SUBDUCTION-ZONE GEODYNAMICS OF CARBON CYCLING: EVIDENCE FROM STUDY OF HP/UHP METAMORPHIC SUITES

HP/UHP metamorphic rocks provide opportunities to evaluate the significance of metamorphism, along the subduction pathway, for the evolution of major geochemical reservoirs. Geochemical study of forearc metamorphic rocks has recently proliferated, largely aimed at making connections with the evidence of subduction output in arc magmas and records of deep mantle compositions. Our work on C behavior in HP/UHP suites integrates field, petrographic, petrologic (including theoretical), and isotopic observations for rocks in the W. Alps and indicates retention of the majority of the C subducted through most modern forearcs in sediments, oceanic crust, and carbonated slab ultramafic rocks (Cook-Kollars et al., 2014, Collins et al., in review; both in Chemical Geology).

Work on subduction zone C cycling is best done for individual margins for which we have the greatest constraints on subduction geometry, lithologies entering the trench (and their chemical and isotopic compositions), accretionary/erosional character, and thermal structure. An analysis of the Central America margin considering C arc output flux estimates from de Leeuw et al. (2007; EPSL) indicates that, at this margin, a large fraction of the initially subducting C (perhaps as much as 50%) is delivered into the mantle beyond the subarc. On a global basis, the oxidized:reduced ratio for C currently delivered into subduction zones (in sediment, AOC, and carbonated ultramafic rocks) could be ~4:1, resulting in an influx with bulk δ¹³C_VPDB near -6‰ (Bebout, 2007; Treat. Geochem). However, the oxidized:reduced C ratio and thus the bulk δ¹³C of the subducting slab varies considerably among modern margins (e.g., oxidized:reduced ratio of ~11:1 for Central America; Li and Bebout, 2005; JGR) and both can be modified by differential metamorphic C release from the two reservoirs.

Work to date on chemical cycling in HP/UHP rocks has been conducted largely on rocks from cool forearcs (15-90 km depths). Our work points to the need to better understand the roles of higher-T devolatilization, partial melting, and carbonate dissolution in releasing C from (and otherwise geochemically modifying) subducting slabs at depths of ~80-100 km where they experience heating by exposure to the convecting mantle wedge (see the recent thermal modeling by van Keken et al., 2011; JGR).