Two-stage hybrid origin of Lachlan S-type magmas: A re-appraisal using isotopic microanalysis of lithic inclusion minerals

Microanalysis of accessory minerals was carried out to elucidate the controversial origin of the classical, ~430 Ma (Silurian) S-type granites and related lithic inclusions from the eastern Lachlan Orogen, southeastern Australia. Inclusions were separated into igneous-derived, mafic microgranular enclaves (MME) and metasedimentary inclusions. Apatite ɛNd(t), and monazite ɛNd(t) isotopic values for granite host and inclusions cover a similar range (~ −4 to −13), but their ɛNd(t) peaks are distinct. Whereas similar unimodal peaks exist for both minerals in the granites, suggesting widespread equilibration during hybridisation of two magmatic components in the host S-type magma, skewed or bimodal peaks characterize the MME and metasedimentary inclusions, indicating their different petrogenetic histories. The isotopic data suggests two endmembers: a mature metasedimentary source with ɛNd(t) of −12, represented by ubiquitous Ordovician quartzose turbidites, and a more radiogenic component represented by a − 5 ɛNd(t) peak in the MME. Zircon ɛ(Hf) values from −3 to −10, and δ18O values from 7.5‰ to 10.5‰, reflect variable incorporation (40–80%) of metasedimentary material with a mantle-like endmember. Temperature estimates for magmatic equilibration are 750-800 °C, though evidence exists for transient higher temperatures in some inclusions. Phase equilibria modelling using the average composition of the Ordovician turbidites as a potential source-rock suggests that melt compositions were uniform, silicic (75–76 wt% SiO2) and peraluminous (ASI = 1.13–1.22), for a wide range of temperatures and water contents. However, neither the modelled residual rock (restite) compositions, nor the analysed metasedimentary inclusions, lie along the Lachlan S-type compositional array. Rather, the compositional array projects to a mafic endmember defined by the MME, suggesting widespread, efficient hybridization with weakly peraluminous, andesitic magmas. Source mixing models (without significant mantle-derived magma input) are rejected on petrological and geological criteria. The andesitic MME-type magmas were hybridized in the lower crust, probably during mafic magma underplating, before incorporation into large, cool, midcrustal S-type magma reservoirs, and the observed MME represent the final mingled stage of that interaction during emplacement at shallow crustal levels. Using a modern example from the Andes, a model is presented of 2-stage hybridization, initially near the Moho, then in the midcrust as hot, hybrid andesitic magmas repeatedly infused a resident, giant, low-T, S-type magma body. Late-stage injection of the MME generated localized inclusion swarms, as at Cowra and Deddick, and promoted eruption of vast S-type ignimbrite sheets during a major magmatic flare-up in the Lachlan orogen.