 1.23 to 1.71), with a magnesian and alkali to alkali-calcic signature.

The major elements composition of the Pedregal granite shows relative low values of SiO₂ (65–69 wt. %) and high concentrations of Al₂O₃ (15.71–16.68 wt. %) and K₂O (5.21–5.96 wt. %) (Fig. 3). This granite has also high values of minor elements like TiO₂ and P₂O₅. The granite shows high contents in Zr (388–435 ppm), La (91.6–130 ppm), Ce (243–320 ppm) and Th (72.3–137 ppm) (Table 1). The Harker diagrams show the values of the Pedregal granite compared to the country rocks (staurolite schist and greywacke), and the adjacent Porto two-mica granite. The contents of TiO₂, P₂O₅, Zr and Th for Pedregal granite are higher than other plotted rocks. On the other hand, the content of alkalis is similar for the Pedregal granite and other plotted rocks.

The REE exhibit an almost flat pattern from La to Pr, but it doesn't continue towards Sm without significant fractionation of LREE (La/Sm)_N = 2.59–4.43) (Fig. 4). The heavy rare earth elements (HREE) show a pattern slightly more fractionated (Gd/Yb)_N = 5.74–8.89). The Eu negative anomaly (Eu/Eu* = 0.20–0.40) is pronounced (Fig. 4). The low HREE content is opposed by the high Zr concentrations (Fig. 5A), which suggests different mineralogical controls. On the other hand the high LREE and also Th content could be controlled by the presence of monazite (Bea, 1996) (Fig. 5B).

The trace element contents of the Pedregal granite has been compared with the average composition of materials similar to the country rocks (SGC): NIBAS (Neoproterozoic Iberian Average Shale) (Ugidos et al., 2010). In the spider diagram the Pedregal granite shows a positive anomaly for Zr, LREE, Hf, Th and U and a negative anomaly in Y, Yb, Lu Ta (Fig. 6A).

The moderate SiO₂ values (65–69 wt. %), high ASI parameter (1.23 to 1.71), high contents of LREE, Zr, and Th, and the contrasts with more differentiated granites such as the Porto two-mica granite (Fig. 6B) put in evidence a peraluminous granite (Pedregal granite) without significant magmatic differentiation.

Field, petrographical and chemical data of the Pedregal granite allow to emphasize the following features: (i) the intrusive character of Pedregal granite in the border of a synorogenic variscan granite-migmatite complex; (ii) the lithology of the host rocks consisting of staurolite-micaschist and banded gneiss-migmatite rocks; (iii) the peculiar structural/textural aspects of the granite, namely the abundance of small biotitic nodules and the corroded biotite crystals point to a textural reequilibrium in solid state; (iv) the geochemical composition, in particular the high peraluminous and low fractionated signature and the high content of Zr, Th, Ti and LREE.

The emphasized features suggest an origin by crustal partial melting with emplacement without magmatic differentiation.

Furthermore, according to Sylvester (1998) and Dong et al. (2013), the CaO/Na₂O and Al₂O₃/TiO₂ ratios give information about the nature of the magmatic source and the temperatures of partial melting in a strongly peraluminous system. In a granitic system, the CaO/Na₂O ratio of strongly peraluminous granites derived from plagioclase-poor and clay-rich sources are lower (<0.3) than that of plagioclase-rich and clay-poor sources (>0.3). The high Al₂O₃/TiO₂ ratios (>100) of strongly peraluminous granites point to cooler (<875 °C) granite melts and low Al₂O₃/TiO₂ ratios (<100) to hotter (>875 °C) ones. Although this information is qualitative, it is in agreement with Zr saturation temperatures (Watson & Harrison, 1983) higher than 890 °C for the Pedregal granite.

In the Pedregal granite, the CaO/Na₂O versus Al₂O₃/TiO₂ and Rb/Ba versus Rb/Sr diagram point to a mixture of pelitic and psammitic sources, and formation of granitic melts at high temperatures (Fig. 7).

Although some of the geochemical peculiarities of the Pedregal granite could be explained by secondary processes (mainly muscovitization), the presence of dispersed corroded biotite and concentrated in banded nodules, point to a diatexite rock. The intrusive character and the presence of metasedimentary xenoliths imply some vertical migration.

The slightly high content of Fe₂O₃ + MgO, and the high content of Zr, Th, Ti and LREE, could be explained by previous partial melting and segregation of a peritectic garnet-rich melt. This garnet was observed in field in the associated gneiss-migmatite rocks (Fig. 1), which may correspond to a first segregated melt. Therefore the source of the Pedregal rock was a residual biotite-rich pelitic/psammitic rock depleted by this first episode of melting. The high content of LREE, Th and U could be explained by the presence of monazite, and the high values of Zr and Ti by the abundance of rutile. The biotite and these associated accessories phases as source minerals may justify the geochemical features of the Pedregal granite.

Taking into account all the above considerations the Pedregal granite may be considered as a secondary diatexite, resulting from a second phase of crustal partial melting in prograde conditions, after the segregation of a first melt with peritectic garnet.