The geochemical and Sr[sbnd]Nd isotopic characteristics of Paleozoic fractionated S-types granites of north Queensland: Implications for S-type granite petrogenesis

Abstract

Moderately to strongly fractionated S-type granites crop out extensively (>2500km2) in the central and eastern parts of the Hodgkinson Province, north Queensland, Australia. The granites have been subdivided in two major supersuites: the garnet-bearing Whypalla and cordierite-bearing Cooktown Supersuites; and a number of minor suites—including the extremely fractionated Wangetti and Mount Alto Suites. Early formed magmatic tourmaline is a feature of the Wangetti and Mount Alto granites. Almost all of the S-type granites contain metasedimentary enclave material, while microdioritic enclaves are mostly notably absent. The S-type granites are felsic with a moderate SiO2 range (68–77%). Most elements are negatively correlated with increasing differentiation, including TiO2, FeOtot, MgO, CaO, Ba, Sr, Th, LREE, Eu, Zr, Hf, and ratios such as K/Rb; many decrease to very low levels. There are very few positively correlated elements: Rb, U, and to some extent Na2O. Geochemical differences between supersuites include higher CaO, Ba, Sr, Pb, and lower Rb, Sn, B, V in the Whypalla Supersuite. Geochemical variation within the granites is largely due to extensive crystal fractionation. Some of the S-type granites have FeO* and MgO contents of 2.5–3.0% or more indicating they do not represent simple sedimentary melts, but rather represent the presence of both cumulate and restitic material. Variable Nd and Sr signatures (εNd between −2 and −6.5; initial Sr ratios between 0.709 and 0.715), suggest multiple components. The S-type granites intrude a very extensive, siliciclastic turbidite sequence that is isotopically evolved (e.g., εNd mostly −12.0 to −15.0 at 270Ma), and generally too mature (too CaO poor) to produce S-type granites. Isotopic and chemical modeling show that although magma-mixing is permissible, the levels permissible (<ca. 20–25% basaltic input), are not large enough to explain the signature of the granites. Instead the data suggest that the S-type granites were derived from unexposed metasedimentary source rocks at depth, consistent with the geology and accretionary tectonic models for the region. Comparison of the S-type granites with nearby similarly extensively fractionated Carboniferous I-type granites shows contrasting geochemical behaviour which appears to relate to ASI-driven changes in solubility of accessory phases. Fractionating peraluminous magmas, like the North Queensland S-type granites, are characterised by increasing ASI, P2O5 and strongly decreasing Th, LREE, HREE, Zr and Y. In contrast, P2O5 strongly decreases while ASI, Th, LREE, HREE, Zr and Y increase or are constant within the I-type granites. These differences, which are also observable between I- and S-type granites elsewhere, provide evidence of how the nature of the source can control the behaviour of even the most fractionated granites.