Intra-oceanic Subduction Zones

Abstract

Modern intra-oceanic subduction zones comprise around 17,000 km (~40%) of the convergent margins of the Earth and are subjects of intense cross-disciplinary studies that are reviewed in this chapter. Most of these subduction zones exhibit trench retreat, do not accrete sediments and are affected by back-arc extension processes. Initiation of intra-oceanic subduction zones is partly enigmatic although two major types of subduction nucleation scenarios are proposed: induced and spontaneous. Internal structure and compositions of intra-oceanic arcs are strongly variable. Both along- and across-arc variation of crustal thickness and lithological structure are inferred based on seismological data. In the base of intra-oceanic arcs, a crust–mantle transitional layer is often detected by seismic velocity studies. This layer of debatable thickness and origin is interpreted as being the result of deep magmatic processes and may potentially include cumulates, replacive rocks and intercalation of various rocks and melts. The vast majority of basaltic magmas erupted in the arcs are too MgO-poor to represent the parental high-MgO mantle-derived magma. Magma fractionation and reactive flow models are suggested to explain this MgO-paradox, the uncertainty is related to yet limited knowledge of deep crustal, and mantle processes under the arcs. Exhumation of high- and ultrahigh-pressure crustal and mantle rocks during intra-oceanic subduction is strongly controlled by a serpentinized subduction channel developing at the plate boundary. This channel composed of intermixed rheologically weak crustal and mantle rocks presumably forms by hydration of the overriding plate and accretion of subducted upper oceanic crust. At a mature stage of subduction the channel may also incorporate newly formed magmatic arc crust and depleted mantle rocks from the base of the arc lithosphere. An array of diverse both clockwise and counterclockwise P–T paths rather than a single P–T trajectory is characteristic for high-pressure rock melanges forming in the serpentinized channels. Crustal growth intensity in intra-oceanic arcs (40–180 km3/km/Myr) is variable in both space and time and may strongly depend on subduction rate as well as on intensity and character of thermal-chemical convection in the mantle wedge driven by slab dehydration and mantle melting. Such convection can possibly include hydrated diapiric structures (cold plumes) rising from the slab and producing hybrid magmatic rocks by melting of subducted rock melanges. Subduction-related arc basalts characteristically have elevated contents of large-ion lithophile element (LILEs) and light rare earth element (LREEs) with depleted heavy REE (HREE) and high field strength elements (HFSEs) compared to subducted oceanic crust. The exact origin of geochemical variations in arc basalts is debatable and may involve a range of processes such as (a) extraction of fluids and/or melts from the subducted slab, (b) fluid-fluxed and decompression melting of the mantle wedge, (c) slab melt–mantle reactions, (d) melting of mantle wedge metasomatized by slab-derived fluid or melt, (e) direct supply of felsic melt from eclogitic slab melting, (f) melting of hydrated mantle and subducted tectonic melanges in thermal-chemical plumes. Water concentrations in back-arc mantle sources increase toward the trench. Back-arc basin spreading combines mid-ocean-ridge-like adiabatic decompression melting with nonadiabatic fluid-fluxed mantle melting depending on the H2O supply from the subducting plate. Numerical modelling results predict that water-rich mantle sources should mainly concentrate at 100–250 km distances from the trench in proximity of water-rich, depleted and chemically buoyant “cold nose” of the mantle wedge. In conclusion, despite recent progress in both observation and modelling many of the first-order features of intra-oceanic subduction remain only partly known and require further cross-disciplinary efforts.