Feedback of Slab Distortion on Volcanic Arc Evolution: Geochemical Perspective From Late Cenozoic Volcanism in SW Japan

Abstract

Southwest Japan is an island arc formed by subduction of the Philippine Sea (PHS) plate. The Quaternary magmatism in this region is characterized by eruptions of high-Sr andesites and dacites, considered to have been derived by melting of the PHS plate. The loci of these volcanoes spatially coincide with seismic discontinuities of the subducted PHS plate. Thus, the magmatism is interpreted as the result of slab melting at the plate tears. However, the processes that promote slab tearing remain unclear. In this study, we applied geochronological and geochemical analyses to late Cenozoic volcanic rocks in southwest Japan as tracers of slab morphology. Two different magma types, ocean-island basalt (OIB) and island-arc basalt (IAB), have occurred over 12 million years (Myr). These two magmas are attributed to different integrations of melts extracted from an originally fertile mantle; the OIBs from high temperature melt (1,300-1,400°C) were extracted at a depth of 40-80 km, whereas the IABs were extracted from a shallower, lower temperature region (30-60 km, 1,200-1,350°C). Secular change in Sr enrichment of IAB likely arose due to a transition of slab-derived fluids, incorporated into magmas as they formed, from water-to melt-dominant one. Progressive shallowing of the subducted PHS plate is responsible for secular change in the properties of slab-derived fluids as well as rollback of OIB volcanoes. Production of chemically variable magmas in the Chugoku district is the surface expression of distorting slab morphology by interaction between mantle and the subducting plate.

Petrogenetic implications and geochronology of middle Miocene Tannayama igneous rocks, Goto Islands, Japan Sea southern margin, northwestern Kyushu, Japan

“…However, these age data are insufficient for a comprehensive examination of the regional differences in magmatic ages and understanding the evolutionary histories of the magmatism represented by each rock. The K-Ar ages reported by Kimura et al (2003) and Nguyen et al (2020) bring relatively high reliability data within ±0.5 Ma. Moreover, the LA-ICP-MS zircon U-Pb geochronology has also provided reliable age data for the middle Miocene igneous rocks on the fore-arc side .…”

“…Error bars (±) for the zircon fission-track (FT) age data are all 2σ. Age data for Sanin are whole-rock or mineral K-Ar ages from Imaoka et al (1997), Kano and Yoshida (1984), Kimura et al (2005), Morris et al (1990), Nguyen et al (2020), Sawada (2018), and Shibata and Nishimura (1989) (1.1-0.3 Ma; Sudo et al, 1998), Iki volcano (5.3-0.7 Ma;Sano, 1995), and Fukue-jima volcano (0.51-0.018 Ma; Nagaoka & Furuyama, 2004) in northwestern Kyushu.…”

“…et al, 1997; Figure 1a), the quartz diorite of Maeubuyu (13.7 Ma ± 0.5 Ma; Imaoka et al, 1997; Figure 1a), and the granitic and gabbroic rocks of Shimane (18.0-14.4 Ma; Kano & Yoshida, 1984;Morris et al, 1990;Shibata & Nishimura, 1989; Figure 1a). Volcanic rocks in the Sanin region include rhyolite and andesite in Susa (14.2-12.1 Ma; Imaoka et al, 1997), and basalt, andesite, dacite, and rhyolite in Shimane (21.8-6.63 Ma;Kano & Yoshida, 1984;Kimura et al, 2005;Morris et al, 1990;Sawada, 2018;Nguyen et al, 2020;Shibata & Nishimura, 1989). The frequency distributions clearly indicate that the ages are concentrated in the last stage of the back-arc opening period (17-12 Ma), with a peak at around 15 Ma.…”

Bilateral Heterogeneity in an Upwelling Mantle via Double Subduction of Oceanic Lithosphere

“…The K‐Ar ages were obtained by analyses of Ar abundance by a noble gas mass spectrometer (Micromass VG5400) and K abundance by a flame photometer (Shimadzu AA‐6200) following T. T. Nguyen et al. (2020). Groundmass fractions were used for both K and Ar analyses.…”

“…Major‐element concentrations and Ni and Cr contents were determined by X‐ray fluorescence spectrometry (XRF) with a Philips PW2400 instrument, using lithium tetraborate glass beads (1:10 ratio of sample and flux) as outlined in T. T. Nguyen et al. (2020). Loss on ignition (LOI) was obtained by the gravimetric method; samples were heated at 1000°C in a furnace (>4 hr), and weight loss or gain values were measured using a balance.…”