Does depleted mantle form an intrinsic part of the Iceland plume?

Abstract

Icelandic basalt ranges in composition from voluminous tholeiite, erupted in the rift zones, to small‐volume, mildly alkaline basalt erupted off‐axis. In addition, small‐volume flows of primitive basalt, highly depleted in incompatible elements, are sometimes found in the actively spreading rift axes. Relative incompatible‐element depletion or enrichment in Icelandic basalt is correlated with variation in radiogenic isotope ratios, implying that the mantle beneath Iceland is heterogeneous and that the relative contribution of the various mantle components relates to eruption environment (on‐ or off‐axis) and hence to degree of melting. Thus small‐degree off‐axis melting preferentially samples an enriched and more fusible mantle component, whereas more extensive melting beneath the rift axes produces magma that more closely represents the bulk Iceland plume mantle composition. The small‐volume flows of depleted basalt represent melts that have preferentially sampled a depleted and more refractory mantle component. A debate has arisen over the nature of the depleted component in the Iceland plume. Some authors [e.g., Hanan and Schilling, 1997] argue that the depleted component is ambient upper mantle, the source of normal mid‐ocean ridge basalt (NMORB) in this region. Others [e.g., Thirlwall, 1995; Kerr et al., 1995; Fitton et al., 1997], however, have used various lines of evidence to suggest that the plume contains an intrinsic depleted component that is distinct from the NMORB source. Hanan et al. [2000] attempt to refute the existence of a depleted Iceland plume (DIP) component through a critical evaluation of the Nb‐Zr‐Y arguments advanced by Fitton et al. [1997] and the Hf‐Nd‐isotopic evidence presented by Kempton et al. [1998]. In this paper we examine the case presented by Hanan et al. [2000] and conclude that their arguments are flawed. Firstly, their trace‐element data set excludes data from depleted Icelandic basalt samples and so it is not surprising that they find no evidence for a DIP component. Secondly, they present two new Hf‐isotope analyses of a single depleted Icelandic basalt sample and show that the data plot in their NMORB field on an εHf versus εNd diagram. However, new data allow the resolution of distinct NMORB and depleted Icelandic basalt fields on this diagram. We conclude that trace‐element and radiogenic isotope data from Iceland require the existence of a DIP component.