Are high ^3He/^4He ratios in oceanic basalts an indicator of deep-mantle plume components?

The existence of a primordial, undegassed lower mantle reservoir characterized by high concentration of ^3He and high ^3He/^4He ratios is a cornerstone assumption in modern geochemistry. It has become standard practice to interpret high ^3He/^4He ratios in oceanic basalts as a signature of deep-rooted plumes. The unfiltered He isotope data set for oceanic spreading centers displays a wide, nearly Gaussian, distribution qualitatively similar to the Os isotope (^(187)Os/^(188)Os) distribution of mantle-derived Os-rich alloys. We propose that both distributions are produced by shallow mantle processes involving mixing between different proportions of recycled, variably aged radiogenic and unradiogenic domains under varying degrees of partial melting. In the case of the Re–Os isotopic system, radiogenic mid-ocean ridge basalt (MORB)-rich and unradiogenic (depleted mantle residue) endmembers are constantly produced during partial melting events. In the case of the (U+Th)–He isotope system, effective capture of He-rich bubbles during growth of phenocryst olivine in crystallizing magma chambers provides one mechanism for ‘freezing in’ unradiogenic (i.e. high ^3He/^4He) He isotope ratios, while the higher than chondritic (U+Th)/He elemental ratio in the evolving and partially degassed MORB melt provides the radiogenic (i.e. low ^3He/^4He) endmember. If this scenario is correct, the use of He isotopic signatures as a fingerprint of plume components in oceanic basalts is not justified