Champion, David Christopher2013-11-11b1815735xhttp://hdl.handle.net/1885/10695North Queensland has been the site of widespread and voluminous recurrent felsic magmatism, with granite emplacement occurring in the Middle Proterozoic, in the Silurian and Devonian, and in the Carboniferous and Permian. In the Georgetown and Dargalong lnliers, the Proterozoic S-type Forsayth and Esmeralda Supersuites (FSS and ESS, respectively) outcrop over 1300 km2 and comprise potassic, Lll...E-, Th- and U-enriched granites. The geochemical variation in the FSS appears to have been produced dominantly via restite-unmixing. Fractional crystallisation was the most operative mechanism in the ESS and in the felsic members of the FSS. Geochemical and Nd-Sr isotopic data suggest that the FSS granites were most probably derived from gneissic metasediments similar to those presently outcropping in the Georgetown Inlier. The ESS granites, which are more felsic than the FSS and very Pe-rich, also appear to have been largely derived from a protolith similar to that of the FSS, but their compositions invoke the input of an Pe-rich component; the latter was most probably similar to Pe-rich pelitic sequences in the Georgetown Inlier. The pre-Carboniferous 1-type granites outcrop over 4000 km2 and consist of the Middle Proterozoic Forest Home and Blackman Gap Supersuites and the Siluro-Devonian White Springs and Dido Supersuites. All these granites have similar overall geochemistry, characterised by high Ah03, CaO, Na20, Ba, Sr, and Eu, and low to moderate K20, Rb, Th, U andY. The relative abundances of Sr andY imply generation of these magmas in thickened crust near or below the gabbro-eclogite transition. Nd and Sr isotope signatures vary from primitive (chondritic) to evolved (tNd of 0.0 to -8.0) requiring the participation of at least two end members. Source rocks for the Forest Home Supersuite were most probably Middle Proterozoic basaltic rocks while those for the Blackman Gap and White Springs Supersuites, which have more potassic end­ members, were most likely mixtures of this basaltic precursor and older (Lower Proterozoic?) basaltic to andesitic rocks. The Dido Supersuite granites were derived from a significantly younger basaltic protolith,that was possibly underplated at 1550 Ma, i.e. at the time of generation of the Middle Proterozoic granites. There is little data to support the direct involvement of contemporaneous mantle-derived material in the pre­ Carboniferous granites. The range in calculated Nd depleted mantle model ages for the pre-Carboniferous I- and S-type granites (1800-2500 Ma.) is consistent with estimates for the age of the Georgetown Inlier (Black & McCulloch, 1984, 1990).The extensive Carboniferous felsic 1-type granites (CIG) outcrop over 8000 km2 in far north Queensland, of which over 80% have Si02 greater than 70%. The granites are post-t ctonic and were emplaced in a tensional environment, either a back-arc or post­ subduction setting, with an overall decrease in age of 330 Ma to 280 Ma from west to east.The CIG have been subdivided into four supersuites: the Almaden, Ootann, Claret Creek and O'Briens Creek Supersuites (ASS, OSS, CCSS, and OBSS, respectively). The OSS and OBSS granites all contain greater than 70% SiD2 and comprise over 90% of the CIG. Dating and field relations suggest that the highly fractionated OBSS are older than or are similar in age to the OSS and ASS granites. The OSS and OBSS are undoubtedly crustal melts with low Sr, Sr/Y, and large negative Eu/Eu* and evolved initial87Srf86Sr and tNd• With increasing differentiation, the OSS and OBSS granites have become strongly depleted in TiD2, Ai203, FeO*, MnO, MgO, CaO, Ba, Sr, Sc, V, Cr, Ni, Eu, (Ce/Y)N, and K/Rb, and enriched in Rb, Pb, Th, U, and Rb/Sr. Although containing petrographic evidence that implies subsolidus re­ equilibration, element migration was largely localised and, consequently, geochemical trends preserve their magmatic signature. Geochemical variation was predominantly controlled by large amounts of crystal fractionation of primarily plagioclase along with, in the most felsic rocks, quartz and alkali feldspar. The OBSS differ from the OSS in having significantly higher HFSE, HREE and F (0.2 to 0.5+ wt%). The ASS granites outcrop over 500 km2. The granites are intermediate to felsic (56% to 72% Si02) and are characterised by high K20, K/(K+Na), Rb, Rb/Sr, Th, U, moderate La/Nb, and relatively low Ba and Sr. Major element contents are similar to some high-K orogenic rocks although the ASS granites lack the high Ba and Sr found in the latter rocks. K20, Na20, Rb, Pb, Th, U, Nb andY behave incompatibly, increasing with increasing Si02. The geochemical and isotopic evidence indicate that the ASS granites are also crustal melts. The CCSS granites form a minor component of the CIG, outcropping over 150 km2. They are geochemically distinct from all other granites of the CIG, being characterised by high Na20 and Sr, and low K20, Rb, U andY. The CCSS granites were derived at greater depths than the other CIG, from a LILE-poor source. The depleted LILE nature of the source may have been due to high grade metamorphism which is known to have have taken place in the lower crust (Rudnick & Taylor, 1987). The granites of the ASS, OSS, CCSS and OBSS have very similar initial 87Srf86Sr and ENd of 0.710 and -7.0 to -8.0, respectively, except those granites west of the Palmerville Fault which outcrop within Proterozoic country rocks; these latter granites have more evolved ENd (-8.0 to -11.0). The differences in ENd across the Palmerville Fault can be explained by minor amounts of assimilation. Depleted-mantle model ages for the CIG cluster around 1.5 Ga, similar to reported ages for the Middle Proterozoic deformation and metamorphic event (Black et al., 1979) and to the Proterozoic granites (Black & McCulloch, 1990). The isotope systematics of the granites are generally less evolved than the Proterozoic and Early Palaeozoic country rocks, and more evolved than contemporaneous, unrelated, mafic (mantle-derived) rocks in the region. Models for the petrogenesis of the CIG all appear to require the involvement of a long-lived, isotopically-homogeneous crustal protolith that was underplated in the Proterozoic. Granites of the OSS and OBSS were either derived by varying degrees of partial melting of this protolith of andesitic to dacitic composition and/or were produced by a two stage process by remelting of intermediate rocks similar in composition to members of the ASS. Regardless of the model invoked, it appears most likely that the very felsic OSS and OBSS granites were derived from a protolith geochemically similar to the mafic end members of the ASS. In the eastern Hodgkinson Province 3000 km2 of Permian felsic 1- and S-type granites outcrop, of which the former comprise over 80%. The S-type granites comprise two major supersuites, the Cooktown and Whypalla Supersuites, and five minor supersuites. The older Whypalla Supersuite (280 Ma) differs from the Cooktown Supersuite (260 Ma) in having higher CaO and Sr, and lower Ti02, Rb, V and Ni. These differences are best explained by invoking a greater pelite involvement in the source rocks for the Cooktown Supersuite granites. Geochemical variation in the S-type granites is largely controlled by crystal fractionation. The exposed sediments of the Hodgkinson Province could not have been the source rocks for the S-type granites, the former have a more evolved Nd isotope signature and are too poor in CaO, P205, and Sr. The source rocks for the S-type granites appear to have been less isotopically evolved and less mature volcaniclastic sediments that were most probably deposited in a Lower to Middle Palaeozoic arc environment. The Permian 1-type granites outcrop over 500 km2 intermingled with the S-type granites. The granites belong to two supersuites, namely the Yates and Cape Melville Supersuites; in many aspects they are geochemically similar to the Carboniferous Almaden and Ootann Supersuites, respectively, with moderate to high levels of K20, Rb, Th and U. The I-type granites are crustal in origin, most probably derived from andesitic to dacitic source rocks. The 1-type granites have similar ENd values to the Permian S­ type granites and similar to less evolved initial 87SrJ86Sr. ENd values for the Carboniferous and Permian granites exhibit a progressive increase from west to east roughly corresponding to the decreases in ages of the granites (330 Ma to 260 Ma). Crustal models based on the granite petrogenesis suggest that this decrease in ENd is due to a corresponding decrease in crustal ages from west to east which reflects both vertical and lateral accretion onto pre-existing Proterozoic basement.en-AU199110.25911/5d7787810520e