Absence of water content contrast between lithosphere and asthenosphere in subduction zone

Abstract

The distribution of water between the rigid lithosphere and ductile asthenosphere is crucial for our understanding of the Earth's dynamics because it has been proposed that the origin of the lithosphere–asthenosphere boundary (LAB; i.e., the bottom of the tectonic plate) is related to a possible elevated water content in the asthenosphere. However, obtaining the depth profile of water content using mantle xenoliths is still challenging because of the rapid diffusive loss of hydrogen during magma ascent and eruption. In this study, we present diffusion profiles of water determined by FTIR mapping and SIMS line analysis for olivine and orthopyroxene of peridotite xenoliths from Ichinomegata maar in Northeast Japan. Extensive analysis of water contents in 17 xenolith samples revealed that these xenoliths exhibit a variety of zoning patterns, from which the timescale of diffusive loss was estimated. The xenoliths from Ichinomegata underwent only limited diffusive water loss (<1h duration), because of the low temperature of andesitic host magma and rapid quench in the maar deposit. Water contents of the mantle value are well-preserved in the homogeneous core parts of large olivine and pyroxene grains. The water contents of olivine suggest clear correlations with petrogenetic factors excepting those marking anomalous high values, implying hydrous metasomatism by a water-rich fluid. Combining the estimated water contents of the minerals and depth estimation of the xenolith samples, for the first time, we obtained the depth profile of water content of the mantle, including the LAB, down to ∼55 km. Our results show that olivine, orthopyroxene, and clinopyroxene in the lithospheric mantle (depth range in 28–38 km) contain 21 ± 2, 302 ± 64, and 616 ± 99 wt. ppm H2O, respectively, which are similar to those in the top of the asthenosphere (depth range in 39–52 km) that contain 20 ± 2, 258 ± 38, and 561 ± 80 wt. ppm H2O. In the region which has experienced local metasomatism, higher water contents are recorded (30 ± 4, 414 ± 48, and 741 ± 43 wt. ppm H2O). This study verifies that there is no water content contrast across the LAB in our studied area. Therefore, we support the ‘partial-melting model’ for the origin of the LAB rather than the ‘olivine-water model’ in the western Pacific Plate subduction zone.