Magmatic and volcanic evolution of a silicic large igneous province (SLIP): the Gawler Range Volcanics and Hiltaba Suite, South Australia

The Gawler Range Volcanics (GRV) and the co-magmatic Hiltaba Suite (HS) granite form a Mesoproterozoic Silicic-dominated Large Igneous Province (SLIP) cropping out over a vast area in the central Gawler craton, South Australia. Only a few SLIP have been recognised in the world; they occur throughout geological history and in both intraplate and plate margin settings. The igneous province in question, or Gawler SLIP, was emplaced in an intracontinental setting during assembly of the supercontinent Laurentia. Since emplacement, the Gawler craton remained an area of positive relief, allowing a good preservation of the Gawler SLIP. Emplacement of the Gawler SLIP is associated with a major mineralising event in the Gawler craton which includes the super-giant Cu-Au-U Olympic Dam deposit. This thesis focuses on the lower part of the Gawler Range Volcanics. By defi nition, SLIP are formed by large volumes of magma (‚Äöv¢‚Ä¢100000 km3) emplaced over a short time, in the order of the millions of years. The emplacement mechanism of large volumes of felsic magma in a short time span is one of the open questions in the study of these igneous manifestations. In this thesis, the volcanic facies in the lower part of the Gawler Range Volcanics have been described in detail, in order to assess the emplacement mechanism of the rocks. The rock units include moderately extensive (up to tens of km in diameter) felsic lavas, associated with pyroclastic deposits (ignimbrite) of comparable extent. Lavas are characterised by evenly porphyritic texture, with medium grained phenocrysts of feldspar ¬¨¬±pyroxene ¬¨¬±quartz in a fi ne microcrystalline quartzofeldspathic groundmass. Flow bands and flow folds, autobreccia domains, elongate vesicles and lithophysae are also present. Some of these large units may be composite lavas, separated by thin breccia layers of possible pyroclastic origin. Ignimbrites are compositionally and texturally homogeneous, and contain mm-scale crystals of feldspar and quartz in fine grained, eutaxitic and vitriclastic matrix. Some of the lavas have previously been interpreted as pyroclastic fl ow deposits in which all evidence of clastic origin have been concealed by welding. This interpretation was mainly based on the extent of the units and inferred high viscosity of felsic lavas. Volumetrically minor mafi c and intermediate lavas are also present and locally form thick piles. The chemical composition of the lower Gawler Range Volcanics has been determined, including major and trace elements. Whole-rock analyses are complemented by melt inclusion analyses, which allowed measuring volatile components (halogens in particular). The rocks are characterised by increasing trends of K2O, REE, Y, Zr, Th and Nb with increasing SiO2, and decreasing trends of CaO, FeO, MgO and TiO2. Fluorine is present in high concentrations in melt inclusions (F ‚Äöv¢¬ß1.3 wt.%, more than 20 times the average upper continental crust), whereas Cl has moderate concentrations. High microprobe totals of melt inclusions are compatible with low water contents, in agreement with the anhydrous parageneses and previous estimates based on petrological considerations (method of Nekvasil, 1988). Plots of Zr and the Zr/Hf ratio versus indicators of magma fractionation (e.g. SiO2 and incompatible elements) indicate that source magmas of the most felsic rocks (SiO2 >68 wt.%) were zircon-saturated. Application of the zircon saturation model (Watson and Harrison, 1983) on these samples yields temperatures up to 950-990¬¨‚àûC for the lower GRV. The combination of high magmatic temperatures, high F and low water in the magma creates conditions of low viscosity and low explosivity. These conditions are favourable for the effusive emplacement of the units, and help explaining how extensive units were emplaced as lavas. Comparison of whole-rock and melt inclusion analyses allowed assessing the degree of alteration of the rock units. Major element compositions are similar in wholerock and melt inclusion analyses, with the local exception of Na, which showed low concentration in a few whole-rock samples. This indicates that alteration did not affect substantially concentrations of the most mobile‚ÄövÑvp, water-soluble major elements. This is also true for most trace elements, and whole-rock analyses can be considered as indicative of the magma composition. Notable exceptions are Pb, U and Sn, which were selectively mobilised and variously depleted. Alteration of Pb, U and Sn is evident from scattered whole-rock compositions and lack of correlation between these elements and other elements. In contrast, melt inclusion compositions show good correlations and indicate incompatible behaviour. Quartz textures and trace element content in the lower Gawler Range Volcanics and Hiltaba Suite were studied by scanning electron microscopy cathodoluminescence (CL) and electron microprobe. Mineral zones retain a record of crystallisation conditions through time (crystal stratigraphy‚ÄövÑvp). Comparisons were made between volcanic units, shallow and deeper intrusions (dykes and granite). Quartz in volcanic units and dykes has sharp-edged CL zones in which CL brightness correlates with Ti content, whereas the granite quartz has smudged‚ÄövÑvp zones with gradational contacts. This difference is interpreted as the result of poor preservation of Ti zones in the granite, for which slow cooling allowed solid-state diffusion of Ti in the quartz lattice. Intra-granular textures in volcanic units and dykes also include truncation of growth textures and reverse zoning (rimwards increase of Ti content). Intra-granular textures indicate a complex history of crystallisation and resorption, and trace elements suggest varying temperature. These results point to pulsating magmatic conditions, compatible with a non-linear evolution of the lower Gawler Range Volcanics magma chamber(s). The volcanic units have contrasting (non-correlatable) zoning patterns among quartz crystals, each pattern indicating different crystallisation conditions. The juxtaposition of quartz crystals with contrasting zoning patterns are consistent with a dynamic regime (convection, stirring, overturning) of the GRV magma chamber. These results point to pulsating magmatic conditions, compatible with a non-linear evolution of the GRV magma chamber. In contrast, quartz crystals in the dykes have similar zoning patterns, suggesting that all the crystals in each dyke experienced a similar crystallisation history. In some rhyolite samples, aggregates of minerals (including fluorite, epidote, REE-F-carbonate, titanite, anatase, and zircon) have crystallised in pockets‚ÄövÑvp such as vesicles, micro-miarolitic cavities and lithophysal vugs. These aggregates of minerals contain significant amounts of rare earth (RE), high-fi eld strength (HFS) elements, and base metals (Cu and Mo). In average, concentration of these elements is higher in these aggregates than in the surrounding host rock. These aggregates are interpreted to have crystallised from a late-stage magmatic volatile (F, CO2, H2O, ¬¨¬±S, ¬¨¬±P)-bearing fluid that exsolved and infilled pockets during the final stages of emplacement and crystallisation. The presence of complexing agents such as F and CO2 can explain how low-solubility, immobile‚ÄövÑvp trace elements were transported in solution. A magmatic, primary origin for this fl uid appears likely, given the F-, REE-, and HFSE-rich composition of the melt shown by melt inclusions. Conversely, the hypothesis of a post-magmatic, secondary origin is considered less likely because of the absence of alteration, mineralisation and veins in these rocks. Magmatic accessory minerals, the alteration of which would have been necessary for secondary remobilisation of RE and HFS elements, appear fresh and unaltered. These data testify the mobility of RE and HFS elements in the lower Gawler Range Volcanics and can have implications in the formation of associated mineral deposits in the Gawler craton, including the Olympic Dam deposit. This deposit is also characterised by high concentrations of F, RE and HFS elements, and a similar F-rich fl uid as the one hypothesised here might have been active in the mineralisation process.