SMALL PROPORTIONS OF REFILL BUT LARGE PROPORTIONS OF MIXING BETWEEN MAFIC AND SILICIC MAGMAS IN SHALLOW MAGMA CHAMBERS

Magma mixing involves thermal equilibration ahead of chemical homogenization. As a result, mafic magma crystallizes many tens of percent of microlites with normal zoning while the host silicic magma heats up and dissolves phenocrysts that develop reverse zoning. The residual melt in mafic magma (enclaves or bands) becomes nearly identical compositionally to the host silicic magma, and no subsequent chemical exchange is required. Thus, mixing of magmas is a physical process of disintegration of a suspension of microlites within enclaves in the main volume of common silicic melt. Here I review evidence of mafic microlites in silicic rocks, and consider natural evidence showing that intervals of normal zoning in mafic minerals are very large, while the intervals of reverse zoning in silicic phenocrysts are small. Given measured crystallinities in mafic enclaves, and intervals of mineral zoning in several volcanoes of Kamchatka, I offer a simple thermal model that allows calculation of proportions of mafic magma injections in silicic chamber. Magma mixing proportions based on these measurements are typically very small (1-8 vol %) in accord with a view that mafic magma can't be injected into the silicic magma chamber without creating a room problem. On the contrary, measured proportions of mixing based on whole-rock chemical compositions are far larger (25-50%). This is a common paradox in many igneous systems worldwide that is explained here by mafic magma vesiculation, flotation, and accumulation near the exit. Vesiculated mafic enclaves are therefore overabundant in extrusive domes, and high proportion of mixing are achieved by their mechanical disintegration before and in the course of eruption. Size distribution of the enclaves measured in many domes worldwide provide some light into their formation and disintegration processes. Plutonic equivalent of volcanic exit channels may be represented by composite dikes where proportion of mafic enclaves changes from 10 to >50% and I present evidence of mechanical disintegration from nicely exposed composite dikes in Transbaikalia. I also consider an implication that flotation of gas-rich mafic enclaves leads to the ρgh increase in pressure in the whole magma system; this, along with vesiculation itself, contributes to the overall pressure build up and eruption.