OXYGEN-ISOTOPE CONSTRAINTS ON THE SOURCES OF HAWAIIAN VOLCANISM

We have measured oxygen isotope ratios in 99 separates of olivine and 14 separates of plagioclase or glass from Hawaiian lavas. These data c onfirm that the source(s) of some Hawaiian basalts are lower in delta( 18)O than peridotite xenoliths and the source region for mid-ocean rid ge basalts (MORE). Our data document correlations between oxygen and r adiogenic isotope ratios and consistent differences in delta(18)O betw een volcanoes. Low values of delta(18)O are associated with a 'deplete d' component that is relatively high in Pb-206/(204)pb, low in He-3/He -4, and anomalously low in (207)pb/ Pb-204 relative to (206)pb/Pb-204. This component is preferentially sampled in lavas from the so-called Kea trend volcanoes (Kilauea, Mauna Kea, Kohala and Haleakala). Low de lta(18)O values in the 'Kea' component suggest that it is hydrothermal ly altered oceanic crust. The similarity of the Kea end member to Paci fic MORE in terms of Sr, Nd, and Pb isotope ratios further suggests th at this component is assimilated from the local Pacific plate in subcr ustal magma chambers. Anomalous Pb-206/ Pb-204- Pb-207/Pb-204 relation ships indicate recent enrichment in U/Pb in this component and further support the hypothesis that this component is young (< 10(8) yr) Paci fic crust. The isotopic distinctions between Loa and Kea trend volcano es implies a systematic difference in the magma supply and plumbing sy stems of volcanoes on these two trends. Samples from Lanai and Koolau have 'enriched' radiogenic isotope compositions (radiogenic Sr and non -radiogenic Nd and Pb) and higher delta(18)O than typical upper mantle values, suggesting the incorporation of recycled sediment and/or ocea nic crust in their sources. Other isotopic end members to Hawaiian lav as (e.g., high He-3/He-4 and post-erosional lavas) have delta(18)O val ues within the range typical of the upper mantle.