Leucogranites of the nw Himalaya: Crust-mantle interaction beneath the Karakoram and the magmatic evolution of collisional belts.

The Karakoram Axial Batholith in N. Pakistan records the magmatic development of the Eurasian continental margin since the late Jurassic. Magmatism prior to the collision of India with Eurasia at c.45Ma is represented by subduction-related, calc-alkaline granodiorite plutonism. The chemical variation within these plutons is caused by high-level fractionation processes. However, heterogeneous isotope data suggests that the source of these magmas was the mantle wedge, enriched by87 a subducted slab component, with the melts being contaminated by a Sr-rich crustal component. There are two types of post-collisional leucogranite. The Sumayar pluton is related to restricted partial melting of sillimanite-grade metapelites triggered by fluxing of fluids from the underthrust Indian crust. This water-saturated, minimum melt is considered to be an analogue of the High Himalayan leucogranites. The other Karakoram leucogranites are related by fractionation to a more basic monzogranitic parent, whose geochemistry indicates a lower crustal source. However, melting of typical crust cannot explain the anomalously high large ion lithophile element (LILE) content of the monzogranites. Associated with the granites are ultra- potassic, LILE-enriched lamprophyres. This LILE-enrichment is attributed to alteration of the mantle wedge by fluids and/or siliceous melts from the slab. Amphibole in the resulting metasomatic assemblage acts as a sink for the otherwise incompatible LILE. As a result of heating and thermal relaxation beneath the thickened continental crust, enriched amphibole, stable in the pre-collisional mantle wedge beneath the Karakoram, dehydrated or melted some 20Ma after collision to give the lamprophyres. Fluid precursors to this melting contaminated the source region of the granites selectively enriching it in LILE and triggering/promoting melting. The identification of the above magma-types, which have different generative processes and magmatic triggers, in other collisional environments will lead to information about the evolution of similar orogenic belts.