UC San Diego

Ultramafic xenoliths and mafic lavas have been instrumental in understanding the composition of Earth’s mantle and processes that influence its composition. These processes can be obscured by overprinting from eruption, emplacement and magma storage processes prior to the rocks coming to the surface. In this dissertation mantle processes are examined using melt depleted peridotite xenoliths and through the eruptive products of the mantle, ultramafic cumulate xenoliths, and ocean island basalts. Similarities between oceanic lithosphere and non-cratonic continental lithosphere (CLM) suggest these domains are linked. To begin, in chapter two, xenoliths from the Ferrel seamount are determined to sample a fragment of ancient Farallon Plate lithosphere at the abandoned spreading ridge. Next, similarities between these xenoliths and non-cratonic CLM are used to develop a new model for the evolution of oceanic lithosphere to non-cratonic CLM. Chapter three explores the isotopically distinct Loa and Kea trends in Hawaii through a novel approach of studying cumulate xenoliths. The Loa and Kea trends show disparities in long-lived radiogenic isotope systems, including Re-Os, that suggest these magmas are sourced from two distinct mantle reservoirs. This study evaluates the potential for mixing between these two trends considering recently published seismic evidence suggesting a shared magmatic system. Chapter four shows the Canary Islands are built atop lithosphere more similar to the Atlantic rather than the adjacent non-cratonic African lithosphere. Lavas erupted here show evidence of lithospheric contamination and likely sample two different mantle endmembers, yet lack the characteristic bilateral zoning observed in the Pacific Ocean islands. Finally, chapter five examines an emerging continental rift zone east of the Sierra Nevada mountains in California. Peridotite and Pyroxenite cumulate xenoliths from the Papoose Canyon volcanic sequence in the Owens Valley have trace element and Os isotopic systematics which suggest crustal contamination has had a significant impact on their parental magmas. A model of Os isotopic composition and HSE fractionation suggests that only a small amount of crustal addition is required to replicate the observed compositions. Despite the apparent crustal contamination, absolute and relative HSE abundances preserve information on the parental melt.