Inferring tectonic provenance of siliciclastic rocks from their chemical compositions: A dissent

Abstract

Chemical compositions of siliciclastic sedimentary rocks are commonly used to infer their tectonic provenance. We have tested the universal applicability of the underlying principle in a small, but controlled study expecting 100% confirmation of the practice. A comparison is made between the chemical composition of the ~1480Ma Butler Hill Granite in an uplifted cratonic block of the St. Francois Mountain Igneous Complex and that of a small ~1-m-thick regolith body, a weathered granite sample, and the basal quartz arenites of the ~520Ma Lamotte Formation immediately above the regolith. The results show that in plots of K2O/Na2O vs. SiO2/Al2O3, the regolith and sandstone samples correctly plot in the Passive Margin field, although the weathered granite plots in the Arc field. In plots of Th-Sc-Zr/10 and La-Th-Sc, the results plot in the Passive and Active Continental Margins and their extensions. In other common plots to discriminate tectonic provenance (e.g., SiO2 vs. K2O/Na2O, Fe2O3+MgO vs. Al2O3/SiO2, Fe2O3+MgO vs. TiO2, Sc/Cr vs. La/Y) a few points plot in the Passive Margin field but scatter into and outside of other fields of tectonic provenances. The chondrite-normalized REE distributions show variable degrees of negative Eu anomalies, with flat HREE, conforming to a felsic source. The LREE distributions show both positive and negative Ce anomalies that can be ascribed to the variability of redox conditions during weathering and diagenesis of the original siliciclastic sediments. The variability of the Eu anomaly was likely affected by post-erosion processes in addition to whatever was inherited from the parent rocks. We conclude that chemical compositions can provide good clues, but are neither strong indicators nor unique identifiers of their tectonic provenance. Rather, they indicate a dominantly felsic or dominantly mafic, or a mixed set of source rocks.