DI31A-2551
Role of the subduction filter in mantle recycling

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Jun-Ichi Kimura, JAMSTEC Japan Agency for Marine-Earth Science and Technology, Kanagawa, Japan, Susanne E Skora, ETH Zurich, Zurich, Switzerland, James Gill, University of California Santa Cruz, Santa Cruz, CA, United States and Peter E Van Keken, University of Michigan Ann Arbor, Ann Arbor, MI, United States
Abstract:
Subduction modifies the descending basaltic and sedimentary oceanic crust and generates felsic arc materials and continental crust. Studies of element mass balances in the subduction zone therefore reveal the evolution of the Earth’s two major geochemical reservoirs: the continent crust and mantle. We use the Arc Basalt Simulator ver.4 (ABS4) to model the geochemical mass balance during dehydration by prograde metamorphism and melting of the slab followed by subsequent flux melting of the wedge mantle caused by the addition of slab-derived liquids. The geochemistry of high-Mg andesite or adakite formed in a hot subduction zone is akin to the present-day bulk continental crust and to the Archean (>2 Ga) Tonalite-Trondjhemite-Granodiorite composition. Therefore, the residual slab and the metasomatized mantle wedge at hot subduction zones should be the most plausible sources for materials recycled back into the deep mantle. Model calculations of isotopic growth in the residual slab and mantle formed in hot subduction zones reproduce fairly well the EM1–FOZO–HIMU isotope arrays found in ocean island basalts (OIBs) of deep mantle plume origin, although FOZO with high 3He/4He is not generated by this slab recycling process. The recycled materials are bulk igneous ocean crust for HIMU and metasomatized mantle wedge peridotite for EM1. In contrast, the EM2–FOZO array can be generated in a cold subduction zone with igneous oceanic crust for FOZO and sediment for EM2 sources. Necessary residence time are ~2 Ga to form HIMU–FOZO–EM1 and ~1 Ga to form EM2–FOZO. The subducted oceanic crust (forming HIMU) and mantle wedge peridotite (forming EM1) may have travelled in the mantle together. They then melted together in an upwelling mantle plume to form the EM1–FOZO–HIMU isotopic variations found frequently in OIBs. In contrast, the less frequent EM2–FOZO array suggests a separate source and recycling path. These recycling ages are consistent with the change in the mantle potential temperature over the past 3.5 Ga, which was high and constant at Tp = 1650 °C during 3.5–2 Ga before decreasing monotonously to Tp = 1300 °C today. This allows the formation of HIMU–EM1 sources in an ancient hot subduction zone, and formation of EM2–FOZO sources in younger, colder subduction systems.