Comment on 'Silicic melts produced by reaction between peridotite and infiltrating basaltic melts: ion probe data on glasses and minerals in veined xenoliths from La Palma, Canary lslands' by Wulff-Pedersen et al.

摘要

In a recent paper, Wulff-Pedersen et al. (1999) interpret amphibole-phlogopite-bearing veins and silicic melts in peridotite xenoliths from San Antonio volcano on La Palma as the products of reactions between infiltrating basaltic melt and peridotite minerals. Their work builds on an earlier paper that was based on major element compositions of minerals and glasses (Wulff-Pedersen et al. 1996). The authors conclude that these infiltration-reaction-crystallization (IRC) processes took place in situ in the upper mantle and that the silicic melts may resemble a metasomatic agent. Their conclusion is largely based on (1) CO_2-rich fluid inclusions associated with silicic glasses that indicate mantle pressures, and (2) the time of vein formation based on diffusion calculations. Their arguments, however, are one-sided and do not consider recent results on similar samples from La Palma, some even from the same xenolith suite, which clearly show the complexity of multi-stage xenolith transport and magma ascent (Klugel et al. 1997; Hansteen et al. 1998; Klugel 1998). They make no attempt to separate processes related to xenolith sampling where heating, decompression and xenolith-magma interaction occur from those representing true in-situ wallrock reactions. Such a distinction is clearly difficult, but is essential if the nature and extent of metasomatic reactions are to be assessed properly. Because mantle xenoliths are derived from the wallrock of an active volcano's magma plumbing system, they inevitably represent biased samples showing overprinting by the host magmas.The purpose of this comment is to encourage a more complete discussion on the formation of veins and glasses in the La Palma xenoliths and their bearing on mantle processes. The comment is concerned with xenolith transport processes and the conclusions drawn by Wulff-Pedersen et al. (1999) that the infiltration-reaction-crystallization processes took place in situ, but does not question their model as a principal mode of peridotite-melt interaction.