Defining Geochemical Signatures and Timescales of Melting Processes in the Crust: An Experimental Tale of Melt Segregation, Migration and Emplacement

Abstract

Both the mechanism of melt segregation and the nature of the short-range transport paths developed during partial melting can exert significant control over melt geochemistry because these variables influence the degree of equilibrium achieved between the melt and the solid restite. We suggest that the physical process of melt segregation will therefore leave a chemical signature that may be detected in the anatectic melts and the rocks which melted to form them. The experimental studies described below suggest that there are distinct differences in composition between melts derived statically and distributed in crack networks from those formed in permeable shear zones. These differences in composition can be exploited by studies of natural anatectic granites to estimate rates of melting, extraction and crystallization. As these melts rise into the upper crust, assimilation can also become an important contributor to final magma composition. We describe the use of radiogenic isotopes because they are a powerful means of assessing crustal melting and contamination models. However, the evidence presented here for disequilibrium suggests that the isotopic composition of crustal melts may not necessarily be the same as their source. We have seen that the use of trace elements may be problematic as well (e.g. REE in our pelite melting experiments) in that melt compositions may deviate from those predicted using established phase equilibrium and mineral-melt partitioning data. If melting in the crust really is a disequilibrium process, would it not make constraining source and contaminant signatures a less tractable problem? We propose that in accepting that this can happen, we are now in a position to make real progress. We propose that the isotopic and trace element composition of crustal melt is a function not only of the composition of its precursor, but also of the processes by which and rates at which it is formed and extracted. Therefore, if we can establish a general framework for modelling the compositions of crustal melts, we then have a powerful tool to constrain the mechanisms and timescales over which melts are extracted or assimilated. The results are consistent with a growing body of theoretical and field evidence indicating that processes, ranging from melting in the lower crust to assimilation in the upper crust, are rapid, efficient and can occur over timescales of decades or less.