The trace element compositions of S-type granites: evidence for disequilibrium melting and accessory phase entrainment in the source

Byline: Arnaud Villaros (1), Gary Stevens (1), Jean-Francois Moyen (1), Ian S. Buick (1) Keywords: S-type granite; Trace element; Partial melting; Geochemical modelling; Accessory phases; Cape Granite Suite Abstract: Within individual plutons, the trace element concentrations in S-type granites generally increase with maficity (total iron and magnesium content and expressed as atomic Fe + Mg in this study) the degree of variability in trace element concentration also expands markedly with the same parameter. The strongly peraluminous, high-level S-type granites of the Peninsular Pluton (Cape Granite Suite, South Africa) are the product of biotite incongruent melting of a metasedimentary source near the base of the crust. Leucogranites within the suite represent close to pure melts from the anatectic source and more mafic varieties represent mixtures of melt and peritectic garnet and ilmenite. Trace elements such as Rb, Ba, Sr and Eu, that are concentrated in reactant minerals in the melting process, show considerable scatter within the granites. This is interpreted to reflect compositional variation in the source. In contrast, elements such as LREE, Zr and Hf, which are concentrated within refractory accessory phases (zircon and monazite), show well-defined negative correlations with increasing SiO.sub.2 and increase linearly with increasing maficity. This is interpreted to reflect coupled co-entrainment of accessory minerals and peritectic phases to the melt: leucocratic rocks cannot have evolved from the more mafic compositions in the suite by a process of fractional crystallisation because in this case they would have inherited the zircon-saturated character of this hypothetical earlier magma. Trace element behaviour of granites from the Peninsular Pluton has been modelled via both equilibrium and disequilibrium trace element melting. In the disequilibrium case, melts are modelled as leaving the source with variable proportions of entrained peritectic phases and accessory minerals, but before the melt has dissolved any accessory minerals. Thus, the trace element signature of the melt is largely inherited from the reactants in the melting reaction, with no contribution from zircon and monazite dissolution. In the equilibrium case, melt leaves the source with entrained crystals, after reaching zircon and monazite saturation. A significant proportion of the rocks of the Peninsular Pluton have trace element concentrations below those predicted by zircon and monazite saturation. In the case of the most leucocratic rocks all compositions are zircon undersaturated whilst the majority of the most mafic compositions are zircon oversaturated. However, in both cases, zircon is commonly xenocrystic. Thus, the leucocratic rocks represent close to pure melts, which escaped their sources rapidly enough that some very closely match the trace element disequilibrium melting model applied in this study. Zircon dissolution rates allow the residency time for the melt in the source to be conservatively estimated at less than 500 years. Author Affiliation: (1) Department of Geology, Geography and Environmental Studies, Centre for Crustal Petrology, Stellenbosch University, Private Bag X1, Matieland, South Africa Article History: Registration Date: 25/02/2009 Received Date: 06/10/2008 Accepted Date: 25/02/2009 Online Date: 24/03/2009 Article note: Communicated by T. L. Grove. Electronic supplementary material The online version of this article (doi: 10.1007/s00410-009-0396-3) contains supplementary material, which is available to authorized users.