PETROGENESIS OF THE GRANDE RONDE LAVAS: HIGH-P MELTING OF BASALTIC ROCK OR CRUSTAL CONTAMINATION?

There is no consensus on the petrogenesis of the Columbia River Basalt Group (CRBG). Much of the controversy centers on the volumetrically dominant Grande Ronde ‘Basalt’, which mostly consists of quite uniform and relatively evolved basaltic andesite. In particular, melting of basaltic compositions at mantle pressures has been proposed as a possible origin for Grande Ronde lavas [1,2]. Such material is unlikely to be ancient subducted basaltic crust recycled in a mantle plume because it would be depleted in LILE with respect to HFSE and this signature should be imparted to the melt, the opposite of what is observed in Grande Ronde lavas [3]. An alternative is melting of basaltic crust delaminated shortly prior to magmatism [2]. However, heavy REE characteristics of the Grande Ronde lavas are inconsistent with residual garnet present during partial melting, a feature required by any plausible eclogite melting scenario. Additionally, Nd isotope characteristics of Grande Ronde lavas require a component of cratonic crust in their source [3]. An important observation is that Grande Ronde lavas form a chemical continuum with the underlying, less differentiated Imnaha basalts; both formations contain a crustal component. A continuous chemical shift from a more ‘mantle-like’ to a more ‘crust-like’ signature occurs from the latest Imnaha flows through Grande Ronde magnetostratigraphic interval R1, representing a period of ~70 ka. This rapid transition is consistent with recent models for the growth and establishment of a large, lower crustal magma chamber on the site of a pre-existing dike complex that heated the crust to the point of large-scale melting above the critical melt fraction required for bulk magma transport. We propose that the Imnaha basalts are the product of such a dike complex , and the Grande Ronde lavas the product of the subsequent magma chamber.

[1] Takahashi et al., EPSL 162, 63-80 (1998); [2] Camp & Hanan, Geosphere 4, 480-495 (2008); [3] Wolff et al., Nature Geoscience 1, 177-180 (2008).