Home | Online Resources | Table of Contents |
Journal of Petrology, Volume 39, Issue 8: August 1 1998.
The Serra do Mar province extends for ~500 km as plutonic complexes, some with volcanic remnants, and dyke swarms along coastal SE Brazil. The igneous rocks are emplaced into metamorphics of the Brasiliano (Pan-African) Ribeira mobile belt, which abuts the southern margin of the Archaean São Francisco craton. Syenites and phonolites dominate most complexes but mafic-ultramafic dykes (MgO <16·4%) are also widespread, together with sparse mafic lavas. The mafic rocks are alkali basalts, basanites and melanephelinites, with rare leucitites and minettes (s.l.). Major and trace elements of all but a few of the strongly potassic rock types are indistinguishable from ocean-island basaltic (OIB, s.l.) magmatism. Isotopically (Sr-Nd), the Serra do Mar suite overlaps the field for OIB ([epsilon]Nd up to +2·7) and aseismic ridges, especially Walvis. In general, the most potassic compositions have the lowest [epsilon]Nd values (-1·4 to -3·3). We consider the most plausible source for Serra do Mar mafic magmas to be predominantly sub-lithospheric convecting mantle, with sporadic input from lithospheric metasomites. Additionally, the two most Mg-rich dykes have high 87Sr/86Sri (>0·707), despite <[epsilon]Nd in the same range as nearby mafic dykes. We attribute this to upper-crust contamination during turbulent emplacement of very hot, low-viscosity magma batches. New K/Ar dates, together with a detailed assessment of published ages, show that the initiation of Serra do Mar magmatism migrated progressively ~500 km ESE, from Poços de Caldas to Cabo Frio, between ~80 and 55 Ma. It is inferred that the Trindade mantle plume impacted at ~85 Ma beneath interior SE Brazil where the lithosphere was too thick (>150 km), except very locally, to permit dry decompression melting and basaltic (s.l.) magma genesis. Instead, strongly alkalic, potassic, lithosphere-source magmatism (including diamondiferous kimberlites) broke out penecontemporaneously over a region of ~600 km radius. Between ~85 and 80 Ma it is suggested that the starting-plume head expanded rapidly beneath the thick continental lithospheric lid. This caused regional kilometre-scale uplift and erosion (evidenced by published apatite fission-track data) for 1200 km to the south, together with strongly alkaline magmatism at Lages, southern Brazil, a site of localized rifting. After ~80 Ma, as the plume tail moved ESE beneath the Archaean São Francisco craton, its magmatic signal at the surface was `switched off' by the thick (200-250 km) cratonic lithosphere. Nevertheless, hot plume-tail mantle `leaked' progressively southward, until it encountered strongly thinned lithosphere beneath SE coastal Brazil, along the passive margin of the South American plate, where decompression melting generated the predominant OIB-like component of the Serra do Mar magmatism. After 55 Ma this magmatism transferred NE from Cabo Frio to the Abrolhos Platform (52 Ma), the landward terminus of the seamount chain leading to the islands of Trindade and Martin Vaz in the South Atlantic.
Keywords:
Pages 1493-1526
Please note that full-text access is only available to current subscribers [Registration & Subscription Info] | ||
Full-text HTML (150 KB) | Full-text PDF (664 KB) |