Insights from trace element geochemistry as to the roles of subduction zone geometry and subduction input on the chemistry of arc magmas

Subduction zones of continental, transitional, and oceanic settings, relative to the nature of the overriding plate, are compared in terms of trace element compositions of mafic to intermediate arc rocks, in order to evaluate the relationship between subduction parameters and the presence of subduction fluids. The continental Chilean Southern Volcanic Zone (SVZ) and the transitional to oceanic Central American Volcanic Arc (CAVA) show increasing degrees of melting with increasing involvement of slab fluids, as is typical for hydrous flux melting beneath arc volcanoes. At the SVZ, the central segment with the thinnest continental crust/lithosphere erupted the highest-degree melts from the most depleted sources, similar to the oceanic-like Nicaraguan segment of the CAVA. The northern part of the SVZ, located on the thickest continental crust/lithosphere, exhibits features more similar to Costa Rica situated on the Caribbean Large Igneous Province, with lower degrees of melting from more enriched source materials. The composition of the slab fluids is characteristic for each arc system, with a particularly pronounced enrichment in Pb at the SVZ and in Ba at the CAVA. A direct compositional relationship between the arc rocks and the corresponding marine sediments that are subducted at the trenches clearly shows that the compositional signature of the lavas erupted in the different arcs carries an inherited signal from the subducted sediments.     

Constraining input and output fluxes of the southern-central Chile subduction zone: water,chlorine and sulfur

In this paper, we constrain the input and output fluxes of H₂O, Cl and S into the southern-central Chilean subduction zone (31°S–46°S). We determine the input flux by calculating the amounts of water, chlorine and sulfur that are carried into the subduction zone in subducted sediments, igneous crust and hydrated lithospheric mantle. The applied models take into account that latitudinal variations in the subducting Nazca plate impact the crustal porosity and the degree of upper mantle serpentinization and thus water storage in the crust and mantle. In another step, we constrain the output fluxes of the subduction zone both to the subcontinental lithospheric mantle and to the atmosphere–geosphere–ocean by the combined use of gas flux determinations at the volcanic arc, volume calculations of volcanic rocks and the combination of mineralogical and geothermal models of the subduction zone. The calculations indicate that about 68 Tg/m/Ma of water enters the subduction zone, as averaged over its total length of 1,480 km. The volcanic output on the other hand accounts for 2 Tg/m/Ma or 3 % of that input. We presume that a large fraction of the volatiles that are captured within the subducting sediments (which accounts for roughly one-third of the input) are cycled back into the ocean through the forearc. This assumption is however questioned by the present lack of evidence for major venting systems of the submarine forearc. The largest part of the water that is carried into the subduction zone in the crust and hydrated mantle (accounting for two-thirds of the input) appears to be transported beyond the volcanic arc.