Abstract

Thermally buoyant mantle, in the form of a plume, rises beneath Iceland creating a major topographic anomaly along the Mid-Atlantic Ridge and in the surrounding ocean basin. However, the influence of the Iceland plume on the composition of lavas erupted on adjacent ridges remains a contentious issue. Trace element systematics and radiogenic isotope ratios of Sr, Nd and Pb suggest that the plume influences a region 1200km in length. In contrast, the 3He anomaly associated with Iceland closely corresponds to the 2400km ridge section affected by thermal uplift. We present evidence that the Sr, Nd and Pb isotope signature of the Iceland plume is in fact as widespread as its thermal and ³He anomalies. Results imply that much of the source of North Atlantic ridge basalts has been contaminated by lateral outflow of asthenosphere from the Icelandic plume. Consequently, estimates of the average composition of mid-ocean ridge basalt (MORB) sources are likely to be biased by including data from plume-contaminated regions. True MORB values and perhaps upper mantle geochemistry can be constrained only by considering data untainted by plume asthenosphere.

Figure 1. Variation of radiogenic isotopes, trace elements and helium with distance from the centre of the Iceland plume. Vertically shaded area marks the extent of subaerial Iceland. Horizontally shaded area for each parameter represents the range of values found in MORB considered to be unaffected by plumes. Data in the section 180-250km and 400-1200km south of the plume is from this study. Other Iceland and ridge data from literature.

Figure 2. 87Sr/86Sr versus 206Pb/204Pb for North Atlantic oceanic volcanism. Reykjanes Ridge and Reykjanes Peninsula data from this study, other data as for Fig.1 and 24,25,26. Plume-free Atlantic MORB field distinguishes 40 samples from 1.5^oS-7^oS, 11^oN-12^oN,17^oN-30^oN and 50oN. Inset graph shows potential mixing lines between plume-free Atlantic MORB and three Icelandic compositions, calculated assuming Sr/Pb(MORB)=Sr/Pb(Iceland)=250.

Figure 3. ¹⁴³Nd/¹⁴⁴Nd versus ⁸⁷Sr/⁸⁶Sr for northern and central Atlantic Ridges and ocean islands. The plume-free Atlantic MORB field distinguishes 28 samples from 1.5^oS-7^oS, 11^oN-12^oN,17^oN-30^oN and 50^oN. DM represents plume-free Atlantic MORB

. Figure 4. (a) 143Nd/144Nd versus 87Sr/86Sr for Reykjanes Ridge basalts grouped according to distance from Iceland as in Fig.2. Enriched Iceland field encloses 50 samples with Zr/Y>2.2, Depleted Iceland field encloses 30 samples with Zr/Y<2.2. (b) 143Nd/144Nd - 87Sr/86Sr mixing models. Three Icelandic lava compositions are used in the modelling: eI is an average enriched Icelandic lava, dI an average depleted Icelandic lava, and BIR is a high Sr/Nd depleted Icelandic lava (International rock standard BIR-1 from the Reykjanes Peninsula, Table 1). These are mixed with the plume-free Atlantic MORB (DM) from table 1 (Sr/Nd~10). Although these components are lavas, it is assumed that the mixing calculations would work equally well for mantle sources, provided that the average degree of melting, and the relevant bulk partition coefficients are broadly similar for the mixing components.