Mid-Cretaceous Wake seamounts in NW Pacific originate from secondary mantle plumes with Arago hotspot composition

Highlights

• Comprehensive dataset of lavas from southern Wake seamount trail (WST).

• Lavas formed from melting of carbonated peridotite or reaction between carbonated eclogite-derived melts and peridotite.

• Lavas show FOZO-HIMU Sr-Nd-Pb-Hf isotopic compositions similar to Arago hotspot.

• WST originated from partial melting of secondary plumes similar to Arago composition.

Abstract

The geochemistry of oceanic intraplate (primarily oceanic island and seamount) lavas can provide essential information on the composition and evolution of their mantle source and geodynamics. Due to very limited rock sampling of the many mid-Cretaceous Wake seamounts in NW Pacific, the mantle source lithology and petrogenesis of their lavas, and the geodynamic mechanism responsible for generating the lavas have not been fully delineated. In order to help resolve these issues, here we present whole-rock major-trace element and Sr-Nd-Pb-Hf isotopic data for twenty-one lava samples collected from Lamont, Dacheng, Xufu, Penglai, Niulang, Zhinyu, and Zhanlu seamounts in the Southern Wake seamount trail (WST). These lava samples are silica-undersaturated alkali basalt and basanites/nephelinite. They have high CaO, FeO^T and TiO₂ contents and CaO/Al₂O₃ ratios, consistent with their derivation from partial melting of a carbonated peridotite or reaction between carbonated MORB-eclogite-derived silicate melts and fertile peridotite. High Zr/Hf and negative Zr-Hf-Ti anomalies in the most mafic lavas further suggest a contribution from carbonated components in their mantle source. These lavas show FOZO (focal zone)-HIMU (high μ = ²³⁸U/²⁰⁴Pb)-like Sr-Nd-Pb-Hf isotopic compositions (e.g., (²⁰⁶Pb/²⁰⁴Pb)_i = 19.36–20.72), falling within the Arago (also known as “Young Rurutu” or “Atiu”) hotspot field. Combined with the sparse previous age and geochemical data, we propose that WST lavas were most likely derived from partial melting of secondary plume clusters emanating from the top of Arago mantle plume trapped at the mantle transition zone. Alternatively, WST lavas could have come from a number of secondary plumelets emanating from the top of the Pacific Large Low Shear Velocity Province (LLSVP). The simultaneously upwelling secondary plumes or plumelets generated chronologically overlapping, compositionally similar and closely-spaced Wake seamounts atop the moving Pacific plate.

Introduction

The great concentration of volcanic seamounts in the West Pacific Seamount Province (WPSP), NW Pacific, is one of the most striking features on the Earth's surface (Smith et al., 1989), but its origin is enigmatic. The mid-Cretaceous (~81–120 Ma) Wake seamount trails (WST) in the WPSP consist of the Northern and Southern WST (Fig. 1; Ozima et al., 1977; Smith et al., 1989; Winterer et al., 1993; Koppers et al., 2003b). The Northern WST was sampled at Alcatraz and Scripps whereas the Southern WST was sampled at Himu, Golden Dragon, Missy, Jennings, Maloney, Lamont, Xufu, Zhanlu, Dacheng, and Miami (Fig. 1b; Koppers et al., 2003b; Yan et al., 2021). The seamounts in the Southern WST do not form a narrow, linear and continuous chain of volcanoes; rather they are characterized by complex and irregular age–distance relationships as they exhibit near-synchronous volcanisms along its trail (e.g., ~96.7 Ma - Missy; ~97.7–100.8 Ma - Maloney; ~96.8 Ma - Miami; Fig. 1b; Koppers et al., 2003b). Such features cannot be explained solely by the classic mantle plume hypothesis which predicts that a linear and clearly age-progressive chain of volcanoes in the direction of plate motion originates above the narrow plume tail (Morgan, 1981; Richards et al., 1989). Competing alternative models proposed that such scattered volcanoes without geographic age progressions can be accounted for by shallow processes, for example as a response to lithospheric cracking (McNutt et al., 1997), small-scale sublithospheric convection induced melting (Ballmer et al., 2010), or melting of fusible components in the upper mantle (Hoernle et al., 2011; Yan et al., 2021).

The WST lavas exhibit a HIMU- to FOZO-like composition (Smith et al., 1989; Staudigel et al., 1991; Koppers et al., 2003b; Konter et al., 2008) which is similar to the Arago hotspot (Konter et al., 2008; Finlayson et al., 2018). The WST can be geographically and geochemically backtracked to the Tertiary Cook-Austral volcanic lineament in South Pacific, indicating that volcanoes in WST and Cook-Austral volcanic lineament most likely share a common mantle source (Koppers et al., 2003b; Konter et al., 2008). The WST, as well as other seamount groups in WPSP and Cook-Austral volcanic lineament, were suggested to be formed from partial melting of secondary plumelets emanating from the top of the Pacific Large Low Shear Velocity Province (LLSVP) (Koppers et al., 2003b). The plumelet hypothesis (typical timescale of ∼30 Ma), however, is not able to explain the longevity (>100 Myrs) of the geochemical anomalies and highly complex and irregular age–distance relationships of volcanoes in WPSP and Cook-Austral volcanic lineament (Ballmer et al., 2010). Instead, Ballmer et al. (2010) proposed that the volcanoes in WPSP and Cook-Austral volcanic lineament can be explained by partial melting of pyroxenite and enriched components in the upper mantle induced by small-scale sublithospheric convection (SSC). More and more evidences, however, demonstrate that the Cook-Austral volcanic lineament is composed of multiple overlapping, age-progressive hotspot tracks (e.g., Chauvel et al., 1997) that are generated by at least two long-lived (at least >70 Ma) primary mantle plumes: Arago and Macdonald (e.g., Konter et al., 2008; Finlayson et al., 2018; Konrad et al., 2018b; Jackson et al., 2020; Buff et al., 2021). Based on their geochemical similarity, the WST lavas were suggested to be formed from partial melting of the long-lived (> 100 Ma) Arago mantle plume (Konter et al., 2008). This model, however, does not explain why the WST volcanoes do not show a clear age progression and/or why near-synchronous volcanisms occur in several locations along the Southern WST. Due to limited rock sampling, the compositional range and mantle source lithology for WST lavas which are critical to understand their petrogenesis and geodynamic model responsible for generating WST have not been fully delineated.

In this study, we attempt to address the above questions by analyzing whole rock major-trace element and Sr-Nd-Pb-Hf isotopic data for twenty-one lava samples collected from seven seamounts, i.e., Lamont, Dacheng, Xufu, Penglai, Niulang, Zhinyu, and Zhanlu, in the Southern WST during the Chinese R /V Dayang Yihao cruises DY105–11 in July 2001, DY105–12/14 in April 2003, DY105–15 in July–August 2003 and DY105-16A in August 2004 (COMRA, 2017). We evaluate the mantle source lithology using major element compositions for these lavas in detail. We observe large Sr-Nd-Hf-Pb isotope variations in these seamount lavas, almost covering the entire isotopic range of WST lavas and falling within the field defined by lavas from old portion of the Arago hotspot track including Tuvalu and Gilbert Ridge islands and seamounts (Konter et al., 2008; Finlayson et al., 2018). Finally, we propose that the WST lavas were most likely partial melts from secondary plume clusters emanating from the top of Arago mantle plume trapped at the mantle transition zone or plumelets emanating from the top of the Pacific LLSVP during the mid-Cretaceous.