#1 The production and modification of continental crust is an integral part of plate tectonics and involves the transfer of melt through the lower crust to mid and upper crustal levels. This talk summarises the different modes of melt transfer recognised in the lower crustal sections of the well-exposed Mesozoic magmatic arc of Fiordland, New Zealand, involving: (1) diffuse and channelized porous melt flow under conditions of low differential stress, (2) syntectonic, channelized porous melt flow and (3) brittle failure allowing melt transfer via dyking. Each mechanism has distinct field, microstructural and geochemical signatures that can be used to identify them. At the same time these signatures inform about the details of the processes involved. Common to all three mechanisms is the inference that the system is open and that the migrating melt is externally derived. Hence, it is likely to be in chemical disequilibrium with the host rocks through which it migrates. The chemical potential drives melt-rock reaction and the development of complex microstructures, microchemistry and rock textures. Analogous to aqueous fluid-rock interaction, features typical of reactive transport of melt through the crust are common and include reaction fronts, finger structures and rapid replacement of the host assemblage by a distinct, high variance assemblage by coupled dissolution-precipitation. The key field relationships and microstructural and microchemical fingerprints of reactive melt ascent are summarised to enable others to recognise pathways of melt migration in other settings.
#2 Locations including East Antarctica are well known for rocks that exhibit complex geochronology based on U-Pb data for zircon and monazite grains that spreads close to concordia over tens to a few hundreds of Myr. Traditionally, the oldest analyses are used to infer the age of igneous crystallisation or a high-T metamorphic event, whereas the youngest ages point toward the timing of a Pb-loss event. While the isotopic and trace element characteristics of zircon and monazite have been well characterised, clear links to microstructural patterns are often lacking, as for example where core domains are dated as younger than rim domains. This talk presents newly published data for granitic melt-monazite reaction experiments and compares the compositions and textures of the reaction products to those of natural samples from East Antarctica and elsewhere. The experiments resulted in a range of complex textures that are attributed to both dissolution and coupled dissolution-precipitation processes. The microstructure of natural zircon from both gabbroic and granitic rocks and monazite from a range of rocks are comparable to the modified grains in our experiments. Additionally, the complexly modified isotopic signatures, including ages, of the monazite reacted in the experiments mimic natural complex data sets. The complex textures and age-data patterns of natural zircon and monazite grains are interpreted as the result of melt-mediated coupled dissolution-precipitation reactions acting on pre-existing zircon and monazite grains. This process skews apparent ages towards the age of melt-mineral interaction. Therefore, significance is placed on the youngest grains to date high-T anatectic events. We highlight that zircon and monazite grains modified via coupled dissolution-precipitation during melt-rock interaction may not faithfully record the age or duration of metamorphism in melt-present systems and caution against relying on complex data sets for such interpretations. These datasets are in some cases unlikely to be geologically meaningful.