Oxygen isotope survey of the northern Canadian lithospheric mantle: Implications for the evolution of cratonic roots

Aleksandar Miskovic (NTGO), Ryan Ickert (U. of Alberta), Graham Pearson (U. of Alberta), and Richard Stern (U. of Alberta)

Problematic
The sub-continental lithospheric mantle (SCLM) is the source region for kimberlitic magmas and host to their precious cargo. It forms the roots to the earliest crust and represents the most ancient mantle domain on Earth. Hundreds of oxygen isotope data have been reported for various mantle domains in the past two decades, yet the full geodynamic significance of these data is still largely unappreciated. When coupled to the major element systematics, the oxygen isotope data have the potential to illuminate the longstanding conundrum of the origin of highly depleted lithospheric roots beneath the Archean crust, specifically in the context of mantle geodynamics. The current debate is centered on quantifying the relative contributions of mantle plumes (asthenospheric source) and subduction zone processes (lithospheric source) in the SCLM genesis (Fig. 1).

Figure 1. A carton depicting one of the proposed mechanisms for generation of the subcontinental lithospheric mantle by subcretion of altered oceanic lithosphere (after Cyn-Ty Lee et al., 2010)

Findings
We analysed 61 grains of olivines extracted from peridotite xenoliths entrained by kimberlites and pericratonic basalts worldwide; the most extensive high-precision in situ δ18O database to date. Olivines analysed by multi-collector secondary ion spectrometry (at CCIM, U of A) in this study come from the following localities: Paleogene to Jurassic Diavik and Jericho kimberlites (Slave Province), Cretaceous Somerset Island kimberlites (Rae Craton), Late Neoproterozoic kimberlites of Western Greenland (North American Craton), Middle Cretaceous to Late Mesoproterozoic Kimberly, Finsch, Letseng-la-Terae and Premier kimberlites (Kaapvaal Craton), Early Carboniferous Udachnaya pipe (Siberian Platform), and its pericratonic alkaline olivine basalts erupted during Miocene (Vitim volcanic field).

Figure 2. The Slave SCLM compared to a global compilation of δ18O values in olivine extracted from mantle xenoliths carried by kimberlites and related rocks

The SCLM δ18O data are normally distributed about the mean value of 5.30 ‰ (±0.22; 2sigma) and largely corroborate previous upper mantle compilations (Fig. 2). However, we find no correlation between the oxygen isotope ratios and the whole-rock major or trace element chemistry, rhenium-depletion (TRD) ages (i.e. minimum age of lithosphere formation), nor equilibrium pressure-temperature conditions from which the mantle xenoliths were derived. Based on the new data we find no statistical difference between the oxygen isotope composition of cratonic peridotites and modern MORB-source mantle. These observations rule out an origin for SCLM via subcretion of serpentinised oceanic lithosphere (Fig. 3). However, the data cannot discriminate against the subduction of relatively un-altered oceanic lithospheric mantle versus a plume origin. Furthermore, the overlap of the SCLM data with the mean global δ18O value of mid-oceanic ridge basalts of 5.17‰ (±0.21; 2sigma) has important implications for the long-term relationship between the convecting asthenospheric mantle and ancient refractory cratonic roots.

Figure 3. Oxygen isotope systematics of serpentinised oceanic lithosphere displays significant variance and cannot be linked to the uniform δ18O SCLM dataset.