Abstract
Poikiloblastic harzburgite xenoliths (P-type) from Borée, France are characterised by large (>1 cm), essentially unstrained olivines and high equilibrium temperatures (>1200 °C). Mineralogical data, trace element abundances and Sr-Nd-O isotopes of the constituent minerals are consistent with formation as a result of melt percolation-reactions in a lherzolite precursor during lithospheric erosion by an upwelling plume. This petrogenetic model contrasts with previous models involving isochemical recrystallisation from a granular lherzolite precursor (G-type) or derivation as metacumulates from tholeiitic magmas. Numerical simulation of percolation reactions at the lithosphere-plume boundary using the plate model of Vernières et al. (1997) indicates that the different textured xenoliths may represent mantle from different levels in a percolation-reaction column. If correct then the P-type harzburgites resulted from pyroxene-dissolving and olivine-producing reactions at increasing melt fraction (>3%) at the lower part of column (base of the lithosphere), whereas the G-type lherzolites were located within the low-porosity domain (<0.1%) above a permeability barrier, and are formed through a melt-rock reaction at decreasing melt mass. Given the very low melt fraction, the REE fractionation in this zone is controlled by chromatographic effects coupled with source effects of reaction. The variations in porosity, melt/rock ratio and melt-rock reaction mechanism are believed to be responsible for the diversity of REE patterns and striking correlation between REE abundance and texture in Borée xenoliths.
Similar content being viewed by others
Author information
Authors and Affiliations
Additional information
Received: 15 June 1997 / Accepted: 7 January 1998
Rights and permissions
About this article
Cite this article
Xu, YG., Menzies, M., Bodinier, JL. et al. Melt percolation and reaction atop a plume: evidence from the poikiloblastic peridotite xenoliths from Borée (Massif Central, France). Contrib Mineral Petrol 132, 65–84 (1998). https://doi.org/10.1007/s004100050405
Issue Date:
DOI: https://doi.org/10.1007/s004100050405