Reactive bulk assimilation: A model for crust-mantle mixing in silicic magmas

Abstract

Abstract Bulk assimilation of small (millimeters to ∼1 km) fragments of crust—driven and (ultimately) masked by reactions during xenolith melting and magma crystallization—is an important mechanism for crust-mantle mixing. Xenoliths containing mica or amphibole undergo dehydration melting when incorporated into a host magma, yielding mainly plagioclase, pyroxene, Fe-Ti oxides, and hydrous melt. The xenolith is physically compromised by partial melting and begins to disintegrate; xenolithic melt and crystals are mixed into the host magma. Xenocrystic zircon is liberated at this stage. The cryptic character of assimilation is greatly enhanced in any hydrous magma by hydration crystallization reactions (the reverse of dehydration melting). All pyroxenes and oxides (phenocrysts, xenocrysts, or crystals having a hybrid signature) will be subject to these reactions, producing feldspars, amphiboles, and micas that incorporate material from several sources, a particularly effective mixing mechanism. Implicit in the model is a reduced energy penalty for bulk assimilation—much of the assimilant remains in solid form—compared to melt-assimilation models. A large role for bulk assimilation supports stoping as a credible mechanism for the ascent of magmas. While the assimilation of low-density crust and concomitant fractionation provide the isostatic impetus for ascent, the wholesale incorporation and processing of crustal rocks in the magma chamber helps create the room for ascent.