Effects of melt loss, melt retention, and protolith composition on differentiation of anatectic metapelites: A case study of the Wet Mountains, Colorado

Abstract

Two metasedimentary migmatite samples from the same outcrop in the northern margin of the Oak Creek pluton in the Wet Mountains, Colorado, have distinct mineralogy, textures and major and trace element geochemistry. Sample W6, a garnet-biotite migmatite, has trace element characteristics consistent with a sedimentary protolith, this sample is also rich in silica and lacks sillimanite and cordierite. Sample W6A, a garnet-sillimanite-cordierite migmatite, has strong heavy rare earth element (HREE) enrichment, low silica, and contains abundant aluminum-rich minerals. Both rocks are relatively oxidized, consistent with sedimentary protoliths for both. Based on textural and geochemical evidence, we interpret the differences in mineralogy between the two samples to be the result of differing degrees of melt extraction and variations in protolith composition. Such differences were assessed using isocon analysis and calculated P-T phase diagrams. Melt retention in sample W6 resulted in the absence of sillimanite and cordierite, and back-reaction between minerals and melt in the peak assemblage. By contrast, melt extraction in W6A left garnet in the residuum, and imparted a strong heavy rare earth element enrichment on the sample, allowing preservation of the peak mineral assemblage. Consequently, W6A records peak conditions of 743 °C and ~7.6 kbar, whereas W6 yields lower P-T estimates that may represent the prograde and retrograde metamorphic paths. Our results indicate that melt extraction can be heterogeneous even at outcrop scale, and differing mineralogy may reflect variable proportions of melt loss and local variations in the composition of the protolith. Moreover, at least at small-scale, migmatization of iron-rich protoliths is a plausible mechanism for the production of ferroan, alkaline granitic magmas.