A review of plagioclase growth rate and compositional evolution in mafic alkaline magmas: Guidelines for thermometry, hygrometry, and timescales of magma dynamics at Stromboli and Mt. Etna

Abstract

Mafic alkaline magmas, such as those feeding the persistent eruptive activity of Stromboli and Mt. Etna volcanoes in Italy, are dominated by the crystallization of plagioclase via cooling and degassing phenomena related to the dynamics of shallow crustal reservoirs and eruptive conduits. Because plagioclase textures and compositions are extremely sensitive to the changes of intensive variables in subvolcanic plumbing systems, the phenomenological variability of erupted crystals preserves detailed evidence of complex growth histories. From this point of view, we reappraise the textural maturation and compositional complexity of plagioclase by allying thermodynamic and kinetic principles to natural and experimental observations, with the purpose of drawing up guidelines for reconstructing magma dynamics in mafic alkaline volcanic settings. A multifaceted statistical method is adopted to parameterize the decay of crystal growth rate with increasing crystallization time, as relaxation kinetics prevails over melt supersaturation effects. This model parameterization is combined with the textural analysis of natural plagioclase crystals to quantify the residence time of phenocrysts in equilibrium with magmas at Stromboli and Mt. Etna and/or the timescale of rapid microlite growth during disequilibrium ascent of magmas within the conduit. The role played by temperature and melt-water content on plagioclase components and major cation substitution mechanisms is also evaluated under both isobaric-isothermal and decompression conditions. The emerging paradigm is that the influence of dissolved water on anorthite-albite exchange between plagioclase and melt is overwhelmingly mitigated by changes in temperature at conditions of P = 30–300 MPa, T = 1050–1150 °C, fO2 = NNO + 1.9-NNO + 2.3, and melt-H2O = 0.6–4.4 wt%. As a corollary, anorthite and albite melt activities are almost fully encapsulated in the variation of anhydrous melt components as the crystallization of plagioclase proceeds during magma cooling. Following this line of reasoning, we propose an integrated modeling approach to decipher complex zoning patterns in natural plagioclase phenocrysts from mafic alkaline eruptions. Key findings from our re-assessment of equilibrium, thermometric, and hygrometric models indicate that temperature and dissolved water can be iteratively estimated for different plagioclase textural patterns if crystals are sufficiently strongly zoned and probability-based criteria are applied to determine the maximum probability distribution from kernel density analysis.