The Samoan mantle plume is thought to host three isotopically (radiogenic) distinct low 3He/4He components including EM2 (enriched mantle 2), dilute HIMU (high μ = 238U/204Pb), and a depleted mantle (DM) component, which were sampled by shield stage lavas from the Malu, Vai, and Upo trend volcanoes, respectively.
Fig1. (a) Topographic map showing the distribution of Samoan islands and seamounts. (b) Simplified geological map of Tutuila Island.
However, it is unclear whether the isotopically distinct components are present as different lithologies. Using new Fe–Sr–Nd-Pb isotope data for Tutuila basalts (Samoa), combined with literature data for other Samoan basalts, we attempt to infer the lithological structure of the Samoan plume. The results show that “Malu trend” basalts have heavier Fe isotopic compositions (δ57Fe = 0.15–0.24‰) than “Vai trend” and “Upo trend” basalts. The latter two groups have average δ57Fe of 0.14 ± 0.07‰ (2SD) and 0.11 ± 0.03‰ (2SD), respectively, similar to normal midocean ridge basalts (N-MORBs, δ57Fe = 0.15 ± 0.05‰, 2SD). The fractional-crystallization-corrected δ57Fe values of all shield lavas are positively correlated with (Gd/Yb)N, Pb/Nd and 87Sr/86Sr ratios whereas negatively correlated with Nb/Th and εNd ratios, which cannot be explained by partial melting of a single garnet peridotite but point to heterogeneous source lithologies. The EM2 lavas are characterized with high δ57Fe and (Gd/Yb)N, low Nb/Th, and enriched Sr–Nd isotopic ratios, requiring a pyroxenitic source component with imprints of both recycled terrigenous sediments and oceanic crust. The Vai- and Upo-trend lavas with MORB like δ57Fe can be explained by partial melting of peridotitic sources, although different extents of refertilization by recycled crust are essential for generating their distinct radiogenic isotope signatures.
Fig2. Sr, Nd and Fe isotopic comositions of the studied Tutuila samples.
These observations highlight the lithological heterogeneity of the Samoan plume and relates the EM2 component with a pyroxenitic lithology.