J. metamorphic Geol., 2016, 34, 821–843 doi:10.1111/jmg.12204 Himalayan leucogranites and migmatites: nature, timing and duration of anatexis R. F. WEINBERG School of Earth, Atmosphere and Environment, Monash University, Clayton, Vic. 3800, Australia ([email protected]) ABSTRACT Widespread anatexis was a regional response to the evolution of the Himalayan-Tibetan Orogen that occurred some 30 Ma after collision between Asia and India. This paper reviews the nature, timing, duration and conditions of anatexis and leucogranite formation in the Greater Himalayan Sequence (GHS), and compares them to contemporaneous granites in the Karakoram mountains. Himalayan leucogranites and associated migmatites generally share a number of features along the length of the mountain front, such as similar timing and duration of magmatism, common source rocks and clock- wise P–T paths. Despite commonalities, most papers emphasize deviations from this general pattern, indicating a fine-tuned local response to the dominant evolution. There are significant differences in – – P T XH2O conditions during anatexis, and timing in relation to regional decompression. Further to that, some regions underwent a second event recording melting at low pressures. Zircon and monazite ages of anatectic rocks range between c. 25 and 15 Ma, suggesting prolonged crustal melting. Typi- cally, a single sample may have ages covering most of this 10 Ma period, suggesting recycling of accessory phases from metamorphic rocks and early-formed magmas. Recent studies linking monazite and zircon ages with their composition, have determined the timing of prograde melting and retro- grade melt crystallization, thus constraining the duration of the anatectic cycle. In some areas, this cycle becomes younger down section, towards the leading front of the Himalayas, whereas the oppo- site is true in other areas. The relationship between granites and movement on the South Tibetan Detachment (STD) reveals that fault motion took place at different times and over different dura- tions requiring complex internal strain distribution along the Himalayas. The nature and fate of mag- mas in the GHS contrast with those in the Karakoram mountains. GHS leucogranites have a strong crustal isotopic signature and migration is controlled by low-angle foliation, leading to diffuse injec- tion complexes concentrated below the STD. In contrast, the steep attitude of the Karakoram shear zone focused magma transfer, feeding the large Karakoram-Baltoro batholith. Anatexis in the Karakoram involved a Cretaceous calcalkaline batholith that provided leucogranites with more juve- nile isotopic signatures. The impact of melting on the evolution of the Himalayas has been widely debated. Melting has been used to explain subsequent decompression, or conversely, decompression has been used to explain melting. Weakening due to melting has also been used to support channel flow models for extrusion of the GHS, or alternatively, to suggest it triggered a change in its critical taper. In view of the variable nature of anatexis and of motion on the STD, it is likely that anatexis had only a second-order effect in modulating strain distribution, with little effect on the general his- tory of deformation. Thus, despite all kinds of local differences, strain distribution over time was such that it maintained the well-defined arc that characterizes this orogen. This was likely the result of a self-organized forward motion of the arc, controlled by the imposed convergence history and energy conservation, balancing accumulation of potential energy and dissipation, independent of the presence or absence of melt. Key words: anatexis; continental collision; leucogranite; migmatite; self-organization. 1998). Decades of research helped elucidate their spa- INTRODUCTION tial distribution, timing and conditions of anatexis, as Miocene granites and migmatites are some of the well as their role in the evolution of the Orogen. Yet most studied features of the Himalayan-Tibetan Oro- their relative timing in relation to major structures, gen. They are central to understanding its evolution the origin of the heat source for melting and the tec- and the basis for many evolutionary models (England tonic impact of melting, remain widely debated et al., 1992; Harris & Massey, 1994; Huerta et al., (Hodges, 2000; Yin & Harrison, 2000; Kohn, 2014). 1996; Harrison et al., 1997a, 1998, 1999b; Hodges, Several tectonic models have been presented to © 2016 John Wiley & Sons Ltd 821 822 R. F. WEINBERG explain the evolution of these mountains (see Kohn, and key features significant to understanding the 2014 for a summary) many of which are fundamen- origin of collisional granitoids and the behaviour of tally linked to crustal melting. For example, channel the Orogen. It is demonstrated how the process of models require the presence of melt, weakening thick granite generation is simultaneously tightly con- sections of the Greater Himalayan Sequence (GHS), strained as well as highly variable responding to to allow the flow of rocks from underneath Tibet local conditions. The paper starts with an introduc- (Beaumont et al., 2001; Grujic et al., 2002; Jamieson tion to the geological setting of the anatectic rocks, et al., 2004). followed by: (i) melting conditions and melting reac- Miocene granitoids are typically Ms–Bt–Grt–Tur tions across the Orogen; (ii) timing and duration of peraluminous leucogranites, and crop out in three magmatism, and significance for granite generation; distinct regions in the Orogen with significant tec- and (iii) granite ages constraining movement dura- tono-thermal and temporal differences: (i) within the tion on the STD. The section on melting conditions GHS along the Himalayan front, cropping out (i) is separated from the section on timing and dura- between the Main Central Thrust (MCT) zone and tion of magmatism (ii) in order to emphasize chang- the South Tibetan Detachment (STD), and known as ing patterns along the Himalayas. Himalayan Greater or Higher Himalayan leucogranites; (ii) north anatectic rocks are then contrasted to those from of the STD, in association with domes (the North the Karakoram before a discussion focusing on how Himalayan granites; Le Fort, 1986; Harrison et al., the data inform and constrain the tectonic evolution 1998; King et al., 2011); and (iii) in the Karakoram of the Orogen. The paper finishes by suggesting that Range, north of the Indus-Tsangpo Suture Zone the variety of conditions and timing of magma gen- (Fig. 1). A fourth and much younger anatectic event eration and movement on the STD had only a (<4 Ma) is exposed in the two syntaxes, at Namche second-order effect in modulating strain distribution. Barwa and Nanga Parbat. These were local responses to a broader pattern of This article reviews the literature on migmatites self-organized motion imposed by the potential and leucogranites (referred to collectively as anatec- energy stored in the Orogen and the continued tic rocks) of the GHS and then contrasts them to northward indentation of India, which allowed for anatectic rocks from the Karakoram Range. The the orderly development of the Orogen and mainte- focus is dominantly on the more recent literature nance of the orogenic arc. Karakoram N Baltoro Transhimalaya Indus-Tsangpo Suture Zone (ITSZ) K akoram a Tethyan Himalayan Sequence (THS) r r a a K k Greater Himalayan Sequence (GHS) o ra High Himalayan and Karakoram m leucogranites fa North Himalayan leucogranites u lt Zanskar Lesser Himalayan Sequence (LHS) ST Pleistocene basins D Main Central thrust (MCT) MC Tangtse MBT T Leo Pargil Main Boundary thrust (MBT) Zanskar South Tibetan Detachment (STD) MB TIBET 32°N T Annapurna Rongbuk Mabja 75°E S Manaslu TD Everest Dinggye Gharwal Him. Makalu Khula Kangri 250 km STD Lhasa MC T INDIA MB Gangotri T STD MCT T STD MCT MB Malari Modi Khola Bura Buri Kathmandu Sikkim MBT Bhutan Arunachal Him. Langtang 90°E Fig. 1. Simplified geological map of the southern part of the Himalayan-Tibet Orogen showing the GHS, between the STD and the MCT, and the two belts of leucogranites, including the Karakoram-Baltoro Batholith on the upper left (NW) of the figure. Locations refer to those mentioned in the text. © 2016 John Wiley & Sons Ltd HIMALAYAN LEUCOGRANITES AND MIGMATITES 823 separates it from the low-metamorphic grade rocks GEOLOGICAL SETTING OF LEUCOGRANITES of the THS. It began moving prior to 22 Ma, and The Himalayas and the Tibetan plateau result from has had a short duration in some places (Hodges collision between India and Asia between 54 and et al., 1996; Carosi et al., 2013; Finch et al., 2014), 50 Ma (Hodges, 2000). The Himalayas form a well- whereas in others it may have continued to move to defined arc, with a sinuous front forming salients and c. 11 Ma (Kellett et al., 2009). recesses at the scale of a few 100 km in length Miocene migmatites are found from within the (Bendick & Bilham, 2001; Mukul, 2010). Hodges MCT zone (Hubbard, 1989; Coleman, 1998; (2000), building on previous work, divided the evolu- Dasgupta et al., 2009) to the base of the STD (Pog- tion of the Himalayas into two deformation phases, nante, 1992; Kohn, 2014). Migmatites are commonly, linked to distinct metamorphic events: the Middle but not solely, developed in the 500–400 Ma gneisses Eocene–Late Oligocene Eohimalayan, recording crus- at the top of the GHS (e.g. Sikkim, Dasgupta et al., tal thickening with metamorphism peaking at c. 33– 2009; Zanskar, Finch et al., 2014; Horton et al., 2015; 28 Ma; and the Early Miocene to present Neohi- E Nepal, Groppo et al., 2012). These are physically malayan phase, associated with sillimanite-bearing linked with granites through complex channel net- rocks, anatexis and leucogranite intrusions. This works. The Higher Himalayan leucogranites form a phase is associated with a substantial change in tec- discontinuous belt along the highest structural levels tonics at the start of the Miocene, and remains of the GHS, generally immediately below or over- roughly unchanged, suggesting a quasi-steady-state printed by the STD.