Melt Inclusions in High-Grade Metamorphic Rocks

Abstract Melt inclusions (MI) in migmatites and granulites are one of the strongest microstructural criteria for the former presence of melt in high-grade metamorphic rocks and represent fundamental repositories to retrieve the composition of anatectic magmas in the source, as well as the nature of the fluid regime during anatexis. In this ‘Perspectives in Petrology’ article, we review what has been done on MI in metamorphic rocks in the last 15 years, revisiting the nomenclature and the recommended practices for their successful investigation. Various examples of metamorphic minerals hosting MI are presented, but the main focus is on garnet. Why garnet? Using phase equilibrium modelling, we explore the advantages of this mineral as the ultimate MI host in metamorphic rocks and contemplate what MI teaches us about garnet’s suprasolidus behaviour. MI commonly form clusters in the internal part of migmatitic and granulitic garnet, in contrast to phase equilibrium predictions that would indicate the beginning of garnet formation under subsolidus conditions. We present two alternative explanations (growth of garnet highly overstepped vs. complete garnet recrystallization under suprasolidus conditions), concluding that the second one is the most plausible. A complete database of major and trace elements and volatiles of anatectic MI is presented and used to discuss the fluid regime of the deep continental crust and the impact of anatexis on lithosphere differentiation. We also provide new insights into the debate ‘conservation vs. depletion’ of heat-producing elements (HPE) in the deep crust. Data suggest that only ultrahigh-temperature (UHT) metamorphism and formation of UHT anatectic melts may mobilize sufficient amounts of HPE, resulting in a HPE-depleted residual lower crust. Controversies on the origin of MI by partial melting of pre-existing mineral inclusions are discussed using phase equilibrium modelling. We conclude by proposing some directions to bridge the existing gaps and direct the future studies on this still promising field of research in crustal petrology.