Carbonation of subduction-zone serpentinite (high-pressure ophicarbonate; Ligurian Western Alps) and implications for the deep carbon cycling

Abstract

Much of the long-term carbon cycle in solid earth occurs in subduction zones, where processes of devolatilization, partial melting of carbonated rocks, and dissolution of carbonate minerals lead to the return of CO2 to the atmosphere via volcanic degassing. Release of COH fluids from hydrous and carbonate minerals influences C recycling and magmatism at subduction zones. Contradictory interpretations exist regarding the retention/storage of C in subducting plates and in the forearc to subarc mantle. Several lines of evidence indicate mobility of C, of uncertain magnitude, in forearcs. A poorly constrained fraction of the 40-115 Mt/yr of C initially subducted is released into fluids (by decarbonation and/or carbonate dissolution) and 18-43 Mt/yr is returned at arc volcanoes. Current estimates suggest the amount of C released into subduction fluids is greater than that degassed at arc volcanoes: the imbalance could reflect C subduction into the deeper mantle, beyond subarc regions, or storage of C in forearc/subarc reservoirs.We examine the fate of C in plate-interface ultramafic rocks, and by analogy serpentinized mantle wedge, via study of fluid-rock evolution of marble and variably carbonated serpentinite in the Ligurian Alps. Based on petrography, major and trace element concentrations, and carbonate C and O isotope compositions, we demonstrate that serpentinite dehydration at 2-2.5 GPa, 550 °C released aqueous fluids triggering breakdown of dolomite in nearby marbles, thus releasing C into fluids. Carbonate + olivine veins document flow of COH fluids and that the interaction of these COH fluids with serpentinite led to the formation of high-P carbonated ultramafic-rock domains (high-P ophicarbonates). We estimate that this could result in the retention of ~0.5-2.0 Mt C/yr in such rocks along subduction interfaces. As another means of C storage, 1 to 3 km-thick layers of serpentinized forearc mantle wedge containing 50 modal % dolomite could sequester 1.62 to 4.85 Mt C/yr.We stress that lithologically complex interfaces could contain sites of both C release and C addition, further confounding estimates of net C loss at forearc and subarc depths. Sites of C retention, also including carbonate veins and graphite as reduced carbonate, could influence the transfer of slab C to at least the depths beneath volcanic fronts

Subduction metamorphism of serpentinite‐hosted carbonates beyond antigorite‐serpentinite dehydration (Nevado‐Filábride Complex, Spain)

“…At different arc depths, reaction of these fluids with slab and mantle wedge serpentinites may produce ophimagnesite (magnesite–antigorite rocks), talc–magnesite rocks and listvenites (Falk & Kelemen, ; Menzel et al., ; Sieber, Hermann, & Yaxley, ), HP‐ophicarbonate (e.g. Scambelluri et al., ) and carbonate–enstatite rocks (Tumiati et al., ). Tectonic mixing and prograde metamorphism of serpentinite and carbonate‐bearing sediments in the slab interface may also generate hybrid rocks with a bulk composition similar to that of ophicarbonates (Scambelluri et al., ).…”

“…Exhumed metamorphic terranes provide essential insights into the mechanisms of deep carbon recycling and mobilization in subduction zones (Bebout & Penniston‐Dorland, ; Ferrando, Groppo, Frezzotti, Castelli, & Proyer, ; Piccoli et al., ; Scambelluri et al., ). Thermodynamic calculations and studies of eclogite facies marbles and carbonated eclogites and serpentinites in palaeo‐subducted metamorphic terranes show that carbonate minerals undergo variable extents of decarbonation and can be stable at high pressure (Collins et al., ; Connolly, ; Cook‐Kollars, Bebout, Collins, Angiboust, & Agard, ; Ferrando et al., ; Proyer, Mposkos, Baziotis, & Hoinkes, ).…”

“…Compilation of peak pressure–temperature (P–T) conditions of meta‐ophicarbonates (white fields) superposed on the pseudosection of a Ca‐bearing serpentinite from the Almirez massif. VM : Voltri massif (Scambelluri et al., ); ZS : Zermatt‐Saas; B: Bracco Unit, Internal Ligurides; UV : Ubaye Valley, Western Alps (all from Collins et al., ). L, L*: Lanzo massif peak conditions and PT of graphite‐formation in ophicarbonate, respectively (Vitale Brovarone et al., ); BA : contact‐metamorphic ophicarbonates, Bergell aureole (Pozzorini & Früh‐Green, ; Trommsdorff & Connolly, ); BD : Bellinzona‐Dascio unit, Central Alps (Stucki, ).…”

Fluid pathways and high‐P metasomatism in a subducted continental slice (Mt. Emilius klippe, W. Alps)

“…Our suggested scenario for the redistribution of carbonate within the Mt. Emilius massif is similar to the HP carbonation process documented for metasomatic marbles in the Ligurian and Corsican Alps (Piccoli et al ., ; Scambelluri et al ., ). At these localities, the C fraction of the metamorphic fluid, released by decarbonation and possibly also carbonate dissolution, was trapped within favourable sites during percolation through heterogeneous lithologies at depth.…”

“…Carbon and O isotope compositions are important proxies that can elucidate fluid transfer processes in metasomatic systems (e.g. Galvez et al, 2013;Ague & Nicolescu, 2014;Collins et al, 2015;Rubatto & Angiboust, 2015;Scambelluri et al, 2016). The d 13 C and d 18 O values of three calcite-bearing HP veins were obtained in order to evaluate the isotopic compositions and potential source(s) of fluids infiltrating the ASZ (see Fig.…”