A rare kimberlite mantle xenolith displaying a direct contact relationship between a MARID rock and metasomatised peridotite is described. There is an abrupt textural transition from a MARID (phlogopite-K-richterite-diopside) portion, to a 2 cm rind of peridotite (olivine pseudomorphs-phlogopite-K-richterite-diopside-chromite). Small, but significant differences in mineral composition exist across the boundary, the most important of which are lower Fe/Mg ratios in the peridotite rind. Our interpretation is that MARID magmas crystallised and expelled fluids which metasomatised surrounding peridotite wall-rock, and the xenolith has sampled a cumulate-wall-rock contact of this nature.
A suite of peridotites from the Kimberley area has been studied to investigate the vertical extent and distribution of metasomatism in the subcontinental lithospheric mantle. Geothermobarometry on garnet-bearing peridotites yields pressures and temperatures that fit closely to a continental geotherm. They define a continuous trend which extends from about 150 km depth and 1000°C (sparsely phlogopitized and phlogopite-free samples), to 100 km depth and 780 °C (mostly phlogopite-rich garnet phlogopite peridotites (GPP), including some edenitic amphibole-bearing samples). Garnet-free phlogopite peridotites (PP) yield a similar temperature range (1010–755 °C) but their average temperature is lower. The most highly metasomatised peridotites, the phlogopite K-richterite peridotites (PKP), yield the lowest average temperatures, with estimates ranging between 885 an 745 °C. While pressures cannot be directly calculated for the garnet-free rocks, mineral stability relationships place the PKP group at a maximum depth of 120 km, and probably around 100 km depth. Hence metasomatized peridotites are believed to be present from ∼ 100 km to at least 150 km depth. Although some relatively unmetasomatized peridotites are present throughout the depth interval indicated, the average degree of metasomatism exhibited by the peridotites appears to increase with decreasing depth. This is shown by negative correlations between temperature (and pressure) and parameters such as modal abundances of hydrous minerals, and concentrations of incompatible elements in whole-rocks and minerals. Ba contents of phlogopites in PKP rocks constitute the only exception to this trend of increasing abundances with decreasing temperature. While whole-rock Ba concentrations are indeed highest on average in the PKP rocks, the Ba contents of PKP phlogopites are unexpectedly low-they average 530 ppm Ba, compared to an 3400 ppm Ba in GPP phlogopites. Preferential partitioning of Ba into Ba-rich titanates such as lindsleyite at shallow depths may be the cause of this feature, but there are problems with such a model. Like PKP rocks, MARID (Mica-Amphibole-Rutile Ilmenite-Diopside) nodules are constrained by the presence of K-richterite to have formed at < 120 km depth. Mineralogical, chemical and isotopic similarities between PKP and MARID magmas dominated the PKP-metasomatism, and the low Ba contents of PKP phlogopites could thus be a feature inherited by equilibration with the MARID fluids. Although MARID-derived fluids are thought to have been responsible for the PKP-metasomatism, most GPP rocks were metasomatized below the depth of MARID crystallization. We consider that there may have been input of metasomatic fluids at two levels in the sub-continental mantle lithosphere below Kimberley; (i) fluids derived from an unknown source at the base of the lithosphere (possibly crystallized melts at the lithosphere/asthenosphere boundary); and (ii) fluids expelled from crystallising MARID magmas at shallow depths (∼ 100 km).