Jasper G. Konter, Valerie Finlayson , Kevin Konrad, Matthew Gerard Jackson, Anthony A.P. Koppers, Paul Wessel, Michael Bizimis, Alexander Alverson, Christopher Kelley

Volcanic hotspots are thought to initially form by melting in an upwelling mantle plume head followed by melting of the plume tail. Plate motion then generates an age progressive volcanic track originating from a large igneous province that connects to an active hotspot. However, the most voluminous large igneous province, the ~120 Ma Ontong-Java Nui Plateau (OJP-Nui) in the mid-Pacific, appears to lack such a volcanic track. Although the Louisville hotspot track was originally proposed as a candidate, limited constraints for Pacific absolute plate and plume motion prior to 80 Ma suggest a mismatch[1]. Existing Pacific models rely on age-distance data from the continuous Hawaii-Emperor and Louisville volcanic tracks, but their seamounts older than ~80 Ma are now subducted, and elsewhere on the Pacific plate only discontinuous and sparse seamount tracks can be found that formed prior to 80 Ma[2–7]. These existing models require ~1,200 km of latitudinal motion for the Louisville plume to also erupt the OJP-Nui1, yet paleolatitude estimates from to ~70 Ma to today remain within error of its present location[8,9] and suggest that any major amount of Louisville plume motion should precede that time. Here we provide evidence from geochemistry and eruption ages[9–14] demonstrating that Samoa and Rurutu-Arago are the longest-lived Pacific hotspots that can be traced back to ~120 Ma (and older) in the West Pacific where they subduct into the Mariana Trench. These newly defined tracks provide for an alternative Pacific absolute plate motion model, with better constraints for a plate rotation between 80-100 Ma, and allow us to establish Louisville as the missing volcanic track for OJP-Nui without requiring major plume motion.