Elemental evidence for crustal contamination of mantle-derived Caledonian syenite by metasediment anatexis and magma mixing

Abstract

This paper presents elemental evidence that the mafic unit of the Caledonian Glen Dessarry syenite (northern Scotland, U.K.) has been variably contaminated with material derived from the surrounding Moine metasediments. Uncontaminated portions are shown to have trace-element abundances and ratios similar to those of the central leucosyenite, and island-arc high-K or shoshonitic magmas from present-day oceanic settings. By comparison, contaminated samples display grossly disturbed trace-element patterns, including higher U, Th; Zr, Hf; Nb, Ta; and heavy rare earths; lower Ba and Sr; and negative Eu anomalies. These features are matched in the elemental systematics of the Moinian envelope, but bulk incorporation of excessive amounts of metasediment would be required to reproduce the observed perturbations in, especially, the high field-strength elements. Xenoliths within the contaminated mela-syenites, and parts of the adjacent country rock show petrographic evidence of melting. Since the source composition (Moine metasediment) and residual mineralogy (xenoliths) are known, simple geochemical models of melting and subsequent mixing with uncontaminated melasyenite melt have been computed. The results provide close approximations to the observed geochemistry of the contaminated melasyenites, particularly for the REE's. Moreover, the residues of this hypothetical process have similar REE patterns to several samples of metasediment from immediately adjacent to the syenite contact. Nevertheless, contemporaneous crystal-liquid separation is required to explain the full range of elemental data. Contamination by this mechanism of melting followed by magma mixing is not immediately apparent in the major-element relationships of the two melasyenite groups. This seems to be the result of the approximate colinearity of the established melasyenite crystal fractionation path and the proposed “granitic” contaminant, and may be a more general problem in the recognition of in situ contamination with metasedimentary melts in plutons whose fractionation itself proceeds towards a broadly granitic terminus.