Magmatism And Mantle Dynamics In The Lau Back-arc Basin,SW Pacific

As a "special type" of divergent plate margin, back-arc spreading centers provide a useful window to understand the interplay between spreading rate and subduction influence, and how they collectively affect back-arc magmatism and mantle dynamics. The Lau Basin, located in the west limb of the Pacific, developed three spreading centers (CLSC, ELSC and VFR) with variable spreading rates, crustal thicknesses and distances to the arc. Besides, the opening of the Lau Basin has close link to the subduction of the Louisville Seamount Chain (LSC). Thus, it presents an unparalleled opportunity to study the issues on back-arc basin crustal accretion, mass transfer and its transport mechanism from subducted slab, as well as the influence of seamount subduction on back-arc magmatism. Here we systematically conducted a series of works on the magma evolution, mantle fertility, mantle domain boundary, the extent of mantle melting and the identification of subduction fluid end-members, by making use of major, trace and Pb-Sr-Nd isotopes of high-density sampling data from published literature and our new lavas.From CLSC-ELSC-VFR, the influence of subduction is roughly increased as the distance to the arc decreases southward. This results in (1) the enhanced influx of subduction fluid and larger mantle flux melting. The computing results show that the extent of mantle melting beneath CLSC [F=10.1±2.44(n=22)] is quite close to that beneath nELSC [F=10.4±1.63(n=52)], but much lower than those beneath sELSC [F=22.8±6.02(n=18)] and VFR [F=24.1±4.75(n=9)]. Additionally, that also results in (2) the increase of H2O content and oxygen fugacity (fO2) in the primitive magma, which makes the magma differentiation path towards enrichment in SiO2 and Al2O3, and depletion in TiO2 and FeO*, thus enhancing crustal differentiation in chemical composition. The crustal structure of CLSC and nELSC resembles the mid-ocean ridges that upper and lower crusts share similar crustal density. However this is not the case for sELSC and VFR, which develops a large thickness of upper crust with lower density (more enriched in Si-Al), and a lower crust with higher density (more enriched in Fe-Mg) than their counterparts of CLSC and nELSC. Consequently, the extent of mantle melting (F) that determines magma budget exerts first-order control on the crustal thickness in Lau Basin, while the magma differentiation path that determines the crustal density also plays a subordinate role.There are three subduction fluid end-members with distinct chemical composition that has been identified in the Lau Basin, each of which has its unique impacting region. "End-member 1" (SC-1) represents the subduction fluid composed of aqueous fluid released from altered oceanic crust and minor sediment melt, and has an influence on the region of CLSC, ELSC and VFR. Both the "end-member 2" (SC-2) and "end-member 3" (SC-3) are subject to the exotic subduction fluid bearing much higher percentage of hydrous melt than SC-1. The region influenced by SC-2 covers VFR and ELSC south of 21.2°S (region 2), while the region influenced by SC-3 is confined within CLSC (region 3). The rest region only influenced by SC-1 is the ELSC north of 21.2°S (region 1). As the projection of subducted LSC coincidently overlaps with "region 2" and "region 3" but northward deviates from "region 1", we conclude that SC-2 and SC-3 are subduction fluids released from the LSC. Another evidence supporting this inference comes from the finding that the composition of SC-1 is coupled with the inferred subduction fluid affecting normal arc islands of Tonga ridge, while the composition of SC-2 is coupled with Louisville-affected arc islands fluid.The onset of transition from LSC fluid to LSC melt occurs at rear-arc depths. LSC melt contributed 18-88% of the Ba budget,18-66% of the Pb budget and 0-83%% of the Th budget respectively to the lavas in "region 2", and contributed 10.8-73.8% of the Th budget respectively to the lavas in "region 3". This indicates that seamount subduction can not only promote the mass supply of subduction components, but also enhance element recycling at great subducted depths. In addition, the preferential melting of basaltic LSC crust instead of the normal basaltic Pacific oceanic crust at similar depth implies that elevated temperature across the subduction interface or seamount erosion and rupture were required to trigger partial melting.