The along-strike pattern of magmatism during breakup in the southern South Atlantic: The relative roles of deep mantle plume and lithospheric segmentation

Whilst our understanding of rift formation and evolution has matured in recent years there remains a fundamental debate on the relative roles of deep processes, and in particular mantle plumes, compared to shallower processes in the lithosphere or uppermost asthenosphere. Current interest in this topic is partly driven by the increased exploration for hydrocarbons in deep- and ultra-deep-water. Mitigating the high risks associated with exploring in these settings requires an improved understanding of the geodynamics of passive margin formation. Here we investigate the pattern of magmatism in the South Atlantic between the Florianopolis and Falkland-Agulhas Fracture Zones. This region is recognized as a prime example of conjugate volcanic margins, with extensive seaward dipping reflector sequences (SDRs) and high velocity lower crust (HVLC), but it remains relatively under-studied and has a somewhat chequered exploration history. Recently, based on closely spaced, but shallow recording academic reflection data, it has been suggested that rather than a systematic increase in the volumes of the SDRs towards the ancestral Tristan da Cunha hot-spot there is an alternating pattern correlating with first order, -500-km long segments bounded by transfer zones which link to onshore continental structures that may have controlled the Cretaceous rifting. Here, we test this observation, which would imply a lithospheric rather than deep plume influence on melting, by adding observations from 18,000 km of long-offset commercial seismic reflection data that images to 40 km depth to the analysis. To explore the along-strike variations in magmatism we have also compiled observations of initial ocean crustal thickness as a measure of the melt production immediately post-breakup. We show that along both conjugate sides. there is a systematic decrease in oceanic crust production with distance away from an assumed ancestral Tristan hotspot impact point, averaging 1.4 km/1000 km. We explore the implications of these observations of melt volumes by comparing against results from a 20 numerical simulation of continental breakup. In these models we include independent estimates of initial lithospheric thicknesses and extension rates to isolate mantle temperature influence. The observed pattern can be matched with numerical models with a 200 km thick hot later beneath the lithosphere, where the excess temperature reduces from 200°C to 50°C over a distance of 1500 km, southwards away from the plume centre. These results suggest that a simple plume concept is appropriate for these margins to a first order, albeit with modification from plate thickness.