The limited extent of plume-lithosphere interactions during continental flood-basalt genesis: geochemical evidence from Cretaceous magmatism in southern Brazil

Abstract

It has been suggested that large areas of the Earth's lithospheric mantle undergo pervasive dehydration melting during the impact of mantle plumes and the Early-Cretaceous Parana-Etendeka continental flood-basalt (CFB) province has repeatedly been cited as evidence of this phenomenon. During the Cretaceous, however, southern Brazil experienced two phases of mafic magmatism. These igneous events occurred ∼50 Ma apart and therefore represent distinct episodes of melt genesis in the underlying mantle. The first phase of magmatism, in the Early Cretaceous, included the emplacement of lava flows associated with the Parana-Etendeka CFB province and also the intrusion of small-volume mafic alkaline magmas (e.g. Anitapolis, Jacupiranga and Juquia) in the Dom Feliciano and Ribeira mobile belts. During the Late Cretaceous, both sodic and potassic mafic magmas were emplaced on the margin of the adjacent Luis-Alves craton and intrude the flood-basalts at Lages. On the basis of variations in incompatible trace-element concentrations (e.g. Ba = 1000 to 2000 ppm), initial 87Sr/86Sr ratios (0.7048–0.7064) and ɛNd values (−3 to −12), we suggest that all of the Late-Cretaceous mafic potassic magmas were derived from the subcontinental lithospheric mantle (SCLM) which was metasomatically enriched during the Proterozoic. We propose that these relatively low temperature, volatile-rich, mafic melts provide direct evidence that the underlying SCLM did not melt wholesale during the previous Early-Cretaceous Parana-Etendeka CFB event. Late-Cretaceous melting of the SCLM beneath southern Brazil may have been caused by heat conduction from either: (1) ponded ∼132 Ma Tristan plume-head material; or (2) ∼85 Ma Trindade plume-head material channelled southwards between the thick cratonic keels of the Amazonas and Sao Francisco cratons. The Late-Cretaceous magmatism appears to have been contemporaneous with uplift across southern Brazil and Paraguay; we suggest that both of these phenomena represent the widespread effects of the impact of the Trindade mantle plume on the base of the SCLM. Plate margin stresses and lithospheric extension associated with the opening of the South Atlantic may also have changed the geothermal gradient beneath southern Brazil and contributed to mantle melting.