Nature of the Shyok (Northern) Suture Zone between India and Asia: petrology, geochemistry and origin of the Tirit granitoids and associated dykes (Nubra Valley Ladakh Himalaya, NW India)

Abstract

The Shyok Suture Zone is an oceanic remnant of the Neo-Tethyan ocean sandwiched between the Ladakh Batholiths to the south and Karakoram Batholith to the north. The Tirit granitoids in this suture are dark-coloured, relatively rich in ferromagnesian minerals and range from granodiorite–tonalite to gabbro–diorite in composition. Mafic igneous enclaves are quite common and they are intruded by NW–SE parallel doleritic and aplitic dykes. The Tirit granitoids have a wide range of major oxide compositions (SiO₂ = 52.1–72.11 wt %, TiO₂ = 0.21–1.23 wt %, Al₂O₃ = 11.42–13.52 wt %, MgO = 1.69–10.69 wt % and CaO = 3.24–9.31 wt %) and show calc-alkaline, metaluminous, I-type characteristics, transitional between primitive and mature arc continental plutons. Rare earth elements (REE) show considerable enrichment in light REE (LREE) as compared to the heavy REE (HREE). Late Cretaceous U/Pb dates (74–68 Ma) show that they formed during the pre-collision northward movement of India. The Tirit dykes are only slightly younger and probably part of the same episode.

1. Introduction

The age and nature of the crust and the timing of ocean basin closure are keys to understanding the evolution of suture zones (Dewey, 2005; Stern et al. 2012; Draut & Clift, 2013). Critical to developing accurate tectonic reconstructions of such sutures are detailed mapping and analysis of their usually complex and highly deformed rock suites (Dewey, 1977; Frassi et al. 2016; Saktura et al. 2021 b).

The Shyok (Northern) Suture Zone (SSZ) is the northernmost of two sutures that can be traced across the NW Himalayan syntaxis, from the Chaman fault zone to the Karakoram fault zone (Fig. 1), both of which are complex mid- to late Cenozoic collisional transform faults which connect the N-directed thrust of the North Pamir to the S-directed thrusts of the Makran and Southern Tibet (Zhang et al. 2011; Brookfield et al. 2017; Wallis & Searle, 2019). East of the Karakoram fault zone, the southern Indus Suture Zone (ISZ) is usually equated with the Yarlung–Tsangpo suture zone (Hébert et al. 2012), while the SSZ is usually equated with the Bangong suture separating southern and central Tibet (Rolland et al. 2009; Parsons et al. 2020). The SSZ, however, may be equivalent to the Luobadui–Milashan Fault with eclogites and blueschists in eastern Tibet, which separates the southern Lhasa block, with the Gangdise Batholiths, from the central Lhasa block, with a complete Palaeozoic section comparable with that of the Karakoram block (Gaetani, 1997; Zhu et al. 2013; Liu et al. 2017; Yogibekov et al. 2020). The early Cretaceous Bangong suture zone is more likely equivalent to the Rushan–Pshart suture between the Central and South Pamir blocks (Schwab et al. 2004). West of the Chaman Fault Zone, the SSZ may be equivalent to the less compressed Kandahar fore-arc and Ras Koh arc (Kassi et al. 2007; Shroder et al. 2021) (Fig. 1). The nature and equivalence of the SSZ is nevertheless debatable. The SSZ is at altitudes of over 3000 m within even higher mountainous terrain, and is difficult to access in places, not only because roads and tracks are limited, but because it lies in politically disputed zones between Pakistan, India and China (Fig. 2a). Studies have been made in the areas on either side of the India–Pakistan line of control (summarized in Pudsey, 1986; Robertson & Collins, 2002; Rolland et al. 2009; Borneman et al. 2015), but the age and nature of components of the SSZ are still unresolved, as illustrated by the very different and incompatible geological maps published of the area (e.g. Rai, 1982; Dunlap & Wysoczanski, 2002; Saktura et al. 2021 b). Estimates of the age of formation of the SSZ range from Cretaceous (Weinberg et al. 2000; Rolland et al. 2006; Rehman et al. 2011) to Eocene (Khan et al. 2009).

Fig. 1. Geological map of Ladakh, showing location of Figure 2 (from Jain, 2014). Inset shows location of map on terrane map of India–Asia collision zone (from Parsons et al. 2020). CF – Chaman fault; KF – Karakoram fault; Ko – Kohistan arc; Ka – Karakoram.

Fig. 2. (a) Satellite view of study area. (b) Geological map of study area, modified from Borneman et al. (2015), with changes from Saktura et al. (2021 b) and authors’ field observations, showing location of samples; cross-section modified from Upadhyay et al. (1999) and Borneman et al. (2015).

Detailed study of the individual components of the SSZ is needed in order to resolve its nature, development and age. We here discuss the petrology, geochemistry, age and nature of the Tirit granitoids and associated dykes at the Nubra–Shyok confluence, using a map based on Borneman et al. (2015) and our own extensive fieldwork in the area (Fig. 2).

2. Geological overview

In the Nubra–Shyok confluence area, the SSZ forms a narrow highly tectonized belt, overthrust in the south by the Ladakh batholith and in the north by the Karakoram batholith (Fig. 2b, c; Rai, 1983; Upadhyay et al. 1999; Rolland et al. 2000; Borneman et al. 2015). The dominant rocks are ophiolitic fragments, basic to acid volcanics and intrusives, pelagic oceanic sediments, cherts and foraminiferal limestones similar to those of the ISZ to the south, with large olistoliths of Asian shelf sediments in places (Brookfield & Gupta, 1984; Rolland et al. 2000; Robertson & Collins, 2002; Upadhyay, 2014; Upadhyay et al. 2014). All these have been highly tectonized and are unconformably overlain in Ladakh by coarse continental clastics of post-Oligocene age (Rai, 1982).

The SSZ has various and divergent interpretations as: an oceanic suture (Gansser, 1977); a subduction zone, older than (Petterson & Windley, 1985) or younger than (Brookfield & Reynolds, 1981; Reynolds et al. 1983) the ISZ; a tectonic repetition of the ISZ (Rai, 1982, 1983; Srimal, 1986); a back-arc complex related to the ISZ (Thakur & Misra, 1984; Dunlap & Wysoczanski, 2002); and a marginal basin (Upadhyay, 2002). A popular view is that the SSZ is a site of probably Early to Late Cretaceous Ocean closure, while the ISZ is the younger main suture between the Indian continental plate and Asian continental plate (Srimal, 1986; Upadhyay et al. 1999; Rolland et al. 2000; Robertson & Collins, 2002; Bhutani et al. 2009; Upadhyay, 2009; Borneman et al. 2015; Kumar et al. 2016).).

South of the SSZ, the Ladakh block is mostly calc-alkaline batholiths and their extrusive equivalents (the Khardung Volcanics), with pendants and enclaves of pre-intrusive country rocks that include the Changmar and Shoyk Volcanics described below (Weinberg et al. 2000; Rolland et al. 2002 b; Thanh et al. 2010; Saktura et al. 2021 a). The Ladakh batholiths consist mostly of coarse- to fine-grained granite to diorite intrusions with roof pendants and enclaves of ophiolites (Reuber, 1990; Rolland et al. 2000), coarse mafic and acid intrusives (Kumar, 2010) and sedimentary rocks (Raz & Honegger, 1989). U/Pb zircon dates indicate that magmatism was concentrated during a 20 million year period from ∼65 Ma to ∼45 Ma (St-Onge et al. 2010; White et al. 2011). More basic intrusives on the north range give U/Pb of ∼66 to 60 Ma, whereas the main granite intrusions give a U/Pb date of ∼50 Ma (St-Onge et al. 2010; Thanh et al. 2010; Shellnutt et al. 2014). These correspond to two main episodes of magmatism, with a change in composition from I-type to S-type granites and adakaites, attributed to crustal thickening during the ‘hard’ collision between Indian and Asian continental crust (Shellnutt et al. 2014).

On the northern side of the Ladakh Range, the Khardung Volcanics overlie, are intruded by and are overthrust by the Ladakh batholith (Upadhyay et al. 1999; Rolland et al. 2000; Weinberg et al. 2000). The Khardung Volcanics consist of basalts, andesites, dacites, rhyodacites and rhyolites with associated pyroclastics, and become more acidic upwards (Upahdyay, 2014). The lower andesites (U/Pb zircon date of ∼70 Ma) switch to rhyolites (U/Pb zircon dates of ∼65 ± 2 Ma), and these persist until ∼52 Ma (Lakhan et al. 2020; Saktura et al. 2021 a), with Ar/Ar whole-rock dates of ∼55 Ma (Bhutani et al. 2009; White et al. 2011). Like the Ladakh batholith, the volcanics become more acidic upwards and are thus likely co-magmatic with it, as they have comparable ages.

From south to north, the units in the Shyok belt are as follows:

1. 1. The Changmar complex consists of norites, gabbronorites, plagiogranites, harzburgites and serpentinites (Saktura et al. 2021 b). It is intruded by the Ladakh Batholith, intrudes the Shyok Volcanics to the north and possibly forms the oceanic substrate to the Shyok volcanic arc (Rolland et al. 2000; Saktura et al. 2021 b).

2. 2. The Shyok Volcanics consist of calc-alkaline basaltic to andesitic volcanics metamorphosed to greenschist facies (Weinberg et al. 2000; Thanh et al. 2012; Sivaprabha et al. 2022). Along the Shyok River, the unit is intruded by gabbroic dykes probably related to the Changmar conplex (Saktura et al. 2021 b). One of the dykes gave a Lower Cretaceous ⁴⁰Ar/³⁹Ar hornblende age of 125.6 ± 6.1 Ma (Aptian) (Borneman et al. 2015).

3. 3. The Saltoro Formation is overthrust by the Kharduing Volcanics south of Khalsar, but may rest unconformably on the Shyok Volcanics to the northwest (Upadhyay et al. 1999; Borneman et al. 2015) (Fig. 2). The lower part, sometimes separated out as the Tsoltak Formation, consists of deeper water thinly and mostly even-bedded, highly fissile and cleaved slates, phyllites, and siltstones, with intercalations of thinly to medium-bedded grey fossiliferous limestones and marbles, with possibly late Jurassic bryozoa (Upadhyay et al. 1999). A tectonically isolated thick recrystallized limestone, called the Hundiri Formation, contains abundant mollusc and Lower Cretaceous (Aptian – early Albian) foraminifera (Juyal, 2006). To the southeast, the lower part of the formation contains Middle Jurassic (Callovian) ammonites (Ehiro et al. 2007). The upper part of the Shyok Formation consists of shallower-water sandstones and mudstones with bivalves and gastropods passing up into Lower Cretaceous (Albian) limestone conglomerates and acidic tuffs (Matsumaru et al. 2006). The age of the Saltoro Formation overlaps with that of the Shyok Volcanics and may be a back-arc basin to a Shyok Volcanic oceanic arc (Upadhyay et al. 1999).

4. 4. The Saltoro Molasse consists of coarse- to fine-grained continental clastic sediments, at least 3.2 km thick, unconformable on, and overthrust onto, the Shyok Volcanics and Saltoro Formation (Rai, 1983). The rounded fluvial orthoconglomerate pebbles range from 0.1 to 1 m, and are dominated by Shyok Volcanic greenstones, with subordinate granite, phyllites and schists (Rai, 1983). The molasse is a typical deposit of near-source alluvial fans (Blair & McPherson, 2009).

5. 5. An ophiolitic mélange is the topmost unit within the Shyok suture zone, overthrusts the Saltoro Molasse and is overthrust by the Karakoram block (Weinberg et al. 2000). It is c. 1 km thick and contains decimetre- to kilometre-scale blocks of phyllite, limestone, red chert, basalt, gabbro and peridotite (Rai, 1983).

Numerous felsic to mafic dykes also cut all SSZ units, except the ophiolitic mélange, and thus predate the southward thrusting of the Karakoram block. Some individual dykes can be traced over 1 km in the field. The felsic to intermediate composition dykes typically have an aplitic texture and are often heavily altered and crumbly in hand specimen.

1. 6. The Karakoram block overthrusts the ophiolitic mélange, and consists of metamorphosed Palaeozoic to Upper Cretaceous host rocks (Rai, 1983; Sinha et al. 1999; Rolland et al. 2002 a) intruded by Lower Cretaceous (110–100 Ma) calc-alkaline and Upper Cretaceous (85 Ma) alkaline granitoids. This Cretaceous magmatism is attributed to the development of an Andean arc on the southern edge of Asia (Rolland et al. 2002 b). Younger Cenozoic collisional magmatism and metamorphism occurred in several phases during the hard collision of India and Asia (Rolland et al. 2001), with Oligocene to Miocene (26 to 21 Ma) leucogranites intruded during thrusting of the Karakoram block over the SSZ and right-lateral displacement on the Karakoram Fault (Schärer et al. 1990; Allen & Chamberlain, 1991; Debon & Khan, 1996; Rolland et al. 2001, 2006, 2009; Horton & Leech, 2013; Brookfield et al. 2017; Pundir et al. 2020).

3. Tirit granitoids and associated dykes

Tirit granitoid and dyke samples were collected at outcrops at Khalsar (KS), Tegart (T), Diskit (DM), Tirit Bridge (TB) and Sati Bridge (SB) (Fig. 2).

3.a. Tirit granitoids

The Tirit granitoids outcrop in four areas: around Tirit, north and south of Diskit and northeast of Hundar. They intrude the Saltoro Formation, and are faulted against the Karakoram block at the Karakoram Fault Zone east of Tirit (Figs 3 and 4a). These outcrops were originally a single plutonic body and are now isolated by a cover of Quaternary alluvium (Rao & Rai, 2009). The granitoids are dark-coloured, relatively rich in ferromagnesian minerals and range from granodiorite–tonalite to gabbro–diorite in composition. Mafic sub-rounded to rounded igneous enclaves up to 30 cm in diameter are quite common in these granitoids and possibly derive from the Shyok Volcanics (Fig. 4).

Fig. 3. View up Nubra valley from Diskit showing eastern outcrops of Tirit granitoids.

Fig. 4. Field photographs showing (a) intrusive contact of Tirit granitoids with Shyok Volcanics near Tirit Village Nubra Valley; (b) Tirit granitoids relatively rich in ferromagnesian minerals; (c) diffuse mafic enclaves in Tirit granitoids at Tirit Bridge, Nubra Valley; (d) xenoliths in Tirit granitoids at Tegar, Nubra Valley; (e) mafic and felsic dykes in Tirit granitoids near Tirit Bridge; (f) displacement of mafic enclave in Tirit granitoids near Tirit Bridge.

3.b. Dykes

Mafic dykes intrude the Tirit granitoids east of Tirit at Sati Bridge (SB) (Fig. 2), The NE–SW-trending parallel mafic dykes are doleritic (Fig. 4e), with thicknesses varying between 30 cm and 1.5 m, and displaced up to 0.5 m by post-emplacement faulting (Fig. 4f). They show no evidence of chilling against the walls of granitoids, indicating that the host rocks were still hot when the dykes intruded. They are comparable in orientation and possibly age, but not in composition, to E–W- and NE–SW-striking late-stage andesitic dykes of the Ladakh batholith (∼45 Ma) west of Leh (Heri et al. 2015).

3.c. Petrography

3.c.1. Tirit granitoids

Major minerals are plagioclase, K-feldspar, quartz, biotite and hornblende, with zircon and magnetite as accessory minerals (Fig. 5). Point counts show that: the granites have 24–27 % K-feldspar, 5–7 % plagioclase and 14–22 % quartz; the granodiorites have 9–34 % K-feldspar, 11–19 % plagioclase and 15–29 % quartz; the trondhjomites have 20–29 % K-feldspar, 29–35 % plagioclase and 2–16 % quartz; and the quartz–diorites have 14–23 % K feldspar, 14–16 % plagioclase, and 20–24 % quartz. Euhedral plagioclase laths within subhedral grains of K-feldspar and quartz signify early increments of growth (Fig. 5a), followed by later development of K-feldspar and quartz grains which show graphic intergrowths (Fig. 5b). Quartz grains have undulose extinction and cataclastic textures in places (Fig. 5c). Some biotite grains show chloritization (Fig. 5d), and some plagioclase grains show sericitization (Fig. 5e); while both show marginal epidotization in places (Fig. 5f), indicating hydrothermal alteration.

Fig. 5. Photomicrographs of Tirit granitoids showing (a) plagioclase laths enclosed within subhedral grains of K-feldspar and quartz (PPL); (b) plagioclase with euhedral crystal faces in contact with K-feldspar and quartz grains (XPL); (c) plagioclase mineral grains exhibiting simple and polysynthetic twinning; (d) epidotization at the margins of plagioclase and biotite mineral grains (XPL); (e) alteration of biotite into green chlorite and the sericitization of plagioclase (XPL); and (f) epidotization of biotite, alteration of plagioclase and zircon inclusions (XPL). Kfs: K-feldspar; Qtz: quartz; Bt: biotite; Zr: zircon; Pl: plagioclase; Cl: chlorite; Ms: muscovite; Mc: microcline; Ap: apatite; Ti: titanite; Hb: hornblende.

3.c.2. Dykes

The mafic dykes are dark grey to black, and fine- to medium-grained in texture. Major minerals are plagioclase, clinopyroxene and magnetite, with accessory amounts of chlorite and opaque mineral phases. They commonly exhibit porphyritic texture with plagioclase and clinopyroxene phenocrysts embedded in fine-grained groundmass of plagioclase, clinopyroxene, chlorite and glass (Fig. 6a). Plagioclase crystals show a preferred orientation due to magmatic flow showing trachytic texture. A few clinopyroxene crystals show alteration into chlorite along grain boundaries and fracture planes (Fig. 6b). Opaque minerals occur as small discrete grains within the fine-grained plagioclase and clinopyroxene groundmass (Fig. 6c).

Fig. 6. Photomicrographs of Tirit dykes showing (a) plagioclase phenocrysts embedded in fine-grained groundmass of clinopyroxene, chlorite and glass (XPL); (b) plagioclase grains with a preferred orientation due to magmatic flow exhibiting trachytic texture (XPL); (c) opaque minerals within the fine-grained groundmass of plagioclase and clinopyroxene. Pl: plagioclase; Mg: magnetite; Op: opaques; Cpx: clinopyroxene.

4. Geochemistry

4.a. Methods

Thirteen representative samples of Tirit granitoids and eight samples of mafic dykes were analysed for major, trace and rare earth elements (REE) at the National Geophysical Research Institute (NGRI), Hyderabad, India. Major elements were determined by X-ray fluorescence (XRF) spectrometer on powder pellets using a Philips Model PW-2440 Spectrometer. United States Geological Survey (USGS) G-2 geological rock standard was used for calibration and quantitative estimation of chemical elements. Trace elements including REE and high-field-strength elements (HFSE) were determined by high-resolution inductively coupled plasma mass spectrometer (HR-ICP-MS). Reference materials from the Geological Survey of Japan (JB-2, JB-3, JB-1a) and USGS (BHVO-1, BCR-1, BIR-1) along with a couple of procedural blanks were also analysed, with the sample batches as controls on accuracy.

Whole-rock major elemental analyses of dykes were carried out by XRF spectrometery techniques. Trace elements including REE were determined by ICP-MS techniques at Wadia Institute of Himalayan Geology (WIHG), Dehradun, Uttarakhand, India.

4.b. Results and interpretation

4.b.1 Tirit granitoids

Geochemical data for the Tirit granitoids are provided in Supplementary Table S1, and CIPW norms in Table 1. The Tirit granitoids have a wide range of chemical compositions, with SiO₂ = 50.91–72.11 wt %, TiO₂ = 0.21–1.23 wt %, Al₂O₃ = 11.42–14.00 wt %, MgO = 1.69–10.69 wt % and CaO = 3.24–9.31 wt % (Supplementary Table S1). The high content of certain oxides like MgO and Al₂O₃ suggests the mafic components in the source region. Major elements exhibit linear trends against SiO₂, confirming the role of magmatic differentiation in their evolution. Harker Variation plots show a positive trend of SiO₂ with Na₂O and Al₂O₃ and negative trend with MgO, FeO^(t), TiO₂ and P₂O₅ (Supplementary Fig. S1). The positive trend of Na₂O with silica reflects increasing modal plagioclase and K-feldspar. However, negative variation trends of MgO, TiO₂ and P₂O₅ indicate the fractional crystallization of hornblende, magnetite, titanite and apatite. Harker Variation trace element plots against SiO₂ show decreasing Ni, Ta, Pb and Sr but no correlation of Rb, Ba, Y and Zr (Supplementary Fig. S2), which may be due to variable amounts of minerals rich in these elements (Pearce & Norry, 1979). The Tirit granitoids fall within the granite, granodiorite and trondhjomite fields on a normative albite (ab) – anorthite (an) – orthoclase (or) plot (Fig. 7a); on the calc-alkaline differentiation trend on an AFM ternary plot (Fig. 7b); in the metaluminous field on A/CNK and A/NK plots (Fig. 7c); and are I-type on the Chappell & White (1974) plot (Fig. 7d). They show similar patterns on a bulk rock REE normalization to chondrite (Boynton, 1984), although with differences in abundance, showing considerable enrichment in the light REE (LREE; (La/Sm)N = 2.20–6.53) compared to the heavy REE (HREE; Gd/Yb)N = 0.86–1.91) (Fig. 8).

Table 1. Tirit granitoids CIPW norms

Fig. 7. Tirit granitoids: (a) An–Ab–Or classification diagram (after O’Connor, 1965); (b) AFM diagram (after Irvine & Baragar, 1971); (c) A/CNK versus SiO₂ wt % diagram (after White & Chappell, 1977); (d) molar A/CNK versus A/NK plots showing their metaluminous nature (after Shand, 1943).

Fig. 8. Rare earth elements (REE) chondrite-normalized pattern diagrams of Tirit granitoids (a) and Tirit dykes (b). Chondrite values of Boynton (1984).

4.b.2. Tirit dykes

Geochemical data for the Tirit dykes are reported in Supplementary Table S2., and CIPW norms in Table 2. The dykes have a low but wide range of SiO₂ (48.95 to 57.49 %), are enriched in Al₂O₃ % (14.62 to 16.98 %), and have a markedly low TiO₂ content (0.64 to 1.11 %). MgO ranges from 5.77 to 10.31 wt % and total iron FeO^(t) from 8.36 to 10.01 %. Na₂O ranges from 0.44 to 3.44 wt %. The K₂O ranges from 1.17 to 5.03 wt %.

Table 2. Tirit dykes CIPW norms

During low-grade metamorphism, Zr (ppm) is considered to be relatively immobile (Winchester & Floyd, 1977; Macdonald et al. 1988). Hence, Zr can be used as a parameter for evaluating the elemental mobility and also to understand the differentiation/fractional crystallization of magma. In order to assess the fractional crystallization and also the mobility of major elements of Tirit dykes during the post-crystallization processes, these elements has been plotted against Zr. Most of the major elements including FeO^(t), MgO, MnO, CaO and TiO₂ show a negative relationship with Zr. The SiO₂ wt %, Al₂O₃ and P₂O₅ wt % show a positive correlation with Zr. The normal negative and positive relationships probably indicate the primary magmatic characteristics for these major elements. However, the scattering of Na₂O wt % and K₂O wt % against the Zr indicate the mobile nature of Na₂O wt % and K₂O wt % during the post-crystallization processes (Supplementary Fig. S3). Binary plots of trace elements against Zr display positive trends with Nb, Th, Y and Sr and negative trends with Rb, Cr, Ni and U (Supplementary Fig. S4). In volcanic suites, the Nb/Y ratio is an indicator of alkalinity and the Zr/Ti ratio is a differentiation index (DI) (Pearce & Cann, 1973). When plotted together they define compositional fields (Winchester & Floyd, 1977), in which the Tirit dykes fall in the andesite – basaltic-andesite field (Fig. 9a). They are calc-alkaline on a (Na₂O + K₂O versus SiO₂ plot (Fig. 9b). They are moderately enriched in LREE ((La/Sm)N = 2.76–2.87) and show relatively flat HREE ((Gd/Yb)N = 1.76–2.05) (Fig. 9c).

Fig. 9. (a) Zr/Ti versus Nb/Y plot on immobile elements classification diagram for Tirit granitoids and dykes showing andesite – basaltic-andesite composition (after Winchester & Floyd, 1977). (b) Na₂O + K₂O wt % versus SiO₂ wt % plot for Tirit granitoids and dykes, showing their calc-alkaline nature (after Kuno, 1968). (c) Rare earth elements (REE) pattern for Tirit dykes showing enriched LREE and relatively depleted HREE (chondrite-normalizing values of Sun & McDonough, 1989).

4.c. Petrogenesis

The Tirit granitoids show a continuous trend from mafic-rich (trondhjomite) to felsic-rich (granodiorite and quartz–diorite), with high Na₂O (av. 4.22) and low K₂O/Na₂O (av. 0.70), average molar A/CNK of 0.67 and A/NK of 1.36 (Table 1). Their metaluminous characteristics, high contents of TiO₂, Al₂O₃, MgO and high Sr and Ni indicate mafic components in the source region (Chappell & Stephens, 1988). Decrease in CaO % in rocks from low to high SiO₂ content is a result of crystal fractionation of plagioclase and calcic amphiboles. Zircon content and the regular pattern of decreasing P₂O₅ with increasing SiO₂ are characteristic of mafic low-temperature I-type granitoids (Chappell & White, 2001; Chappell et al., 2004), whose petrogenesis is commonly attributed to variable interactions of mantle-derived mafic magmas with the continental crust, or to simple remagmatization of older crust with subsequent fractional crystallization and/or restite crystal fractionation (Chappell, 1996). The negative Eu/Eu* (0.59–0.91) anomaly in these granitoids suggests the removal of plagioclase feldspar from the source magma by fractional crystallization or residual feldspar in the source.

HFSE are least mobile, but show moderate degrees of fractional crystallization and are sensitive to partial melting and source inhomogeneities (Ahmad & Tarney, 1991). The ratios of Th/La, Th/Nb, Zr/Nb and Ce/Nd reveal the source characteristics and are higher in Tirit dykes than those of the primitive-mantle ratios (Sun & McDonough, 1989), indicating their derivation from enriched mantle sources (Table 2). The occasional higher Sr in these dykes suggests the fractionation of calcic plagioclase (Sun & McDonough, 1989). The moderate degrees of LREE enrichment, and flat HREE of the dykes (Fig. 9c) are typically shown by arc-related basalts and indicate an origin in a shallow (spinel–lherzolite) mantle (Murphy, 2007).

4.d. Tectonic setting

Maniar and Piccoli (1989) distinguished seven types of granitoids based on their tectonic setting. However, the Tirit granitoids show divergent settings. On the tectonic discrimination parameters R1 (= [4Si − 11(Na + K) − 2(Fe + Ti)]) and R2 (= [6Ca + 2Mg + Al]), most of the Tirit granitoids samples are in the post-collisional uplift, syn-collision and pre-plate collision fields (Fig. 10a). On the relative trace element Y + Nb versus Rb plot, however, they are in the Volcanic Arc Granites (VAG) field (Fig. 10b). Furthermore, on the Rb/Zr ratio against Nb and Y ppm plots of Brown et al. (1984), the samples are scattered and show no pattern (Fig. 10c).

Fig. 10. (a) Multi-cationic R1 versus R2 tectonic discrimination diagram for Tirit granitoids of Nubra–Shyok Valley showing pre-plate collision and post-collision uplift of Tirit granitoids (after Batchelor & Bowden, 1985); (b) Rb versus Y + Nb tectonic discriminant diagrams for Tirit granitoids and dykes (after Pearce et al. 1984); (c) Rb/Zr versus Nb and Y for Tirit granitoids showing transitional nature (after Brown et al. 1984).

Primitive-mantle-normalized spider diagrams for Tirit granitoids show negative anomalies of Nb, Ti, Zr, Ba and P along with positive anomalies of Th, U, K, Nd, Sm and Pb (Fig. 11), which suggests that they were derived from partial melting of juvenile crustal material possibly within a supra-subduction setting (Pearce et al. 1984; Qi et al. 2014). Fluids and melts derived from the subducted oceanic slab led to metasomatism of the upper mantle wedge and negative anomalies of Nb (Cox et al. 1979; Chappell, 1999). The positive anomalies in Pb and K, however, are attributed to metasomatism of mantle wedge by fluids derived from the subducted slab and/or contamination with continental crust (Kamber et al. 2002).

Fig. 11. (a) Hf–Tb–Ta plot of Tirit granitoids, showing dominantly in calc-alkaline field and extreme Hf and Ta depletion; (b) Y–La–Nb plot of Tirti granitoids and dykes showing dominantly calc-alkaline nature.

On the ternary 2Nb – Zr/4 – Y plot of Meschede (1986), the Tirit dykes scatter across several fields (Fig. 12a), while on the binary Zr/Y – Zr plot of Pearce (1983) they plot firmly in the continental arc basalts field, with higher Zr/Y than oceanic basalts (Fig. 12b). The primitive-mantle-normalized plot of the dyke elements is similar to the Tirit granitoid plot, and shows enrichment in U and Pb, and depletion in Nb and Ti (Fig. 13a), with the the negative Nb anomaly in both granitoids and dykes caused by subduction-related enrichment of lithospheric mantle (Kepezhinskas et al. 1997).

Fig. 12. Multi-element spider plots of Tirit granitoids and dykes showing negative anomalies of Rb, Ba, Nb, Pr, P, Zr, Ba and P along with positive anomalies of U, K, Pb, Nd and Sm (normalized values of Sun & McDonough, 1989).

Fig. 13. (a) 2Nb – Zr/4 – Y ternary plot for Tirit dykes in dominantly volcanic arc basalt and within-plate tholeiite fields (after Pearce, 1983). (b) Zr (ppm) versus Zr/Y plot for Tirit granitoids and dykes: oceanic arc for granitoids, continental arc for dykes (after Pearce, 1983).

5. Geochronology

Previous U/Pb ion-microprobe ages from a tonalite at Tirit range from 66.8 ± 2.0 to 70.7 ± 0.8 Ma (11 spots; 6 slightly younger ages are considered altered) with a weighted mean of 68 + 1 Ma (Weinberg et al. 2000). A best-fit intercept U/Pb date from a granodiorite at Tirit gave 71.40 ± 0.36 Ma (Upadhyay, 2008); an Ar/Ar integrated total gas cooling age on hornblende from a diorite north of Khalsar was 73.6 ± 2.0 Ma, which is statistically the same as the age spectrum date of 73.1 ± 2.2 Ma (Weinberg et al. 2000); and an albite porphyroblast from a greenschist NW of Tirit gave a date of 65.4 ± 3.3 Ma, though abundant phengite inclusions may have affected the date (Thanh et al. 2010). Zircons from a granite at Hundar – supposedly from the Ladkah batholith, but probably from enclaves within it (Kumar, 2020) – gave a weighted mean average (of three dates) of 66.6 ± 2.1 Ma (Upadhyay et al. 2008). All errors have been increased, where necessary, from cited 1σ to 2σ, in keeping with normal practice, as 1σ gives an unacceptable 1-in-3 chance of being wrong, while 2σ gives a more acceptable 1-in-20 chance (Wyatt et al. 1998).

We analysed zircons in a tonalite (TB3) from Tirit Bridge (TB3) locality, using LA-ICP-MS with methods described in Liu & Stockli (2020) (Table 3; Fig. 13b). The most concordant zircon gives an age of 69.24 ± 0.96 Ma, with a lower intercept age for the 26 acceptable zircons of 64 ± 14 Ma, and MSWD = 4.6.

Table 3. LA-ICP-MS geochronology, sample TB3 of Tirit granitoids

The dates are consistent with intrusion of the more mafic Tirit granitoids being somewhat older than the more felsic granitoids, spanning ∼74 to 66 Ma. The dykes are younger than this.

6. Tectonic significance

The northward motion of India from the Mid Cretaceous to Eocene was accommodated by northward subduction of Neo-Tethys lithosphere along the ISZ and SSZ (Coward et al. 1986; Debon et al. 1987; Heuberger et al. 2007; Kumar et al. 2017), sporadically producing very large volumes of diversified granitoids. The Tirit granitoids were intruded during a relatively short episode of latest Cretaceous island arc development that is contemporary with the Spong island arc of the Indus suture zone to the south (Catlos et al. 2018) prior to the latest Cretaceous obduction of the Spongtang and other ophiolite nappes onto the northern Indian margin in the middle of the Tethys Ocean (Corfield et al. 2001; Gibbons et al. 2015). In that case a Palaeocene – early Eocene magmatic arc is required to obliterate the oceanic lithosphere between northern India and Asia, and this might be represented by the calc-alkaline, island arc plutonic bodies of the Leh pluton (60 ± 5 Ma), the Chang La pluton (57.6 ± 1.4 Ma to 53.4 ± 1.8 Ma) (Debon et al. 1987; Upadhyay et al. 2008) and the Palaeocene Ras Koh arc in SW Pakistan (Nicholson et al. 2010) (Fig. 14). The final ‘hard’ collision of India and Asia then occurred in the early Eocene between 50 and 45 Ma (Rex et al. 1988; Rolland, 2002; Upadhyay, 2008; Khan et al. 2009; Jain, 2014), following which further northward movement of the Indian Plate and compression and uplift formed the Himalaya (Fig. 15). The younger mafic dykes trending NW–SE in the Tirit granitoids may be related to the activity on the strike-slip Karakoram fault. Evaluating the complex and competing versions of this history, however, requires a more detailed (and expanded) study (Andjić et al. 2022).

Fig. 14. Concordia ²⁰⁷Pb/²³⁵U versus ²⁰⁶Pb/²³⁸U diagram for zircons from TB3 sample of the Tirit granitoids.

Fig. 15. Diagram showing (a) 65–45 Ma emplacement of Tirit granitoids during initial collision of extinct Ladakh arc with Karakoram Andean arc and development of Palaeocene arc to south, followed by (b) 45 Ma – present hard collision of India with Asia and partial melting of subducted Indian slab to give Ladakh batholith and associated volcanics with emplacement of the Tirit dykes.

7. Conclusion

The integrated field, petrographical, geochemical and geochronological data indicate that the calc-alkaline Tirit granitoids were emplaced between ∼71 and 58 Ma, during the later stages of an Upper Cretaceous to Palaeocene arc. They were further deformed during the India–Asia collision along the Indus Suture Zone around 45 Ma. The younger Tirit dykes, with high Th/La, Th/Nb, Zr/Nb and Ce/Nd ratios, have subducted components, inherited from the subcontinental mantle lithosphere in a post-collisional setting, and could also be related to the development of the NW–SE Karakoram fault.