Coherent He-Nd-Sr isotope trends in high He-3/He-4 basalts: implications for a common reservoir, mantle heterogeneity and convection

• Published in: Earth and Planetary Science Letters
• Main Authors: Ellam, R.M., Stuart, F.M.
• Format: Article in Journal/Newspaper
• Language: unknown
• Published: 2004
• Subjects: QE Geology; Baffin Island; Baffin; Iceland; Ocean Island
• Online Access: http://eprints.gla.ac.uk/926/; https://doi.org/10.1016/j.epsl.2004.10.020

Early-mid-Tertiary picritic basalts associated with the proto- Iceland plume (PIP) show correlated He-Nd and He-Sr isotope trends. Other intraplate basalts with olivine phenocryst He- 3/He-4 gt 10 R-a either (1) fall on the PIP trends, (2) lie in triangular fields defined by the PIP trend and a depleted mantle field or (3) are displaced to low, and high Sr-87/Sr-86, indicating the involvement of an ancient enriched component, but are consistent with a high He-3/He-4 end-member similar to that required by the Baffin Island data. Neither end of the PIP trend represents isotope compositions that have previously been recognised as global geochemical components. A concentration of data around Nd-143/(144)Ndapproximate to0.5128, with low He- 3/He-4 suggests this may be an important composition and we speculate that it might represent an enriched mantle average (EMA) composition preferentially sampled at the melt fraction operating in ocean island basalt (OIB) generation. The high He- 3/He-4 component has Nd and Sr isotope compositions that are indistinguishable from mid-ocean ridge basalts (MORB) and more "depleted" than other common mantle components that have between postulated from the global OIB-MORB data base. We suggest that this composition is best explained as a mixture between depleted and (He-rich) primordial mantle and we therefore refer to this mixed composition as He-recharged depleted mantle (HRDM). Even the Baffin Island lavas require only a small proportion ( lt 10%) of primordial mantle and most OIB demand lt 2%. If high He-3/He-4 is a lower mantle signal, then only limited lower mantle involvement in plumes is indicated. However, the low proportion of primordial material also removes the requirement for an extensive primordial reservoir. These observations can be reconciled with whole mantle convection provided (1) a small volume of relatively undegassed and undepleted mantle has survived intact and (2) an inventory of incompatible trace elements. similar to that housed in the continents, but probably very minor in volume ( lt 5% of the mantle), resides in enriched domains in the mantle.