Intellectual Merit. The role of water in subduction zones is largely understood based on behavior of fluid-mobile elements such as B, Ba, K, Rb and U, rather than on measurements of water directly. This is because most studies of arcs rely on data from subaerial lavas which have suffered volatile loss through degassing. Published experimental studies and melt inclusion studies across the Central American arc suggest that water is released continuously from the slab across the arc. This is contrary to the behavior of fluid mobile elements, which are progressively stripped from the slab and show systematic cross-arc changes. A fundamental aim of this study is to evaluate the release of water, trace elements, and stable isotopes of hydrogen and boron across two contrasting segments of the IBM system - the Marianas and the Izu-Bonin arcs. These sites have different dip angles and subduction rates, thus we will determine whether these physical parameters have any impact on fluid behavior during progressive dehydration. The second objective of this study is to test how well trace element and B-isotopic proxies for slab fluids track fluid delivery to IBM mantle sources, and whether D/H and B isotopes trace similar fluid release processes or whether each system uniquely traces different stages of dehydration. We propose a systematic study of volatiles (CO2, H2O, S, Cl, and F), fluid soluble elements (B, Cl, Ba, K, Rb and others) and stable isotopes (H, B) in glasses and olivine hosted melt inclusions (MIs) from submarine lavas using ion microprobe techniques. This work will be complemented by multicollector ICPMS analyses of B isotopes, as well as electron microprobe measurements of major elements. The samples include a new suite of cross-chain volcanoes from the Marianas, back-arc lavas and glasses from the Mariana Trough, and lavas from the Izu-Bonin arc and back-arc. The correlation (or lack thereof) between arc magma water contents and fluid-mobile trace elements will have an important impact on the interpretation of geochemical signatures at other subduction zones with existing trace element and isotopic data but where water contents have not yet been measured.
Broader Impacts. This project will complement ongoing MI studies investigating the melting process in the Lau Basin, as well as the scientific objectives of the MARGINS Subduction Factory initiative, and will foster collaboration between scientists at five institutions in the US and Japan. The project will support a female early career scientist at Woods Hole Oceanographic Institution (WHOI). As part of a WHOI and NSF-REU summer intern programs, the PI will mentor an undergraduate student during the 2nd year of the project, introducing them to ion microprobe techniques. This work will also result introduce new analytical protocols to help revitalize the NSF-supported WHOI ion probe facility which was seriously damaged in a fire in October of 2002.
The main goal of this study was to evaluate the release of water, trace elements, and stable isotopes of hydrogen and boron across contrasting segments of the Izu-Bonin-Mariana system (Fig. 1). The three cross arc segments investigated (Torishima-Horeki, Suiyo and Guguan) have different dip angles and subduction rates, thus allowing us to determine whether physical parameters have any impact on fluid behavior during progressive dehydration. Based on a systematic analysis of melt inclusions and submarine glasses from our focus sites, we find that both hydrogen and boron isotopes decrease with progressive dehydration across all arc systems. Likewise, certain trace element proxies for fluids including Ba/La and B/Zr show a general decrease across both the Mariana and Suiyo arc systems. However, the Torishima-Horeki system shows an increase in concentrations of trace element proxies for water across the arc system towards the rear-arc. We attribute this finding to a delay in the bulk of fluid release due to a lower temperature gradient at this northernmost site leading to higher concentrations in rear-arc lavas. To further investigate the relationship between H and B isotopes in subduction settings, we analyzed a suite of glasses from the Manus Back-arc Basin. We found strong correlations between B and H isotopes (Fig. 2) indicative of an ancient dehydrated slab in the mantle source. Both of these isotope systems are positively correlated with water and trace element fluid proxies. These results have important implications for the fate of water in the mantle and the timescales over which diffusive equilibration are thought to occur. Based on our data and current experimental data, subduction-related H isotope and B isotope anomalies can be preserved over timescales of several hundred million years (Shaw et al., 2012). Finally, we received a supplement award to investigate volatile and trace element relationships in olivine-hosted melt inclusions from the fast-spreading East Pacific Rise (EPR) and the intermediate-spreading Juan de Fuca Ridge (JdFR). Through this work, we were able to provide constraints on melt compositions and depths of crystallization beneath the two ridges. Results from this work indicate that the average crystallization depth of the olivines is at or near the seismically-imaged melt lens depths on both spreading centers. However, a significant proportion of the melt inclusions (40% on the JdFR and 30% on the EPR) have pressures consistent with crystallization in the lower crust or mantle (Wanless and Shaw, 2012). Therefore, these results suggest that all models of crustal accretion on mid-ocean ridges must include crystallization in the lower crust, effectively ruling out all purely top-down or "gabbro glacier" models of ocean crust formation.
