The nature of upper mantle is important to understand the evolution of the South China Sea (SCS); thus, we need better constrains on its mantle heterogeneity. Magma water concentration is a good indicator, but few data have been reported. However, the rarity of glass and melt inclusions and the special genesis for phenocrysts in SCS basalts present challenges to analyzing magmatic water content. Therefore, it is possible to estimate the water variations through the characteristics of partial melting and magma crystallization. We evaluated variations in Fe depletion, degree of melt fractions, and mantle source composition along the fossil spreading ridge (FSR) using SCS basalt data from published papers. We found that lava from the FSR 116.2° E, FSR 117.7° E, and non-FSR regions can be considered normal lava with normal water content; in contrast, lava from the FSR 117° E-carbonatite and 114.9–115.0° E basalts have higher water content and show evidence of strong Fe depletion during the fractional crystallization after elimination of the effects of plagioclase oversaturation. The enriched water in the 117° E-carbonatite basalts is contained in carbonated silicate melts, and that in the 114.9–115.0° E basalts results from mantle contamination with the lower continental crust. The lava from the 117° E-normal basalt has much lower water content because of the lesser influence of the Hainan plume. Therefore, there must be a mantle source compositional transition area between the southwestern and eastern sub-basins of the SCS, which have different mantle evolution histories. The mantle in the west is more affected by contamination with continental materials, while that in the east is more affected by the Hainan mantle plume.

Highly depleted peridotites from the Yap Trench in the western Pacific Ocean have been studied for Re-Os elements and Re-Os isotopes. These peridotites have a low Re-Os content and variable ¹⁸⁷Os/¹⁸⁸Os ratios (0.12043–0.14867). The highest ¹⁸⁷Os/¹⁸⁸Os ratio is far higher than that of the primitive upper mantle and the lowest ¹⁸⁷Os/¹⁸⁸Os ratio is comparable to the most unradiogenic ¹⁸⁷Os/¹⁸⁸Os ratio (0.11933) discovered in subduction zone peridotites. The suprachondritic ¹⁸⁷Os/¹⁸⁸Os ratios of the Yap Trench peridotites results from modification of the mantle wedge by slab-derived fluid and melt. This is consistent with the observation that high ¹⁸⁷Os/¹⁸⁸Os ratios generally occur in oceanic peridotites with low Os content (<2 ppb) since Os may be reduced during late processes such as fluid alteration and melt refertilization. The sub-chondritic ¹⁸⁷Os/¹⁸⁸Os ratios of the Yap Trench peridotites correspond to a Re depletion age of 0.24–1.16 billion years, which means that these peridotites represent old mantle residue of ancient melting events. This ancient melting, combined with probable back-arc melting and forearc melting during subduction initiation, indicates that the Yap Trench mantle has a complex evolutionary history. The amount of old mantle residue in the oceanic asthenosphere was underestimated because the ¹⁸⁷Os/¹⁸⁸Os ratio in mantle peridotites is elevated during late processes. Therefore, old depleted mantle fragments may contribute substantially to the chemical heterogeneity of the oceanic mantle.

The Kyushu–Palau Ridge (KPR) is the result of the subduction of the Pacific Plate, separated from the proto-Izu–Bonin–Mariana arc by the back-arc spreading, showing variations in arc structures. The volcanism related to the subducted slab at the remnant arc stops before the initial back-arc spreading. However, the KPR accretion did not stop right when the back-arc spreading started. The internal structure not only reveals the arc characteristics but also contains crisis information about how the arc formed. The less revealed velocity structure in the southern KPR might hinder an understanding. Herein we studied the crustal magmatism at the southern KPR based on the detailed velocity structure. Results show that the KPR’s crust to the west is 6.0–7.6-km-thick with one summit at the basement and a flat Moho boundary; to the east, the crust has a thickness of 7.0–11.0 km, and three summits are observed at the basement sharing one depressed Moho boundary. These structural diversities indicate the presence of two independent seamounts (west and east), with different periods of magmatism. By calculating the duration of volcanism, we hypothesized that the formation of the western seamount of the KPR requireed a magmatism duration of ~4 Myr and one stable volcanic source. In contrast, the eastern seamount needed two volcanic sources and a total magma loading time of ~10 Myr with ~19-Myr period gap in magmatism. By analyzing the KPR evolution, we conclude that multiple volcanic sources and episodic magmatism caused the diversities in arc structures at the KPR; the initiation of back-arc spreading at ~26 Ma might have inhibited the magmatism related to the subduction; however, the back-arc-related partial remelting of the lower island arc would have led to a new magma source and reconstructed the preexisting island arc, thus leaving multiple conduits inside it.