Evolution of the melt source during protracted crustal anatexis: An example from the Bhutan Himalaya

Abstract

The chemical compositions of magmatic zircon growth zones provide powerful insight into evolving magma compositions due to their ability to record both time and the local chemical environment. In situ U-Pb and Hf isotope analyses of zircon rims from Oligocene– Miocene leucogranites of the Bhutan Himalaya reveal, for the first time, an evolution in melt composition between 32 and 12 Ma. The data indicate a uniform melt source from 32 Ma to 17 Ma, and the progressive addition of an older source component to the melt from at least ca. 17 Ma. Age-corrected ɛHf ratios decrease from between −10 and −15 down to values as low as −23 by 12 Ma. Complementary whole-rock Nd isotope data corroborate the Hf data, with a progressive decrease in ɛNd(t) from ca. 18 to 12 Ma. Published zircon and whole-rock Nd data from different lithotectonic units in the Himalaya suggest a chemical distinction between the younger Greater Himalayan Series (GHS) and the older Lesser Himalayan Series (LHS). The time-dependent isotopic evolution shown in the leucogranites demonstrates a progressive increase in melt contribution from older lithologies, suggestive of increasing LHS involvement in Himalayan melting over time. The time-resolved data are consistent with LHS material being progressively accreted to the base of the GHS from ca. 17 Ma, facilitated by deformation along the Main Central thrust. From 17 Ma, decompression, which had triggered anatexis in the GHS since the Paleogene, enabled melting in older sources from the accreted LHS, now forming the lowermost hanging wall of the thrust. INTRODUCTION Crustal melting is a fundamental process both for chemical differentiation and for facilitating ductile deformation of the Earth’s continental crust. In the Himalaya, late Oligocene–Miocene leucogranites provide a well-documented example representing crustal melting induced by continental collision. Oxygen and hafnium isotope data from Himalayan leucogranite zircon rims demonstrate that these granites formed via pure crustal melting with no detectable mantle input (Hopkinson et al., 2017). Melting took place along the Himalaya orogen from at least 25 Ma to 9 Ma (Guo and Wilson, 2012), although in southern Tibet, crustal melting of Eocene age has been reported (Aikman et al., 2012). Despite this extensive period of regional anatexis, no time-dependent change in the magmatic source has been recognized thus far. The evolving thermal regime and/or the continued burial of different material into a stable thermal regime during collisional orogenesis should lead to migration of melt source regions or change in source rocks through time, and to the involvement of sources from different crustal levels (Fiannacca et al., 2017). The recognition of the temporal evolution of melt zones would contribute to more refined models of thermal and structural evolution of the continental crust during orogenesis. The apparent absence of time-dependent trends in the Himalaya may result from the lack of chronological resolution. Here we evaluate the Hf isotope composition of zircon sampled from leucogranites in Bhutan (Eastern Himalaya), along with new whole-rock Nd isotope data. We observe a temporal evolution in the source region, and these data provide, for the first time, direct evidence for a time-dependent change in the mid-crustal material undergoing melting and decompression during Himalayan crustal thickening. REGIONAL SETTING Across the orogen, leucogranites intrude the Greater Himalayan Series (GHS), an upper amphibolite-facies metasedimentary package of primarily Neoproterozoic protolith age (Ahmad et al., 2000; Gehrels et al., 2011). We collected 12 samples from nine localities in Bhutan, where the leucogranites form discrete sheets within the uppermost GHS, close to the tectonic contact with Tethyan sediments. Samples include two-mica (samples 1G03, 3A03, 4D01, 1247, 1251, CWB16, and CWB23), tourmaline-bearing (samples 1G02 and 1215), garnet-bearing (samples 1D01 and 3A02), and pegmatitic two-mica (sample 1G01) types. Samples 1G01, 1G02, and 1G03 are from intersecting sheets from the same exposure, as are samples 3A02 and 3A03 (Fig. 1). We have previously analyzed the isotopic composition of zircon growth zones from all of these samples (Hopkinson et al., 2017). The base of the GHS is marked by the Main Central thrust, an orogen-parallel thrust zone that has facilitated India-Asia convergence since at least the late Paleogene (Mottram et al., 2014, 2015). Structurally below the Main Central thrust is the Lesser Himalayan Series (LHS), a primarily Paleoproterozoic-sourced stack of metasedimentary rocks (Ahmad et al., 2000; *E-mail: [email protected] †Current address: Department of Geological Sciences and Geological Engineering, Queen’s University, Kingston, Ontario, K7L 3N6. §Current address: School of Earth and Environmental Sciences, University of Portsmouth, Ports-