ARE PROVENANCE INTERPRETATIONS FROM MAJOR, TRACE AND RARE EARTH ELEMENT DATA ACCURATE? RESULTS FROM FOUR ACTIVE MARGINS

The chemistry of clastic sediments, particularly fine-grained muds, is assumed to reflect a homogenized representation of the source rocks. It can therefore be used to determine the tectonic setting in which the sediment was deposited. However, there have been some concerns about the accurately of these assumptions, particularly the use of major element discriminant function diagrams and the risk of self-correlation in trace element plots using common denominator ratio pairs. This small study uses samples collected from ODP cores from four subduction zone accretionary wedges: the Nankai Trough (Japan), the northern Barbados Ridge, the Cascadia margin and the Costa Rica accretionary wedge.

Samples of hemipelagic muds from each location were analyzed to determine major, trace and rare earth element concentrations. From these data, provenance interpretations can be made using various established plots. Discriminant function diagrams for major elements have samples plot in fields representing mafic, intermediate, felsic or quartzose provenance. Similarly, trace and rare earth element cross plots divide the samples among mafic, intermediate or felsic provenance, and spider diagrams of normalized trace and rare earth element concentrations have profiles characteristic of the tectonic environment of deposition.

The samples from the Nankai Trough plot in the intermediate provenance field in all cases, as would be expected of sediments derived from a mature island arc like Japan. Barbados samples plot in both the mafic and intermediate fields for both major and trace elements, which is typical of an oceanic island arc. Cascadia and Costa Rica are both examples of active continental margins. However, their geochemistry and provenance interpretations are very different. The Cascadia margin sediments show a mixed source, while the Costa Rica sediments plot in the mafic provenance field. The agreement between the major and trace element plots is strong, suggesting that continental arc volcanic deposits are not the main source of sediment to the Costa Rica accretionary wedge.