Eocene–Oligocene granitoids in southern Tibet: Constraints on crustal anatexis and tectonic evolution of the Himalayan orogen

Abstract

Tectonic models for the evolution of the Himalayan orogen interpret the Greater Himalayan crystalline complex (GHC) to be the result of either thick-skinned thrusting involved Indian basement, thin-skinned thrusting involving exotic terranes, middle-crustal ductile flow, or wedge extrusion of the Indian crust during India–Asia collision. Two key pieces of information needed to test the validity of these models is the temporal–spatial distribution of, and the identification of the dynamic mechanisms involved in, regional melting under southern Tibet. Here, we document an Eocene–Oligocene melting event in southern Tibet, which forms a 150-km-long, NW–SE-trending granitoid belt along the Zedong–Lhunze traverse between the Indus–Yarlung suture (IYS) and the south Tibetan detachment (STD). U–Pb dating of magmatic zircons indicates that this granitoid belt youngs northward from ∼46 Ma (in Lhunze) to ∼30 Ma (in Zedong). ⁴⁰Ar/³⁹Ar dating of deformed biotite within 42–46 Ma granitoids constrains the timing of shearing to ∼39–41 Ma.

The granitoid belt of southern Tibet is dominated by Eocene two-mica granites in the Tethyan Himalaya, with minor ∼30 Ma granodiorites along the IYS and ∼35 Ma granites in the Yelaxiangbo dome, where Indian mid-crustal rocks are exposed. The ∼35 Ma granites are characterized by variable Na₂O/K₂O ratios (1.03–4.44), relatively high Sr concentrations, and high Sr/Y (14.0–126.3) and La/Yb (11.1–42.8) ratios, which distinguish these granitoids from Miocene leucogranites in the Himalaya. Comparison of the Sr–Nd isotopic compositions of these granites with mid-crustal amphibolites exposed in the Yelaxiangbo dome suggests that the granites were derived from melting of the amphibolites at ∼880 °C and ∼10 kbar. The ∼30 Ma granodiorites and ∼42–46 Ma two-mica granites are Na-rich and peraluminous, and are adakitic. They contain inherited Proterozoic zircons, and have a much wider range in ε_Nd(t) of –14.9 to –2.5 and (⁸⁷Sr/⁸⁶Sr)_i of 0.7062–0.7188, and have a Nd isotopic model age of 1486–1978 Ma, indicating that these magmas were derived from a thickened Indian lower crust and were subsequently mixed with amphibolite-derived granite melts or were contaminated by the middle crust under southern Tibet. An apparent northward-younging age trend and shearing of the Eocene–Oligocene granitoids requires the southward migration of slices of middle crustal material, in which the Eocene granitoid magmas were emplaced and stored. Our data, along with structural, metamorphic, and intrusive histories of the Himalaya, lead us to propose a model for crustal anatexis and tectonic evolution of the Himalayan orogen, controlled by a number of large-scale events, such as slab break-off, buoyancy-driven uplift, lateral movement, and subsequent exhumation of slices of the subducted Indian crust during Indo-Asia collision at 55–40 Ma.

Introduction

The Himalaya mountain system, which is the largest active collisional orogen on Earth, was initiated by India–Asia collision at ∼55–50 Ma or earlier (Yin, 2006 and references herein); however, growth mechanisms within the orogen remain uncertain. The Himalaya collisional orogen has been subjected to intense Cenozoic deformation, high-grade metamorphism, and widespread syncollisional anatexis (cf. Le Fort, 1996; Yin, 2006), which has led to many uncertainties about the tectonic evolution of the orogen. More recent debate has focused on mass exchange between the Himalayan (Indian plate) and Tibetan crust (Asian plate) during collision (Yin et al., 2010b, Yin et al., 2010a), and on the origin of the Greater Himalayan crystalline complex (GHC), a high-grade metamorphic unit that forms the core of the Himalaya (Fig. 1A). Four potential models for the origin of the GHC and the formation of the Himalaya have been proposed: (1) Indian basement-involved thick-skinned thrusting (Le Fort, 1975, Yin, 2006, Yin et al., 2010b), (2) an exotic terrane-involved thin-skinned thrusting (DeCelles et al., 2000, DeCelles et al., 2001, DeCelles et al., 2002, Robinson et al., 2006), (3) large-scale horizontal channel flow of the Tibetan middle crust (Nelson et al., 1996, Beaumont et al., 2001, Beaumont et al., 2004, Searle et al., 2003, Godin et al., 2006), and (4) wedge extrusion and exhumation of a slice of subducted Indian crust during India–Asia collision (Chemenda et al., 1995, 1996, 2000).

Crust-derived granitoids can be generated by the distinctive dynamic processes that each of these above models predicts, but the timing, spatial relationships, source region, and dynamic mechanisms involved in crustal anatexis differ among the models (Harris and Massey, 1994, Thompson and Connolly, 1995, Harrison et al., 1997, Searle and Szluc, 2005, Harrison, 2006, King et al., 2007). Thus, constraining the onset and duration of regional melting under southern Tibet, and determining the spatial distribution and source region of anatectic magmas during collision, is a crucial test of the validity of these models. All of these variables can be constrained by studying the granitoids produced during India–Asia collision and that intruded various tectonic units in southern Tibet (Fig. 1).

Miocene granites within the parallel high Himalayan leucogranite (HHL) and the North Himalayan granite (NHG) belts have contrasting ages, petrogenetic histories, and emplacement styles (Harrison et al., 1997 and references therein). In addition, recent studies have reported an Eocene (∼42–44 Ma) regional melting event that formed two-mica granites in southern Tibet (Aikman, 2007, Aikman et al., 2008, Aikman et al., 2012a, Aikman et al., 2012b, Qi et al., 2008, Gao, 2009, Zeng et al., 2011). These studies promote an understanding of crustal anatexis related to continental collision, but a genetic link between crustal anatexis and the growth mechanisms involved in Himalayan orogenesis has not been well established.

In order to explore the relationship between crustal anatexis and Himalayan orogenic processes, and to critically compare the predictions of each of the models mentioned above with direct observations, a NW–SE-striking, 200-km-long tectonic traverse through the Indus–Yarlung suture (IYS), the north Himalayan antiform (NHA), the Tethyan Himalayan thrust–fold belt, and the south Tibetan detachment system terranes (STD; Fig. 1), was undertaken during this study, combined with an analysis of existing data (e.g., Zeng et al., 2011). Along this traverse, representative samples of granitoids, exposed within the IYS, the NHA, and the THS (Fig. 1), were taken for age dating and geochemical analysis to further constrain the temporal and spatial distribution, geochemical variation, and possible origin of these granitoids. This paper presents new ion microprobe U–Pb zircon and biotite ⁴⁰Ar/³⁹Ar dates, major and trace element compositions, and the Sr–Nd isotope systematics of granitoids from this traverse. These results constrain the onset and duration of crust-derived magmatism and the origin of Eocene–Oligocene granitoids of southern Tibet, and provide new insights into the exhumation of high-pressure metamorphic rocks and the development of the Himalayan orogen.