Short-lived and discontinuous intraplate volcanism in the South Pacific: Hot spots or extensional volcanism?

South Pacific intraplate volcanoes have been active since the Early Cretaceous. Their HIMU-EMI-EMII mantle sources can be traced back into the West Pacific Seamount Province (WPSP) using plate tectonic reconstructions, implying that these distinctive components are enduring features within the Earth's mantle for, at least, the last 120 Myr. These correlations are eminent on the scale of the WPSP and the South Pacific Thermal and Isotopic Anomaly (SOPITA), but the evolution of single hot spots emerges notably more complicated. Hot spots in the WPSP and SOPITA mantle regions typically display intermittent volcanic activity, longevities shorter than 40 Myr, superposition of hot spot volcanism, and motion relative to other hot spots. In this review, we use ⁴⁰Ar/³⁹Ar seamount ages and Sr-Nd-Pb isotopic signatures to map out Cretaceous volcanism in the WPSP and to characterize its evolution with respect to the currently active hot spots in the SOPITA region. Our plate tectonic reconstructions indicate cessation of volcanism during the Cretaceous for the Typhoon and Japanese hot spots; whereas the currently active Samoan, Society, Pitcairn and Marquesas hot spots lack long-lived counterparts in the WPSP. These hot spots may have become active during the last 20 Myr only. The other WPSP seamount trails can be only "indirectly" reconciled with hot spots in the SOPITA region. Complex age distributions in the Magellan, Anewetak, Ralik and Ratak seamount trails would necessitate the superposition of multiple volcanic trails generated by the Macdonald, Rurutu and Rarotonga hot spots during the Cretaceous; whereas HIMU-type seamounts in the Southern Wake seamount trail would require 350-500 km of hot spot motion over the last 100 Myr following its origination along the Mangaia-Rurutu "hotline" in the Cook-Austral Islands. These observations, however, violate all assumptions of the classical Wilson-Morgan hot spot hypothesis, indicating that long-lived, deep and fixed mantle plumes cannot explain the intraplate volcanism of the South Pacific region. We argue that the observed short-lived and discontinuous intraplate volcanism has been produced by another type of hot spot-related volcanism, as opposed to the strong and continuous Hawaiian-type hot spots. Our results also indicate that other geological processes (plate tension, hotlines, faulting, wetspots, self-propagating volcanoes) may act in conjunction with hot spot volcanism in the South Pacific. In all these scenarios, intraplate volcanism has to be controlled by "broad-scale" events giving rise to multiple closely-spaced mantle plumelets, each with a distinct isotopic signature, but only briefly active and stable over geological time. It seems most likely that these plumelets originate and dissipate at very shallow mantle depths, where they may shoot off as thin plumes from the top of a "superplume" that is present in the South Pacific mantle. The absence of clear age progressions in most seamount trails and periodic flare-ups of massive intraplate volcanism in the South Pacific (such as the one in the Cretaceous and one starting 30 Myr ago) show that regional extension (caused by changes in the global plate circuit and/or the rise-and-fall of an oscillating superplume) may be driving the waxing and waning of intraplate volcanism in the South Pacific.