Abstract

Intra-plate basalt volcanism derived from the largest mantle plumes appear to be characterized by 3He/4He that are typical higher than mid-ocean ridge basalts (MORB). There is little argument that this reflects the contribution from a deep mantle reservoir that has experienced less degassing of the primordial noble gases (and therefore all volatiles) than the convecting upper mantle. A convectively-isolated lower mantle appears to satisfy several global geochemical observations and was the favoured explanation for many years. However this difficult to reconcile with geophysical observations and alternatives models have been proposed, including isolated blobs in a whole convecting mantle, a discrete core-mantle boundary layer and, most recently, a convectively slow deep mantle. In these models high 3He/4He is associated with a geochemically-distinct mantle reservoir that is less depleted than the convecting upper mantle. We have previously identified the highest terrestrial 3He/4He (up to 50 Ra) in the earliest basalts erupted by Iceland plume. Contrary to expectation these picrites exhibit lithophile radiogenic isotope and trace element compositions that span nearly the complete range of global MORB and OIB . This is in strong contrast to the expectation that the high 3He/4He basalts are derived from a primordial gas-rich, compositionally unique mantle reservoir. Noble gas (Ne, Ar and Xe) isotope compositions cover the range of values that are typical of MORB and OIB , consistent with the radiogenic isotope and trace element data but strongly decoupled from He isotopes. 3He/21Ne and 3He/36Ar are 1-2 orders of magnitude higher than measured in MORB and OIB indicating that the initial stages of the plume sampled mantle that was enriched in high 3He/4He helium. In this presentation I will review the possible mechanisms and implications of these observations.