Origin of trondhjemite leucosome in a kyanite-grade migmatite from the Higher Himalaya Crystalline (Bhagirathi valley): Insight into leucosome forming processes

K2O-poor or trondhjemite leucosome is a common feature of migmatites in the Himalayas and other orogens. The origin of K2O-depleted leucosomes has been attributed to several melting scenarios, such as residuum left after melt extraction and magma accumulation. However, the process leading to the formation of leucosome of trondhjemite composition needs to be better understood. In the kyanite zone of the Higher Himalaya Crystallines (HHC) along the Bhagirathi valley, partial melting of metapelite resulted in the formation of two types of leucosomes: K2O enriched granitic L-1 leucosomes formed stromatic veins of variable thickness and K2O poor trondhjemitic L-2 leucosomes occur as patches and lenses. The L-1 leucosomes are characterized by positive and negative Eu anomaly, variable depleted REE content, and represent in-source leucosomes. The L-2 leucosomes are REE depleted, have small negative Eu anomaly (0.8-0.9), and show a clear trend for anatectic melt-residuum-protolith in major-oxide plots of compatible versus incompatible elements. The overall depletion of REE, Sc, V, Cr, Ni, Co, Ti, Th, Nb, Hf, P2O5, and Zr/Zr* <1 in both leucosomes is indicative for disequilibrium melting. Plagioclase composition in the L-1 leucosomes and the associated melanocratic layers show a narrow range for XAb (0.81-0.86). Pl composition measured for two leucosome layers in L-2 leucosomes show near-constant values between XAb =0.81 and XAb =0.85, but the melanosome shows reverse-zoned plagioclase grains with variable XAb (0.69-0.80). Phase equilibria modelling in the system MnNKCFMASTH (Perplex version 7.0.6, thermodynamic database: Holland and Powell (2011)) using the protolith composition correlated to L-2 leucosome shows a clockwise P-T path in the kyanite field. The evolution of melt composition modeled as a function of pressure and X(H2O) at constant temperatures of 690, 710, and 750° C shows that X(H2O) has little to no effect on melt composition. Melting at low pressure produces a granitic composition, and with increasing pressure at constant temperature and X(H2O), melt composition evolves from granite to trondhjemite. The higher the temperature, the smaller the chemical effect. We conclude that the K2O-poor trondhjemite L-2 leucosomes were produced at incipient partial melting by disequilibrium melting of plagioclase by the reaction Mus + Pl1 (Ab -rich) → melt + Pl2 (Ab-poor). The melt volume is too small to form an interconnected network on which melt could escape, and thus, the melt formed patches and lenses. With increasing temperature, the melting reaction evolved to muscovite dehydration melting, producing a large melt fraction of granitic composition (L-1 leucosome). While previous studies consider K2O poor-leucosomes as residuum melt after melt extraction or represent melt accumulation, we suggest that K2O poor melts could be anatectic melts produced at the beginning of partial melting at high pressure in low melt fractions. References: Connolly, J.A., 2005. Earth and Planetary Science Letters, 236(1-2), pp.524-541. Holland, T.J.B. and Powell, R.T.J.B., 2011.  Journal of metamorphic Geology, 16(3), pp.309-343.