Expedition NBP1808 on the R/V Nathan B. Palmer completed 32 dredges between October and December, 2018 from locations across the Rio Grande Rise (RGR)—a largely unstudied oceanic plateau on the South American plate—and several seamounts located between RGR and the Mid-Atlantic Ridge (MAR). Eighteen samples from 10 dredge locations on RGR were dated to better understand the geochronological history of this large igneous province and to provide clues to its relationship with the Walvis Ridge and Tristan-Gough hotspot(s) on the conjugate African plate. 40Ar/39Ar results from plagioclase separates (and one biotite) show a prolonged emplacement history throughout RGR ranging from ~84 to 48 Ma. Ages in general decrease towards the MAR in accord with plate motions showing that RGR as a whole was emplaced over at least several Ma and not as a single pulse like some other oceanic plateaus. Using the recently published tectonic reconstruction of Sager et al., most volcanism in the NW and NE sectors on RGR was emplaced off-axis while that in the SE sector was erupted on-axis. This suggests that the plume source for RGR changed from more intraplate to more ridge-centered as the system evolved through time. There is evidence of a possible reversed age progression in the NE RGR which could provide evidence for micro-plate activity that has been suggested in this region, though more ages are needed to confirm this trend. Geochemistry studies are ongoing and will be used in the future to better understand the eruptive processes. Additional age analyses are also ongoing and will focus on the other dredge locations throughout RGR as well as the seamounts to complete the geochronological picture of the emplacement of RGR.

Most efforts to characterize the size and composition of the mantle that complements the continental crust have assumed that the mid-ocean ridge basalt (MORB) source is the incompatible-element depleted residue of continental crust extraction. The use of Nd isotopes to model this process led to the conclusion that the “depleted MORB reservoir” is confined to the upper ~30% of the mantle, leaving the lower mantle in a more “primitive” state. Here we use Nb/U and Ta/U to evaluate mass and composition of the mantle reservoir residual to continent extraction and find that it exceeds 67% of the total mantle. Thus the (Nb,Ta)/U-based mass balance conflicts with the ε(Nd)-based mass balance, and this invalidates the classical 3-reservoir silicate Earth model (continental crust, depleted mantle, primitive mantle). Including the combined MORB + ocean island basalt (OIB) sources in the ε(Nd)-based mass balance does not reconcile the conflict as it would require their average ε(Nd) to be ≤3.0, much lower than observed MORB+OIB ε(Nd) averages. We resolve this conflict by invoking an additional, “early-enriched reservoir” (EER), formed prior to extraction of significant continental crust, but now hidden or lost. This EER differs from EERs previously invoked by having no Nb-Ta anomaly. We suggest that it originated as an early mafic crust, which had unfractionated (Nb,Ta)/U but fractionated Sm/Nd ratios. The corresponding “early-depleted” reservoir (EDR) generated the present-day continental crust and the “residual mantle” MORB-OIB reservoir, which occupies at least 70% of the present-day mantle and is only moderately depleted in incompatible trace elements.

Most previous efforts to characterize the size and composition of the upper mantle, the source of mid-ocean ridge basalts (MORBs), have assumed that this MORB source is the residue of continental crust extraction. The use of Nd isotopes to model this process led to the near-consensus that the “depleted MORB reservoir” is more-or-less confined to the upper mantle (above 670 km, ~30% of the mantle), with a severe degree of depletion of incompatible elements, leaving the lower mantle in a more primitive state. Here, we reassess the mass and composition of the mantle reservoir depleted by continental crust extraction. We initially apply simple mass balance considerations, using alternatively ε(Nd) and “canonical” (Nb,Ta)/U tracers, to a conventional three-reservoir silicate Earth consisting of primitive mantle, continental crust, and depleted mantle. The (Nb,Ta)/U tracer yields a ‘depleted reservoir’ exceeding 60% by mass of the total mantle (X(DM) > 0.6) with average ε(Nd) ≤ 3, whereas the ε(Nd)-based mass balance, using ε(Nd) = 8.5, yields a “depleted reservoir” of X(DM) ≤ 0.3. This discrepancy requires additional processes/reservoirs that impact the fractionation of Sm/Nd in the depleted mantle. Simple segregation of enriched OIB sources is shown to be insufficient. Permanent sequestration of a fourth, early-enriched, mafic reservoir (EER), leaving behind an early-depleted reservoir (EDR) can resolve the dilemma. Segregation of the present-day continental crust from EDR generates a moderately depleted, “residual-mantle” reservoir (RM), which occupies 80-98% of the total mantle (X(RM) = 0.8-0.98). This leads to concordant results for the two crust-mantle mass balances.