Date of Award
Open Access Dissertation
Earth and Ocean Sciences
Oceanic island basalts (OIB) which are derived from mantle plumes sample a compositionally heterogeneous mantle, as revealed by their radiogenic isotope systematics. Isotope compositions recurring in distinct hotspots have been used to characterize the composition and heterogeneity of the mantle as well as the thermochemical structure of upwelling plumes, with the assumption that the isotope composition of erupted lavas directly reflects that of their mantle source.
Here I test how source transport, melting and hybridization in the upper mantle through space and time control the expression of deep mantle sources in erupted lavas. This work is focused on the Azores and Hawai‘i hotspots, which erupt in two distinctly different tectonic settings: at a triple divergent plate boundary for Azores and under a thick, fast moving plate for Hawai‘i.
I demonstrate that Azores lavas sample mantle sources of distinct age and composition, including two generations of recycled oceanic crust. Τhe spatial distribution of these components in erupted lavas is controlled by the tectonic structure of the lithosphere, which is in turn controlled by the interaction of the ultraslow Terceira Rift and the adjacent Mid-Atlantic Ridge. This peculiar tectonic setting generates small-scale (70 km) westwards migration of mantle asthenosphere under a thinning lithosphere, as evidenced by the decrease in enriched components in lavas from east to west over ~ 70 km. The spatial expression of mantle sources in Azores lavas is therefore controlled by the tectonic features of the Azores Plateau, and not uniquely by the spatial distribution of mantle sources in the upwelling plume.
Hafnium isotope data for shield and rejuvenated Hawaiian lavas reveal a switch from a spatially homogeneous expression of mantle sources in lavas across the plume track to a bilateral isotope zonation in both shield and rejuvenated lavas after 3.5 [Ma]. This transition coincides in time with a change in direction of the Pacific Plate’s motion, which has been theorized by geodynamicists to produce isotope gradients in lavas due to acrossplume gradients in melting conditions affecting a plume source heterogeneous on a small scale (veins, minerals). Our data therefore suggest plume–lithosphere dynamical interaction in Hawai‘i is the cause of bilateral isotope zonation in lavas, rather than a bilateral source zonation in the upwelling plume.
A major conclusion of this work is that the structure and motion of the lithosphere in Azores and Hawai‘i exerts a first order control on the spatiotemporal systematics of isotope signatures in erupted lavas. Therefore, mantle plume-related erupted lava compositions do not necessarily reflect the chemical structure of the underlying mantle plume.
Another contribution of this work is the development of new isotope tracers of mantle sources and processes. I present radiogenic cerium isotope measurements (138La decay to 138Ce) for a selection of Azores and Hawai‘i lavas. This technique is of interest as La and Ce are as robust to fluid interaction as Sm, Nd, Lu and Hf, while also decoupling from Sm/Nd and Lu/Hf during melting and in sediments. I demonstrate the coupled Ce- Nd-Hf isotope systematics of OIB lavas is more sensitive to the sampling of recycled oceanic crust and sediments compared to Nd-Hf alone, and the occurrence of curved mixing in Ce-Nd and Ce-Hf isotope space can be used as a proxy for the modalities of source melting and hybridization.
Finally, I present a method to measure the stable Fe isotope signature of OIB lavas using a Thermo Neptune MC-ICP-MS with a precision of 0.02‰, on par with the most accurate published data, however through a significantly more time and cost-efficient method. This procedure uses the double-spike technique to monitor instrumental mass fractionation. I demonstrate that the condition of the source defining slit and the ability to move detectors are the most important factors influencing data quality. The iron isotope signature of erupted lavas is a direct tracer of the lithology of their source, which when compared with time-integrated tracers (radiogenic isotopes) can constrain the lithology of recycled sources in the convecting mantle, and the formation of hybrid sources prior to eruption. Preliminary data on Hawaiian shield lavas are at odds with the presence of a mafic component in the plume (shield source), while suggesting rejuvenated lavas sample a mafic or hybridized source. Inverse correlations between 87Sr/86Sr and 56Fe/54Fe in both shield and rejuvenated lavas go against the paradigm of recycled differentiated material existing under the form of mafic lithologies in the convecting mantle.
Béguelin, P.(2019). Mantle Heterogeneity from the Mantle Plume Perspective: Sources vs. Processes. (Doctoral dissertation). Retrieved from https://scholarcommons.sc.edu/etd/5415