THE MANTLE LITHOSPHERE IN NORTH-EASTERN BRAZIL AND THE RECORDS OF THE FERNANDO DE NORONHA PLUME

Abstract

The Cenozoic alkali basalt centres in north-eastern Brazil are believed to represent the surface track of the passage of the craton over the Fernando de Noronha plume (O’Connor and Duncan, 1990). This hypothesis is at present largely unconstrained because the only one previously studied xenolith occurrence, Pico Cabuji, showed large isotopic differences with xenoliths from the Fernando de Noronha island (Rivalenti et al., 2000). We revisit this hypothesis on the basis of a larger geochemical and isotope data set on the xenoliths of unstudied Brazilian localities (Fig. 1). Texturally, the xenoliths define three group: 1) porphyroclastic (lherzolites, harzburgites and one wehrlite); 2) protogranular (lherzolites and harzburgites); 3) transitional between groups 1 and 2 (lherzolites and harzburgites). Equilibrium temperature and pressure ranges 950 – 1280°C, 1.9 – 2.5 GPa in group 1, 830 – 990°C, 1.3 – 1.8 GPa in group 2, 890 – 980°C, 1.5 – 2 GPa in group 3. . The three textural groups have a similar mg# range in clinopyroxene, but the porphyroclastic clinopyroxenes have, at any given mg# value, lower Ti, Al, Ca and Na and higher Fe concentrations (a.f.u.), with respect to the protogranular ones, and the cpx of the transitional xenoliths define an array intermediate between the others. Orthopyroxenes have opposite characteristics with respect to cpx, so that the porphyroclastic ones are the richest in Ti, Al, Ca, Na and the lowest in Fe. In group 1 the REE patterns of clinopyroxenes vary from convex upward, typical of phases in equilibrium with alkaline melts, to spoon shaped, to steadily LREE-enriched in wehrlite (Fig. 2). Group 2 clinopyroxene are predominately LREE – depleted and enriched only in Th and U and group 3 clinopyroxenes are similar to those of group 2, but have higher LREE concentrations and in some cases spoon-shaped profiles. Sr and Nd isotopes of the group 1 clinopyroxenes form an array between DM and an EMI-like component, present also in the host basalts, whereas the protogranular xenoliths exhibit large spreading of the Sr-isotopic values (Fig. 3). The samples with higher 87Sr/86Sr trend towards EMII. Potential melts in equilibrium with the group 1 clinopyroxenes are similar to the alkali basalts hosting the xenoliths and those estimated from protogranular clinopyroxenes are similar, except for a higher concentration of the highly incompatible elements Th and U, to Mesozoic tholeiitic basalts of north-eastern Brazil. These results are consistent with the following conclusions: a) the xenoliths sampled a mantle section whose deepest and shallowest regions are represented by the group 1 and 2 xenoliths, respectively, being those of group 3 derived from an intermediate region; b) the geothermal gradient of the section was as high as 15°C/Km; c) geochemical modelling supports that the deepest region underwent reactive porous flow of alkali basalt, isotopically closer to EMI, whereas metasomatism was controlled by percolation and crystallisation under decreasing melt mass in the regions represented by the group 3 and 2 xenoliths; d) lithosphere predating metasomatism is largely represented by the group 2 xenoliths, which isotopically plot in between DM and an EMII component, thus considered as lithospheric. Similarity with Mesozoic basalts of the group 2 and 3 potential melts calculated from clinopyroxene and high 87Sr/86Sr in clinopyroxene suggests that these xenoliths predominately record mantle processes predating the Cenozoic event. A comparison of the characteristics of north-eastern Brazil and Fernando de Noronha shows that the respective lithospheres record similar stratigraphy, interaction with an asthenospheric EMI component and also similar heterogeneity in the cooler lithosphere section documented by the protogranular xenoliths. This similarity lends support to the notion that the lithosphere beneath Fernando de Noronha is a detached piece of the continental one. None of the characteristics of north-eastern Brazil xenolith suite provides unquestionable evidence of plume interference. For example, the mantle stratigraphy documented by the textural, geochemical, isotopic and thermal variations and its high geothermal gradient, often considered a typical result of plume impingement on the lithosphere are also found in other mantle sections where processes are not controlled by plume. If the existence of the Fernando de Noronha plume is given as granted, the only evidence of its interaction with the north-eastern Brazil lithosphere is the similarity of isotope compositions and processes of the lithosphere of the two localities. The isotopic arrays of the porphyroclastic xenoliths in both cases indicate that a lithosphere having DM characteristics was enriched by an asthenospheric basaltic melt having EMI-like characteristics. If asthenosphere melting was triggered by the addition of plume material, then the enriched component may represent a plume characteristic.