Age of the Ponta Grossa dike swarm (Brazil), and implications to Parana´flood volcanism

The Ponta Grossa Dike Swarm (PGDS) occurs in a NW-trending, 200 km wide zone exposed just east of the Parana´basin in southeastern Brazil. The predominantly basaltic dikes intrude crystalline basement, Paleozoic-Mesozoic sediments, and (rarely) flows of the Parana´-Etendeka flood volcanic province (PEP). The PGDS resembles the failed arm of a rift-rift-rift triple junction, related to the separation of South America and Africa. Detailed geochemical studies of the dikes (including major/minor/trace element and Sr-Nd isotopic analyses) indicate that they probably represent feeders for the voluminous phase of flood volcanism, represented by relatively uncontaminated, predominantly high-TiO2 lavas of the northern PEP, where lava accumulations reach 1700 m thick.40Ar/39Ar stepwise degassing data, using both laser and radiofrequency induction furnace, on plagioclase separates from eighteen dikes and one sill yield seventeen plateau ages: three are between 120.7 ± 1.3 Ma and 125.8 ± 0.6 Ma, and fourteen are clustered between 129.2 ± 0.4 Ma and 131.4 ± 0.5 Ma. Isochron ages are not significantly different from the plateau ages, and plateau ages are adopted in all but two cases as being the most accurate age estimates. The age-probability distribution for the dominant pulse (131.4 ± 0.4 to 129.2 ± 0.4 Ma) shows a pronounced peak at 130.5 Ma; this distribution probably reflects the magma production history in the region. The new geochronologic data are consistent with conclusions based on paleomagnetic and chemical-stratigraphic data that the PGDS is younger than the volumetrically dominant majority of volcanism in the southern PEP, which occurred at 133-132 Ma. The younger (commonly NE-trending) dikes may reflect the initiation of full drift, which was coincident with major basin development in the adjacent continental borderland at 125-120 Ma. The PGDS may indeed represent the failed arm of a rift-rift-rift triple junction, but the triple junction did not coincide exactly in time or space with the site that would be inferred for plume impact.