Co-Occurrence of HIMU and EM1 Components in a Single Magellan Seamount: Implications for the Formation of West Pacific Seamount Province

Abstract

ABSTRACT Our general understanding of mantle composition and dynamics mainly comes from the composition of lavas from oceanic intraplate volcanoes. They are generally accepted to originate from deep, relatively stationary mantle plumes. Many groups of seamounts comprising the West Pacific Seamount Province (WPSP), however, do not form long-lived, narrow and continuous chains of volcanoes with clear age progression; thus, their origin does not seem to fit this general model. Here we show a wide compositional spectrum of lavas from the Pako guyot within the Magellan seamount trail (MST), an age-progressive but short-lived volcanic chain in the WPSP. For the first time, both extreme high μ = 238U/204Pb mantle (HIMU)- and enriched mantle 1 (EM1)-like mantle plume components occur in a single seamount in the Pacific Ocean. Based on alteration resistant trace element and Sr–Nd–Pb–Hf isotopic compositions, the Pako lavas fall into three distinct groups. Group 1 has a distinctive HIMU composition with high (206Pb/204Pb)i ratios (20.41–20.94) similar to the Arago (also known as ‘Young Rurutu’ or ‘Atiu’) hotspot composition, suggesting derivation from a HIMU mantle source. Groups 2 and 3 EM1-like lavas comprise the dominant rock type and display the largest Sr–Nd–Pb–Hf isotopic variations that can be best explained by mixing between melts of focal zone (FOZO) and EM1-like mantle components. Additionally, olivine phenocrysts from Group 3 EM1-like lavas have high Ni contents, Fe/Mn and Mn/Zn ratios and low Zn/Fe*10 000 ratios, which can be explained through fractional crystallization of high-pressure partial melts from fertile peridotite. The EM1-like Groups 2 and 3 lavas are compositionally similar to the Rarotonga hotspot composition, suggesting that the MST is possibly an old trace of the Rarotonga hotspot, which is most likely a long-lived hotspot generated above a deep mantle plume. Combined with existing geochemical and tectonic data, we propose that the Magellan seamounts were likely derived from partial melting of a heterogeneous mantle plume containing HIMU, FOZO and EM1-like components. Alternatively, they could have been derived from the Arago and Rarotonga mantle plumes, each having its distinct compositional signature. The occurrence of HIMU- and EM1-like mantle plume components in a single volcano suggests that the superposition of compositionally different hotspot volcanic trails in the South Pacific could have played an important role in the generation of the WPSP.