The Rio Grande Rise (RGR) in the South Atlantic is an oceanic plateau that appears to have been formed together with the Walvis Ridge (WR) although both are now residing on two different tectonic plates, namely the South American and African Plates. For a long time this was based on meager evidence, because RGR had been greatly understudied. Major uncertainty therefore existed in understanding how these two volcanic provinces could have formed together, when, for how long, and in what tectonic configuration. It was also unclear if the primary Tristan-Gough hotspot system, widely considered to be consistent with the existence of a thermo-chemical mantle plume, did indeed originate along one of the edges of the African Large Low Shear Velocity Province (LLSVP) in the deepest regions of the Earth's mantle. This potential relationship to the edge of the African LLSVP makes the RGR-WR system a globally unique, volumetrically significant, and 130 Myr lasting window into the core-mantle boundary region.

Because RGR had been sampled very sparsely, many critical questions remained: Did Rio Grande Rise Form Along a Plate Boundary and Micro-plate? Is Rio Grande Rise Age-progressive and What Causes its High Volcanic Output? Did Rio Grande Rise Seamounts Form in Relation to the Tristan-Gough Plume? Can Rio Grande Rise Add to our Understanding of Plume Zonation? To answer those questions we have been testing the following hypotheses: Rio Grande Rise is Age Progressive Throughout the Plateau Itself; Rio Grande Rise and Walvis Ridge formed at a Micro-plate; Late-Stage, Diffuse Volcanism occurs after ~70 Ma in Two Seamount Groups Equivalent to the Guyot Province in the WR; and Rio Grande Rise is a Zoned-Plume Product from the Edge of an LLSVP.

To address these questions and hypotheses we collected submarine basaltic samples from RGR in 32 dredge hauls, of which 27 contained materials potentially suitable for dating with the 40Ar/39Ar geochronology technique. In total we carried out 51 analyses from 46 samples. Our new ages demonstrate that RGR was emplaced over a prolonged period, longer than 30 million years, which is in stark contrast to the conventional understanding that large igneous provinces are rapidly emplaced, for example the Deccan and Siberian Traps which both were formed in less than one million years. The oldest RGR ages are 81.8 and 84.2 million years old. The youngest RGR ages 41.3 to 51.8 million years old. Reconstructions of the combined RGR and WR age data sets shows that with the exception of some late-stage volcanism that occurred off the ridge axis (e.g. RGR Eocene volcanism), volcanism north of the Peleg Fracture Zone lasted from ~120 until ~70 million years ago and was largely centered along the Mid Atlantic Ridge (MAR). In this time period, our reconstructions also show that some discrepancy in the reconstructions could be solved by the presence of a temporary microplate that was later incorporated into the South American plate. The less voluminous SE-RGR massif was formed between 70 and 55 million years ago, during a time period when the Tristan-Gough system was transitioning from being ridge-centered to off axis or intraplate. After 55 million years ago, the volcanism was away from the MAR and located beneath the African plate only.

We estimate the entirety of RGR has an erupted volume of ~0.6 million km3 similar to the Deccan Traps but an order of magnitude less than the Ontong Java Plateau. We also calculate that the SE-RGR formed at lower rates of about 0.03 million km3, further suggesting that the Tristan-Gough hotspot was waning in strength even before the plume had become entirely intraplate. Production rates fall lower still in the WR Guyot Province after the plume moved to an intraplate setting. This makes WR-RGR combined region similar to the present analog of Iceland, which is a modern example of a mantle plume centered at (or near) a spreading ridge axis, but completely unlike the voluminous flood basalts erupted on land.

In conclusion, our new ages expand the known influence of the long-lived Tristan-Gough mantle plume on the South American plate back to at least 85 Ma and possibly further. The Tristan-Gough plume produced widespread volcanism centered under the Mid-Atlantic Ridge until ~55 Ma. The plume moved south between ~75 and 45 Ma after which its position stabilized off axis on the African plate. This southward off axis movement corresponds to a decrease in the spatial extent of volcanism and the creation of the Tristan geochemical subtrack and could reflect significant changes at the LLSVP edges.

Last Modified: 06/28/2024
Modified by: Anthony Koppers