Crustal forensics in arc magmas.

Abstract The geochemical characteristics of continental crust are present in nearly all arc magmas. These characteristics may reflect a specific source process, such as fluid fluxing, common to both arc magmas and the continental crust, and/or may reflect the incorporation of continental crust into arc magmas either at source via subducted sediment, or via contamination during differentiation. Resolving the relative mass contributions of juvenile, mantle-derived material, versus that derived from pre-existing crust of the upper plate, and providing these estimates on an element-by-element basis, is important because: (1) we want to constrain crustal growth rates; (2) we want to quantitatively track element cycling at convergent margins; and (3) we want to determine the origin of economically important elements and compounds. Traditional geochemical approaches for determining the contributions of various components to arc magmas are particularly successful when applied on a comparative basis. Studies of suites from multiple magmatic systems along arcs, for which differentiation effects can be individually constrained, can be used to extrapolate to potential source compositions. In the Lesser Antilles Arc, for example, differentiation trends from individual volcanoes are consistent with open-system evolution. However, such trends do not project back to a common primitive magma composition, suggesting that differentiation modifies magmas that were derived from distinct mantle sources. We propose that such approaches should now be complemented by petrographically constrained mineral-scale isotope and trace element analysis to unravel the contributing components to arc magmas. This innovative approach can: (1) better constrain true end-member compositions by returning wider ranges in geochemical compositions among constituent minerals than is found in whole rocks; (2) better determine magmatic evolution processes from core–rim isotopic or trace element profiles from the phases contained in magmas; and (3) constrain rates of differentiation by applying diffusion-controlled timescales to element profiles. An example from Nguaruhoe Volcano, New Zealand, underscores the importance of such a microsampling approach, showing that mineral isotopic compositions encompass wide ranges, that whole-rock isotopic compositions are consequently simply element-weighted averages of the heterogeneous crystal cargo, and that open-system evolution is proved by core–rim variations in Sr isotope ratios. Nguaruhoe is just one of many systems examined through microanalytical approaches. The overwhelming conclusion of these studies is that crystal cargoes are not truly phenocrystic, but are inherited from various sources. The implication of this realization is that the interpretation of whole-rock isotopic data, including the currently popular U-series, needs careful evaluation in the context of petrographic observations