No.4 - Los leucogranitos equigranulares del Domo del Tormes (Zona Centro Ibérica): discriminación geoquímica mediante Biplot Canónico y significado petrogenético

Miguel López-Plaza, Francisco Javier López-Moro, José Luis Vicente-Villardón


About 65 % of plutonic rocks from the Variscan Tormes Dome (Central Iberian Zone) consist of equigranular leucogranites, including coarse-grained, medium-grained and fine-grained granite types. A crustal affinity (S-type granite melts) can be inferred, among others, from the following geochemical features: 1) a high A/CNK index; 2) negatively-slopped normalized spidergrams, showing an enrichment of incompatible and mobile elements, such as Rb and K, as well as negative anomalies in Nb, Ti, Sr, Ba and Eu; 3) high REE fractionation values, (La/Lu)n ranging from 28 to 75, and 4) high δ18O (10.3-12.9 ‰). Despite textural differences, mainly in relation to grain size variation, a compositional overlapping is worth noting, making any geochemical approach difficult to discriminate the two-mica granite groups. Variable partial melting degrees after glandular gneisses is unlikely to be the only reason to account for the linear trend on the ACF diagram, since such a trend is characteristic of leucogranite belts from NW Iberian Massif. To allow a better discrimination of the two-mica granite groups, a statistical methodology, the Canonical Biplot, was applied to a matrix consisting of 32 columns (major and trace-elements) and 42 rows (granite samples from the Tormes Dome). These latter, in turn, were firstly divided into seven granite groups (first analysis), including biotite granites and tourmaline granites, and secondly, into four two-mica granite groups (coarse-grained granites from central areas, coarse-grained granites from south western areas, medium-grained granites and fine-grained granites). Several biplot representations were obtained for the matrix adding confidence circles to the granite group markers in such a way that their circles were projected onto the direction representing a hypothetical magmatic differentiation vector. If no overlap is found we can conclude that there is a certain difference between granite groups, allowing us a geochemical discrimination. The biplot representation for the first analysis results in a 95.9 % inertia absorption for the first three axes. The main plane 1-3 allows a discrimination to be made by considering the projection onto a general differentiation vector that can include information on crustal emplacement level, mineralogy and the geochemical features themselves. In particular, the more shallowly emplaced medium-grained granites display a higher degree of differentiation than the coarse-grained granites from the deeper central area, both granite groups having similar chondrite-normalized patterns. The main plane 1-2 provides information that allows the role played by fluids to be discriminated, affording a fluid-enrichment factor in relation to the content in Cs and Ta, both elements showing similar behaviour. Tourmaline-bearing granites plot on the highest fluid-enriched side, followed by the peripheral coarsed-grained granites from the SW area, whereas the fine-grained granites appear on the “depleted side”, this latter suggesting a possible fluid-exsolution process linked to the late-orogenic decompressional evolution of the Tormes Dome. The biplot representation for the second analysis permits us to confirm these magmatic processes, leading to a perfect discrimination between two-mica granite groups using the residual plane 2-3 (23.7 % of the total inertia). The comparison of the statistical results with the chemical variation diagrams allow us to infer that the incompatible-element enrichment cannot be accounted for by variable partial melting degrees. Additional processes such as pelitic xenolith assimilation, restite unmixing and fluid enrichment are possible to have been involved to explain the enrichment in mobile elements.

PDF downloads:

685 (during 2009)


ariscan anatectic Tormes Dome, Equigranular leucogranites, Geochemical discrimination, Canonical Biplot, Differentiation factors

Full Text: PDF