Extended Data Fig. 3: Summary of new and previously published isotopic data for the Rurutu-Arago hotspot track.
From: Pacific hotspots reveal a Louisville–Ontong Java Nui tectonic link

Non-age-corrected plots of A) 207Pb/204Pb vs. 206Pb/204Pb, B) 143Nd/144Nd vs. 206Pb/204Pb, C) 208Pb/204Pb vs. 206Pb/204Pb, D) 143Nd/144Nd vs. 87Sr/86Sr, E) 143Nd/144Nd vs. 176Hf/177Hf, F) 206Pb/204Pb vs. 87Sr/86Sr, and G) 87Sr/86Sr vs. 176Hf/177Hf of our new Wake seamount samples (hexagons) compared to published data for the Tuvalu, Gilbert Ridge, Marshall/Ratak, and Wake Seamount portions of the Rurutu-Arago track (circles). Background data for young segments of the Cook-Austral plumes (Macdonald, Rurutu-Arago, Rarotonga), the Samoan plume, and Pacific MORB are given as 2σ contours of kernel density estimates (KDEs) of their respective datasets. Rarotonga lacks enough overlapping Sr, Pb, and Hf isotope data to be shown in panels E and G. Jurassic Pacific MORB data with seawater U alteration92 is also shown as small black open circles as a reference for how Pb isotopes may be disrupted in old seafloor basalts – and which is not evident in our new Wake samples (panels A-D). In both plots, the new Wake data plots within the known compositional range for the Wake Seamounts and the greater extent of the track. One of the new samples (outlined in cyan in panels A-C, F) has unusually high 208Pb/204Pb; this is a signature occasionally expressed in older Rurutu-Arago lavas and persists after strong leaching14. Here, 208Pb/204Pb is positively correlated with Th/U (see Extended Data Fig. 10), indicating U loss during alteration. Data are not corrected for post-eruptive radiogenic ingrowth (see Methods). We also identify a clear phosphatization signature (high Y/Y*; see Supplementary Data Table 1) in another of the new Wake samples (outlined in red in panels B, D, and G). As discussed in detail in the Supplement, strongly leached samples that have undergone significant phosphatization tend to have highly radiogenic 143Nd/144Nd of unclear origin. In such samples, Nd isotopes decouple from Sr isotopes (panel D) and Hf isotopes (panel E); obscuring some source mantle information. Here, 87Sr/86Sr vs. 176Hf/177Hf provide a more useful “isotopic fingerprint” of the HIMU-to-FOZO provenance characteristic of other Rurutu-Arago seamounts than plots using 143Nd/144Nd, and are consistent with interpretations from the Pb isotopes.