Skip to main content
eScholarship
Open Access Publications from the University of California

UC Santa Barbara

UC Santa Barbara Electronic Theses and Dissertations bannerUC Santa Barbara

Mantle source versus process for the origin of geochemical heterogeneity in olivine-hosted melt inclusions from Hawaiʻi and Samoa

No data is associated with this publication.
Abstract

Olivine-hosted melt inclusions provide a multitude of opportunities to glimpse magmatic processes at depth because olivine is an early-crystallizing mineral in mantle-derived melts, and melt inclusions often capture a period of magmatic history that is rarely recorded and often obscured by whole rock or matrix glass chemistry of lavas. My research uses geochemical analyses of olivine-hosted melt inclusions to differentiate between the influences of mantle source heterogeneity versus magmatic processes on the composition of mantle-derived magmas at oceanic hotspots.

(I) Olivine-hosted melt inclusions from Puʻu Wahi, Mauna Loa, Hawaiʻi have unusually high (Sr/Ce)N ratios though lack elevated Al2O3, suggesting that plagioclase was not assimilated. This was dubbed a “ghost plagioclase” signature. The high (Sr/Ce)N ratios have been interpreted in the past to be signatures of recycled plagioclase-rich oceanic crust in the mantle source. However, the 87Sr/86Sr of individual olivine-hosted melt inclusions, whether with a high or normal (Sr/Ce)N ratio, are within the range expected for Mauna Loa lavas. Thus, the high (Sr/Ce)N ratios were likely the result of diffusive interaction with gabbro beneath Mauna Loa and not from incorporation of recycled oceanic crust in the mantle source.

(II) In Chapter 2, I argue that the enriched mantle 2 (EM2) mantle source is just as damp with respect to H2O as depleted mantle (DM) or non-enriched mantle (non-EM) sources. Previous studies suggested that the EM2 source is dry with respect to H2O because lower H2O/Ce corresponded with higher 87Sr/86Sr in submarine glasses. The H2O/Ce ratios of Samoan olivine-hosted melt inclusions are similar to those of non-EM submarine glasses, suggesting that closed-system degassing, where both CO2 and H2O degas, led to lower H2O/Ce in Samoan pillow glasses compared to melt inclusions from the same seamounts. I use the correlation between CO2/Nb and (La/Sm)N in OIB and MORB glasses to estimate primary melt CO2 concentrations. In general, pillow glasses with higher 87Sr/86Sr have higher (La/Sm)N, so I predict high CO2 concentrations in the primary melts from EM sources, which leads to more extensive degassing compared to primary melts with lower CO2 concentrations. A major implication of this study is that pillow glasses from EM melts do not reliably retain the H2O content of the melts.

(III) Olivine-hosted melt inclusions from Oʻahu rejuvenated volcanism are extremely alkalic and incompatible trace element-enriched. The H2O/Ce of these inclusions is very low (12–63) and deviates from the OIB pillow glass array of lower H2O/Ce with higher 87Sr/86Sr (Oʻahu rejuvenated lavas have low 87Sr/86Sr), so it is not the enrichment of the mantle source that influences the H2O/Ce ratio in erupted glasses or melt inclusions, but degassing of H2O in melts. I propose that low-degree melts (with high (La/Sm)N) are more alkalic and CO2-rich. CO2-rich melts degas more H2O compared to CO2-poor melts, resulting in melts with low H2O/Ce compared to higher-degree melts that are tholeiitic and CO2-poor. Using mantle melting models, I find that the CO2 content of primary OIB melts is largely influenced by the degree of melting.

Main Content

This item is under embargo until November 1, 2025.