The initiation and magmatic evolution of a juvenile island arc : the Kohistan arc, Pakistan Himalaya
The Kohistan arc, situated in the northwestern Himalaya of North Pakistan, is a Cretaceous intra-oceanic island arc which was erected during subduction of the Tethys Ocean consequent on the northward flight of India. Following accretion to the Asian margin, the arc behaved as an Andean-type continental arc prior to the collision of India with the Asian continent, which uplifted the arc and tilted it, thus providing a unique opportunity to study a complete succession of rocks from the very earliest stages of arc evolution. A combination of fieldwork, geochemical and radiogenic isotope analysis, and rare earth element modelling are used to determine the main magma sources in the mantle beneath the arc. The Kamila Amphibolites fall into two successions. The 'E-type' rocks have a MORB-type signature and were formed from 6% partial melting of a primitive, garnet¬bearing, mantle source. The 'D-type' Kamila Amphibolites have an arc signature and represent the earliest arc volcanic rocks. They originated from 15.5% partial melting of a primitive spinel-bearing, mantle source. The Jaglot Group (Gashu Confluence Volcanics and Peshmal Volcanics) and Western Volcanics of the Chalt Volcanic Group were generated by 7.5%, 13% and 2.5% partial melting, respectively, of a primitive, garnet¬bearing, mantle source. The Hunza Valley Volcanics of the ChaIt Volcanic Group, which contain boninites, have MORB-type chemistries and an arc signature. These rocks were generated from 15% melting of a depleted, spinel-bearing mantle source which may be the residuum from partial melting which produced the 'E-type' Kamila Amphibolites. That the Hunza Valley Volcanics were generated from this source is consistent with the 'E-type' Kamila Amphibolites fonning the basement and the Hunza Valley Volcanics occurring in the back-arc. Radiogenic Nd, Sr and Pb isotope analyses address an hypothesis that the magmas were generated from 'Dupal' -type mantle. The results indicate that this is not the case, but show that the isotopic chemistry of the rocks is the result of fluids from dehydration and melting of sediments carried on the downgoing ocean crust into the subduction zone affecting magma chemistry. A recent controversy concerns the polarity of subduction beneath the arc. This is partly' based on the presence of boninites, and the prevalent understanding that this rock-type occurs solely in the fore-arc. Evidence from this thesis suggests that the boninites of the Hunza Valley Volcanics were erupted into a back-arc setting, and arguments are made that boninites may also be erupted into the back-arc, and that the polarity of subduction was to the north. A model is presented for the erection of the Kohistan arc in which subduction was initiated by gravitational instability at an oceanic transform fault which connected two mid-ocean spreading ridge segments. The first arc magmas were generated by decompression melting beneath extending lithosphere during initial subsidence, and as this turned into subduction, magmas were generated at progressively greater depths in the spinel and garnet lherzolite facies, respectively. Lithospheric extension and rifting behind the volcanic front provided the setting for deompression melting to produce the back-arc Hunza Valley Volcanics, including high-Mg basalts and andesites. The arc signature in these rocks became reduced as the spreading centre developed and became progressively removed from the vicinity of the subduction zone.