The First 10 Million Years of Rear‐Arc Magmas Following Backarc Basin Formation Behind the Izu Arc

Abstract

IODP Site U1437 is located in the Izu rear-arc region, approximately 330 km west of the Izu-Bonin trench axis. The oldest four units (Units IV through Unit VII) include volcaniclastic sediment and in situ hyaloclastites. They have ages of about 6-15 Ma, shortly after cessation of Shikoku backarc basin opening. Three magma types are identified by their distinct geochemistry; they are similar types to those found in the modern Izu arc (Rear Arc Seamount Chain [RASC]-type, Rift-type, and volcanic front [VF]-type). RASC-type has the most enriched Nd and Hf isotope and fluid-immobile trace element ratios and dominates from 9 to 6 Ma. Rift-type, dominant from 15 to 9 Ma, is similar to VF-type in Nd-Hf isotopes but has the least radiogenic Sr and Pb, and intermediate La/Yb and Nb/Yb, indicating a more fertile mantle source and less hydrous slab component than VF-type. Less common and randomly distributed VF-type sediments have the most radiogenic Sr and Pb, and the highest Ba/(Th, LREE [light rare earth element]) ratios, and are interpreted to be distally derived. The genesis of mafic Unit VII samples (~15 Ma) was modeled using the Arc Basalt Simulator. Results are most similar to those for basalts in the modern rift environment indicating the addition of~1% of a melt-rich slab component generated at~125 km, to a Philippine Sea Plate ambient mantle that was more depleted than DMM (depleted MORB mantle). The initial post-Shikoku basin magmatism in the Izu rear-arc generated Rift-type magmas for about 6 million years before the distinctive RASC-type magmatism began, which then became increasingly enriched. Plain Language Summary We explore the history of magma genesis in the Izu rear-arc, Japan, by studying the geochemistry of volcaniclastic sediments that were obtained by ocean drilling. This paper presents their Sr-Pb-Nd-Hf isotope ratios and numerical models of magma genesis. We identify the presence of three geochemical types that change through time. Rift-type formed soon after cessation of seafloor spreading in the Shikoku Basin at~15 Ma and came from a more fertile mantle source with a less hydrous slab component than at the contemporaneous volcanic front. This type is most like basalts from rifts behind the current volcanic front and was unknown before drilling. RASC-type dominates from 9 to 6 Ma and has more enriched Nd and Hf isotope and fluid-immobile trace element ratios that are like those of rocks dredged from the coeval rear-arc seamount chains. VF-type is uncommon, randomly dispersed, and interpreted as derived from the distal volcanic front. The results confirm that the arc front has migrated trenchward since the Miocene.

Basalt Geochemistry and Mantle Flow During Early Backarc Basin Evolution: Havre Trough and Kermadec Arc, Southwest Pacific

“…Those features reflect a distinctive combination of the extent of prior mantle depletion, the depth and degree of mantle melting, and the nature and amount of the slab component. In Izu, the S‐shaped REE pattern in particular has been shown to indicate 2%–4% melting at ∼1.5 GPa of mantle more depleted than DMM to which ∼0.5% of hydrous melt was added from a slab 120–140 km deep (Miyazaki et al., 2020). We infer variations on this theme along and across the strike of the HT and suggest that the S‐shaped REE patterns may characterize the rifting stage of backarc development in general because they seem to be absent from the spreading stage of the Lau Basin and Mariana Trough.…”

“…They predate the current phase of rifting in Izu and cannot be traced across its ∼100‐km wide extensional zone to connect to arc front volcanoes. They did not develop until several million years after “rift‐type” volcanism (Miyazaki et al., 2020; Sato et al., 2020). Therefore, the messy spatial distribution of slab components in the HT may have a temporal dimension as well, and be commonplace during rifting and prior to spreading.…”

Formation of back-arc basins by lithospheric warping: examples of the Andaman, Bismarck and Banda Sea basins

“…But formation of back-arc basins in front of advancing hinges at Izu-Bonin (Miyazaki et al, 2020) and Mariana (Wu et al, 2016) subductions raises concerns about the universality of such a mechanism. Furthermore, the fact that in some cases (Japan Sea Basin, South Fiji Sea Basin) back arc extension came to an end while subduction continued indicates that subduction is not a sufficient condition for back-arc opening.…”