Petrogenetic Significance of Isotope and Trace Element Variations in Volcanic Rocks From the Mid-Atlantic

• Published in: Journal of Petrology
• Main Authors: O'NIONS, R. K., PANKHURST, R. J.
• Format: Text
• Language: English
• Published: Oxford University Press 1974
• Subjects: Articles; Jan Mayen; Reykjanes; Tristan; Kolbeinsey; Bouvetøya; Single Island; Iceland; Ocean Island
• Online Access: http://petrology.oxfordjournals.org/cgi/content/short/15/3/603; https://doi.org/10.1093/petrology/15.3.603

High precision 87Sr/86Sr analyses, together with determinations of Rb, Sr, K 2 O, Na 2 O and, in a few cases, other elements, are reported for about fifty volcanic rocks (mainly basaltic) from the Atlantic Ocean basin. Results for dredged basalts from the Reykjanes Ridge and Charlie Gibbs Fracture Zone, and an enstatite-forsterite basalt from Kolbeinsey islet, support the general observation that ocean-ridge tholeiites have uniformly low 87Sr/86Sr ratios (0.70294±4) and lithophile element contents compared with the most primitive basalts on ocean islands, including the Neovolcanic zones of Iceland, although progressive decrease in these quantities away from Iceland has not been confirmed. In contrast, the ocean island alkali basalts generally have higher 87Sr/86Sr ratios (0.70334±5 for the Snaefellsnes peninsula of Iceland, 0.70343±4 for Jan Mayen, 0.70509±4 for Tristan da Cunha and 0.70369±3 for Bouvetøya). The chief exception is Ascension Island, where volcanic rocks ranging from alkali-olivine basalt to trachyte give a mean value of 0.70284±4. The constancy of this ratio throughout eruptive sequences on any single island indicates that Sr-isotope characteristics are primary features. These variations, which are far outside analytical errors, are considered in the light of the geochemistry and isotope systematics of ocean basalts in general. The implied isotopic (and lithophile element) heterogeneities of the source regions have to be interpreted according to either equilibrium or disequilibrium melting models. The former, which is normally assumed, requires large-scale (domain) isotopic inhomogeneities within the mantle, which must have existed over thousands of m.y. unless the Rb/Sr ratio of extracted liquids is lower than that of the bulk source (as would be the case if phlogopite were a residual phase). In the case of disequilibrium melting, the inhomogeneities are reduced to the mineral scale, as observed in some studies of ultramafic nodules. It is shown that disequilibrium melting models could generally ...