Low δ18O in the Icelandic mantle and its origins: Evidence from Reykjanes Ridge and Icelandic lavas

Oxygen isotope ratios have been determined using laser fluorination techniques on olivine and plagioclase phenocrysts and bulk glasses from the Reykjanes Ridge and Iceland. δ18O in Reykjanes Ridge olivines shows hyperbolic correlations with Sr–Nd–Pb isotope ratios that terminate at δ18O = +4.5‰ at compositions almost identical to those of moderately enriched lavas on the Reykjanes Peninsula, Iceland. Samples with low δ18O show no indication of contamination by oceanic crust such as elevated Cl/K, and are too deep to have been influenced by meteoric water hydrothermal systems. They cannot represent Icelandic melts contaminated in the crust and transferred laterally along the ridge since fissure systems are strongly oblique to the ridge axis. It follows that Icelandic mantle advected along the ridge has low δ18O. The hyperbolic 143Nd/144Nd–δ18O correlation appears to be more strongly curved than magma mixing trajectories and suggests that melt fractions are ∼4.5× greater and source Nd contents ∼9× greater in the mantle at 63°N compared with that at 60°N. Primitive lavas from the Reykjanes Peninsula show linear correlations between olivine δ18O and 143Nd/144Nd or 206Pb/204Pb, extending to δ18O of +4.3‰ at 143Nd/144Nd close to the lowest ratios observed in Icelandic magmas. These correlations cannot be produced by melt mixing or crustal contamination because these would yield strongly hyperbolic trajectories. Lower δ18O seen in more evolved samples from the Eastern Rift Zone may reflect crustal contamination, though there is some evidence of a mantle source with lower δ18O in eastern Iceland. It is very difficult to explain the low δ18O of enriched Icelandic mantle sources on current understanding of mantle and crustal oxygen isotopes. There is no obvious reason why such low-δ18O sources should not contribute to other ocean islands. No oceanic crustal lithologies exist that could produce the low-δ18O enriched sources by recycling into the mantle, and there is no evidence for changes in δ18O of ophiolite suites with time, nor of changes during high-P metamorphism. Low δ18O appears to be associated with high 3He/4He, and we speculate that this signature may be characteristic of the host mantle into which ocean crust was recycled.