Introduction

Oceanic basalt is the most widespread terrestrial volcanic rock and occurs in a range of tectonic settings such as mid-ocean ridges, intraplate islands, oceanic plateaus, island arcs and back-arc basins (Batiza and White 2000; White et al. 2015; Perfit and Soule 2016). Volcanic processes currently forming new oceanic crust can be directly observed using devices such as manned submersibles. This has the advantages of being able to sample fresh and unaltered basalts and to map the areal distribution and surface morphology of the lavas, with both the eruption depth and tectonic setting well constrained (White et al. 2015). In contrast, neither eruption depth nor tectonic setting is directly known for ancient oceanic crust, preserved in ophiolites which are slices of obducted oceanic lithosphere found in continental orogenic belts (Coleman 1977; Dilek 2003). Both parameters are important for understanding the geological histories of orogenic belts and for plate tectonic reconstructions.

Basalt geochemistry and the so-called tectonic discrimination diagrams (e.g. Pearce and Cann 1971; Shervais 1982; Verma 2010; Dilek and Furnes 2011; Pearce 2014) have been the most widely used tools for inferring the palaeotectonic settings of ophiolites (e.g. Beccaluva et al. 1994; Ghazi et al. 2004; Kakar et al. 2014). However, considerable submarine hydrothermal alteration and even metamorphism (in the greenschist to amphibolite facies) are commonly seen in ophiolites. The primary chemical and isotopic compositions of the basalts, including abundances of even the relatively alteration-resistant elements (such as Nb, Zr and Ti), may have been significantly modified. Additionally, modern basalt lavas erupted in very different tectonic settings, specifically ocean island basalts and continental intraplate and rift basalts, have closely similar geochemical compositions (Fitton 2007; Sheth 2008). Geochemical plots proposed to serve as tectonic discrimination frequently do not achieve the purpose (Li et al. 2015), and Xia and Li (2019) emphasise that these diagrams cannot be used in isolation. Thus, it is arguable whether geochemical plots can indicate tectonic setting, because igneous rock compositions are strongly determined by sources and processes and not by tectonic setting (Rollinson 1993; Sheth et al. 2002).

Here, we ask whether volcanic facies can provide clues to the palaeodepth and palaeotectonic setting of an ophiolite, and serve as an independent check on geochemistry-based interpretations. Our chosen case study, the Nidar ophiolite, is exposed > 4100 m above mean sea level in the Ladakh region of the Indian Trans-Himalaya (Fig. 1). It is an ~ 8-km-thick sequence of ~ 140 Ma Neo-Tethyan oceanic lithosphere. Its mantle section is relatively well studied, but the volcanic section is poorly described. We describe the salient features of the volcanic section and discuss how the combined volcanological features are a guide to the environment of formation of this spectacular slice of oceanic lithosphere.

Fig. 1
figure 1

Landscape of the Nidar ophiolite in summer (September 2019). View is approximately towards south

Full size image

Regional geology

The Indus–Tsangpo suture of southern Tibet marks the collision zone between India and Asia (e.g. Gansser 1980; Virdi 1986; Thakur 1990). It is marked by several volcanic and ophiolite belts thought to represent intra-oceanic island arcs in the Neo-Tethys Ocean (e.g. Honegger et al. 1982; Dietrich et al. 1983; Aitchison et al. 2000). The westward extension of the Indus–Tsangpo suture into the Ladakh region of the Indian Trans-Himalaya is represented in the Dras, Spontang and Nidar ophiolites (Thakur and Misra 1984; Corfield et al. 2001; Mahéo et al. 2004) (Fig. 2).

Fig. 2
figure 2

(Inset) Map showing the Indian subcontinent, Tibet, and the Himalaya mountains in between, with the small box showing the part of the Ladakh region of the Indian Trans-Himalaya enlarged in the main figure. The main figure is a geological map of Ladakh, modified after Virdi (1986) and Mahéo et al. (2004), showing the major lithological and stratigraphic units and the localities mentioned in the text. In the map legend, the units are listed from north to south. The Indus River and various high-altitude lakes (including Kyun Tso, Tso Morari and Pangong Tso) are shown in blue

Full size image

The Nidar ophiolite, the focus of the present study, is exposed in southeastern Ladakh in a NW–SE-elongated outcrop (the regional structural trend of the Himalaya) just south of the Indus (Sindhu) River. The river flows along the suture between the Indian and the Asian plates (Fig. 2). Just north of the suture is the 110–50 Ma, calc-alkaline, Ladakh granite batholith and the Dras volcanic arc, which represent subduction of the Neo-Tethyan ocean floor under the Asian active continental margin, followed by the mid-Eocene India–Asia collision and underthrusting of the northern Indian continental margin under Asia (Honegger et al. 1982; Dietrich et al. 1983; Bhutani et al. 2004; Kumar et al. 2016). The Nidar ophiolite is thrust over the Zildat melange, which contains basalts, argillaceous rocks and “exotic” limestone blocks (Colchen 1999). South of the ophiolite and the melange, the ultrahigh-pressure Tso Morari gneisses and eclogites of the northern Indian continental margin are exposed, surrounded by the Mata unit composed of Palaeozoic, greenschist-facies metasedimentary rocks (Mukherjee and Sachan 2001; de Sigoyer et al. 2004). Further southwards, the Zanskar unit or Tethyan sequence is composed of Palaeozoic and Mesozoic continental shelf sediments of the northern Indian continental margin, and the Higher Himalayan crystallines represent Precambrian rocks of the Indian shield forming the central axis of the mountain range (Gansser 1964).

The ~ 8-km-thick Nidar ophiolite sequence consists of cherts, pillow lavas and gabbros (intruded by plagiogranite dykes) forming the crustal section, and peridotites constituting the mantle section (Thakur and Bhat 1983; Rameshwar Rao et al. 2004; Das et al. 2015; Buchs and Epard 2015, 2019). The ophiolite became highly dismembered during obduction, as pillow lavas of the ophiolite are also found at Mahe 20 km northwest of Nidar at 4178 m elevation, and halfway between Nyoma and Mahe at 4150 m (Fig. 2). There has been considerable interest in the mantle peridotites, which are harzburgites and dunites, partly serpentinised and frequently chromite-bearing (e.g. Ravikant et al. 2004; Duraiswami et al. 2014; Das et al. 2015; Ray et al. 2017; Nayak and Maibam 2020). The gabbros have experienced ocean floor metamorphism to greenschist and amphibolite facies, and Mahéo et al. (2004) obtained 40Ar/39Ar cooling ages of 130–110 Ma on gabbros with metamorphic amphiboles. The basalts are also hydrothermally altered, and their vesicles are filled with chlorite, carbonates and quartz.

Linner et al. (2001) obtained a Sm–Nd isochron age of 140.5 ± 5.3 Ma on a Nidar gabbro, but this is only a two-point (plagioclase–clinopyroxene) isochron. Using Sr–Nd isotopic data on gabbros and basalts, they considered that the Nidar ophiolite formed in a marginal basin. Based on geochemical and Sr–Nd isotopic data for mafic and ultramafic rocks, Mahéo et al. (2004) and Ahmad et al. (2008) inferred a depleted mantle source metasomatised by subduction-derived fluids. Ahmad et al. (2008) obtained an imprecise mineral–whole rock Sm–Nd isochron age of 140 ± 32 Ma for the Nidar gabbros, whereas radiolarian cherts that form the sedimentary cover of the Nidar volcanic sequence have been dated biostratigraphically to 132–112 Ma (Thakur and Virdi 1979; Kojima et al. 2001; Zyabrev et al. 2008). These age constraints and geochemical arguments have led to the inference that the Nidar ophiolite represents a slice of Neo-Tethyan oceanic lithosphere ~ 140 Ma in age, and the Nidar ophiolite and the Spontang ophiolite to its northwest (Fig. 2) represent the crustal and mantle sections of an intra-oceanic island arc (Mahéo et al. 2004; Ahmad et al. 2008).

Field and petrographic observations

Pillow lavas

The Nidar ophiolite is best exposed around Nidar village, located 4216 m above sea level between the town of Nyoma and the lake of Kyun Tso (Figs. 1 and 2). Elongated exposures of pillow lavas are found on the mostly smooth, scree-covered slopes of linear mountain ridges around Nidar (Fig. 3a). Pillow size is variable (Fig. 3b). The larger pillows at Nidar are ~ 1 m in diameter and are generally roughly spherical and bulbous with glassy surfaces (Fig. 3c, d). Pillow breccia is usually present between the pillows, and many small, filled feeder tubes are seen around the pillows and in the pillow breccia (Fig. 3e, f). Pillows exposed in cross section commonly depict a very fine-grained, highly vesicular interior with a glassy rind (Fig. 3d), and some pillows show radially arranged pipe vesicles along their marginal parts (Fig. 3f). A petrographic study of the Nidar pillow lavas shows that the glassy rinds are dominantly composed of spherulitic growths of plagioclase forming sheaf-like, bow-tie and fully radial patterns (Fig. 4a–c).

Fig. 3
figure 3

a–f Photographs of pillow lavas and associated features of the Nidar ophiolite at Nidar village, with GPS coordinates N 33° 09′ 55.9″, E 78° 36′ 27.0″, elevation 4216 ± 3 m. a Typical summer landscape and pillow lava exposure. Geologists (encircled) for scale. b Cross section of the elongated pillow lava outcrop in a. Geologists for scale. White object above mound is a small religious construct. c Enlarged view of the pillows at the exposed base of the outcrop in a and b. d Highly vesicular pillow (centre) and small feeder tubes. Vertical face. Pen for scale is 15 cm long. e Feeder tubes in the pillow breccia in vertical section. f Feeder tubes and cross sections of numerous small pillows exposed in plan view. Pickhammer in e and f is 33 cm long

Full size image
Fig. 4
figure 4

a–c Thin-section photomicrographs of the glassy rinds of small pillows from Nidar village. Photomicrograph is a in plane polarised light and b and c in cross-polarised light. The views in b and c are dominated by plagioclase spherulites

Full size image

Basalt pillows of an elongate shape, and overlain by hyaloclastite (described below), are exposed 5 km southwest of Nidar village, at 4304 m elevation, on the road to Kyun Tso (Fig. 5a). These pillows have highly vesicular interiors with hundreds of tiny (≤ 1 mm) vesicles of subspherical shape, many of which have been filled by secondary calcite and zeolites (Fig. 5a–c). In some pillows, the vesicle concentration is markedly greater along the marginal parts just under the rind, compared to that in the pillow interior (Fig. 5b). Similarly, pillows with highly vesicular (now amygdaloidal) interiors, and vesicle-poor glassy rinds, are found in roadside boulders 10 km northwest of Nyoma on the way to Mahe, at 4150 m elevation (Fig. 5d–f).

Fig. 5
figure 5

a Elongate pillows underlying hyaloclastite, exposed 5 km southwest of Nidar village, on the eastern side of the road to Kyun Tso. Location N 33° 08′ 06.1″, E 78° 35′ 01.6″, elevation 4304 m. Vertical section, pickhammer is 33 cm long. b, c Highly vesicular basalt pillows (vertical section views) at the location in a. Pen in c is 15 cm long. d Highly weathered pillow lava forming roadside boulders 10 km northwest of Nyoma on the Nyoma–Mahe road. Location is N 33° 12′ 51.1″, E 78° 34′ 04.4″, 4150 m. e Highly vesicular–amygdaloidal and weathered basalt pillow in the boulder shown in d. Vertical face. The vesicles have been filled up by chlorite and other secondary minerals. Coin for scale is 2.5 cm wide. f Thin-section photomicrograph (in cross-polarised light) of an individual spherical amygdule in the pillow shown in e. The amygdule contains radial aggregates of secondary zeolite (stilbite). Note the fine grain size of the host basalt and veins of secondary minerals (white)

Full size image

At Mahe, pillow lavas form sections hundreds of meters thick (Fig. 6a). We observed a tumulus at road level with a characteristic domal shape (Fig. 6b). The overlying pillow lava sequence has abundant pillow breccia, and the pillows range in size from ~ 50 cm to only a few centimeters (Fig. 6c–e). Several pillows may lie juxtaposed, or an individual pillow may be enclosed in pillow breccia. Small, elongate, filled lava tubes traverse the pillow breccia (Fig. 6f). Some pillows evidently grew from these tubes into the surrounding breccia matrix, as shown by well-exposed necks (Fig. 6e).

Fig. 6
figure 6

a–f Photographs of pillow lavas and associated features of the Nidar ophiolite at Mahe, with GPS coordinates N 33° 15′ 40.0″, E 78° 27′ 10.0″, 4172 m. a Pillow lavas forming sections hundreds of meters thick. b Tumulus at the exposed base of the pillows. Vertical face. Pickhammer (encircled) is 33 cm long. c Numerous pillows of variable size enclosed in pillow breccia. Vertical face, person for scale. Photo has been edited on computer to enhance the contrast between pillows and breccia matrix. Bright orange patch of light in lower right results from the sun being in front of the camera. d Detail of the pillows and pillow breccia shown in c. e Close-up of a pillow surrounded by pillow breccia and connected by a neck to its feeder tube. Curved black arrow indicates the inferred direction of lava being fed into the pillow. Ruler is calibrated in inches (length 3 inches) and centimeters (width 5 cm). f Small feeder tubes in pillow breccia

Full size image

The Mahe pillow lava sequence grades upwards into massive sheet lava (Fig. 7a, b) without a distinct eruptive break. The boundary between the pillowed and the massive types is irregular but subhorizontal (Fig. 7a, b). The sheet lava is many meters thick with its top eroded (Fig. 7c) and is intruded by a 5-m-wide dolerite dyke which shows blocky jointing and two distinct chilled margins (Fig. 7c, d).

Fig. 7
figure 7

a, b Pillow lava at Mahe grading upwards into massive sheet lava. Vertical section views, person for scale. Location is N 33° 15′ 40.0″, E 78° 27′ 10.0″, 4172 m. c, d Short, 5-m-thick dolerite dyke cutting sheet lava; geologists for scale, within ellipse in panel c. Dyke has an even texture in the interior and two distinct glassy chilled margins and strikes N 280° and dips steeply south. Location is N 33° 15′ 39.8″, E 78° 27′ 07.9″, 4177 m

Full size image

Hyaloclastite

At the location 5 km southwest of Nidar village on the road to Kyun Tso (Figs. 2 and 5a), hyaloclastite is seen overlying a stack of elongate pillow lavas (Fig. 8a–d). The hyaloclastite contains innumerable angular fragments of relatively unaltered basalt that range in size from 1- or 2-cm- to decimeter-sized blocks, and more rarely blocks > 1 m in size, in a fine, sand- to clay-sized, grey to green matrix which is extensively altered from the original basalt (Fig. 8b, c). Parts of the hyaloclastite with the smaller basalt fragments show relatively well-developed centimeter-scale, ungraded bedding (Fig. 8c). In other parts of the same deposit, however, bedding is absent or is very crude at best, and many fragments and blocks of basalt are dispersed within a largely structureless and chaotic deposit (Fig. 8c, d). The ash-sized matrix contains millimeter- to centimeter-sized chips of basaltic glass (Fig. 8d). Small (centimeter-thick), subvertical, en echelon basaltic dykelets intrude the hyaloclastite. A petrographic study of the hyaloclastite matrix and one of the basalt blocks (Fig. 8e, f) shows abundant phenocrysts of plagioclase and clinopyroxene, and a few phenocrysts of olivine, in a very fine-grained groundmass containing numerous elongated plagioclase laths.

Fig. 8
figure 8

a–d Photographs of hyaloclastite sequence in the Nidar ophiolite. Location is 5 km southwest of Nidar village on the eastern side of the Nidar–Kyun Tso road, at N 33° 08′ 06.1″, E 78° 35′ 01.6″, 4304 m. a 10-m-high cliff section of hyaloclastite. Geologists for scale. b Crudely bedded hyaloclastite with basalt blocks in a fine-grained ash-sized matrix. Vertical cliff face, pickhammer 33 cm long. c Well-bedded hyaloclastite (lower part of section) overlain by structureless hyaloclastite containing large basalt blocks. Person for scale. d Close-up (vertical face) of the hyaloclastite showing an irregularly shaped block of weathered basalt, and small angular chips of basaltic glass (tachylite) in a fine-grained ash-sized matrix. Pen is 15 cm long. e Thin section photomicrograph (in plane-polarised light) of the hyaloclastite, showing abundant plagioclase (Pl) forming phenocrysts and groundmass laths, and clinopyroxene (Cpx) crystals, some of which are indicated. “gl” is a glass shard. f Thin-section photomicrograph (in cross-polarised light) of one of the basalt blocks shown in panel c. The basalt contains abundant phenocrysts of Cpx (brilliant colours), fewer phenocrysts of Pl and a few phenocrysts of olivine (Ol, altered), in a fine-grained groundmass of the same minerals with abundant plagioclase laths

Full size image

The sedimentary cover

Thick deposits of Early Cretaceous radiolarian chert, forming the sedimentary cover of the ophiolite, crop out around Nidar (Fig. 9a, b). The cherts have been dated biostratigraphically to the Hauterivian (132 ± 2 to 127 ± 1.6 Ma) to the Aptian (121 ± 1.4 to 112 ± 1.1 Ma) (Kojima et al. 2001; Zyabrev et al. 2008) and provide an upper limit to the age of the basaltic crust below. The chert sometimes shows a clear depositional contact with pillow lava (Fig. 9a). Strongly tilted and tightly folded chert layers near Nidar village enclose layers of soft, white, silicic volcanic ash (Fig. 9b, c). A petrographic study of the ash (Fig. 9d) shows numerous crystals of quartz and feldspar in a very fine-grained matrix in which it is difficult to identify additional features, but this is mainly a crystal tuff.

Fig. 9
figure 9

a Bedded radiolarian cherts with a depositional contact above pillow lava, near Nidar village. Geologists for scale. b Thick, steeply inclined (75°–65°), N 330°-striking radiolarian chert sequence near Nidar village, enclosing layers of white silicic volcanic ash (two of which are indicated). Geologists for scale. Location N 33° 08′ 37.7″, E 78° 35′ 44.7″, 4275 m. c Close-up of the silicic ash. Coin for scale is 2.7 cm in diameter. d Thin-section photomicrograph of the ash (in plane-polarised light) containing abundant quartz and feldspar crystals (white) and a few Fe–Ti oxide grains (black), some of which are indicated

Full size image

Also found in the vicinity is a deposit in which centimeter- to decimeter-sized globular clasts (rarely, angular fragments) of basalt are present within a matrix of sandstone which dominates the deposit (Fig. 10a–f). The contact of this deposit with the well-bedded cherts exposed nearby (Fig. 9a, b) is hidden under the widespread scree cover. In contrast to the cherts, the deposit is structureless and lacks bedding (Fig. 10a–f). A petrographic study of the sandstone host shows clasts of glassy, phyric basalt lava along with numerous rounded grains of quartz and feldspars, grains of highly fractured quartz, as well as clasts of volcanic rocks and sandstone (Fig. 11a–d), suggesting similarities to greywackes. Some of the larger clasts are composite, made up of numerous smaller rounded to angular grains (Fig. 11c, d).

Fig. 10
figure 10

a–f Outcrop of peperite near Nidar village, at location N 33° 08′ 39.8″, E 78° 35′ 55.7″, 4265 m. Geologists for scale in a and b. c Globular clasts of basalt in sandstone matrix. Vertical face, pen is 15 cm long. d Numerous globular basalt clasts and a blocky basalt clast with fractures in sandstone matrix. Vertical face. e Close-up of a globular basalt clast in chert matrix (vertical face); hand lens for scale. There are many tiny basaltic fragments in the matrix. f Large, irregularly shaped basaltic clast and several smaller ones in the chert matrix (vertical face). Pickhammer is 33 cm long

Full size image
Fig. 11
figure 11

a–d Thin section photomicrographs (in plane-polarised light) of the peperite matrix. The photomicrographs show glassy fragments (gl) showing quenched juvenile magma, besides grains of quartz (Qz), feldspars (Fs), sandstone (sst) and volcanic rocks (vr), some of which are indicated. Some of the phenocrysts present in the glass and volcanic rock fragments are indicated. Note composite clasts in panels c and d

Full size image

Volcanological interpretations

Pillow lavas

Pillow lavas on the seafloor are emplaced in the same manner and at the same low effusion rates as small-scale compound pāhoehoe flows of subaerial settings (e.g. Lewis 1914; Jones 1968; Moore 1975; Self et al. 1997; Duraiswami et al. 2013; Perfit and Soule 2016; Sheth 2018). The dominantly pillowed facies of the volcanic crust of the Nidar ophiolite thus indicates low eruption rates. Excellent evidence for their endogenous growth is provided by the numerous filled lava tubes (Fig. 3d–f) and the tumulus observed at the exposed base of the pillow lava section at Mahe (Fig. 6b). As in subaerial compound pāhoehoe flows, the tumulus, with its crust domed up, suggests localised inflation due to a feeder lava tube blocked downstream (e.g. Rossi and Gudmundsson 1996; Self et al. 1997; Duraiswami et al. 2001). The common occurrence of pillow breccia (Figs. 3e, f and 6c–f) indicates that individual pillows frequently tumbled down slopes under gravity and fragmented, as new pillows continued to form from tubes that carried fluid lava to different parts of a growing edifice in the manner of distributaries (e.g. Moore 1975; White et al. 2015; El Ghilani et al. 2017; Duraiswami et al. 2013, 2019). The abundant plagioclase spherulites in the chilled rinds of the pillows (Fig. 4a–c) attest to considerable undercooling of the lava in contact with cold seawater. As shown by Lofgren (1971, 1974), plagioclase crystal morphology is strongly dependent on the degree of undercooling ΔT, such that crystals are tabular at small ΔT and change to skeletal crystals, dendrites and spherulites with increasing ΔT.

Whereas eruption rates were generally low, they also locally increased greatly as reflected in the transition from pillow lava to thick sheet lava observed at Mahe (Fig. 7a–c). Submarine lobate or sheet flows (Batiza and White 2000; White et al. 2015; Perfit and Soule 2016) are not common in the Nidar ophiolite.

Submarine basalt lava does not extensively degas and vesiculate at considerable depths due to the high confining pressure, unless excessively volatile-rich, and in general, highly vesicular submarine basaltic lavas suggest a shallow (≤ 2 km) emplacement depth (e.g. Moore 1965; Jones 1969; Batiza and White 2000; White et al. 2015). Therefore, the highly vesicular nature of the pillows (Figs. 3d, 4a, b, and 5b, c, e, f) suggests a shallow (≤ 2 km) submarine setting for the Nidar ophiolite volcanic section.

Hyaloclastite

Hyaloclastite has not been described in the Nidar ophiolite so far, but it is known from other Tethyan ophiolites such as the ~ 95 Ma Andaman ophiolite (Jafri et al. 2010). Hyaloclastite is a clastic aggregate generated by the non-explosive quench fragmentation of hot basaltic lava in contact with cold seawater (Batiza and White 2000; White et al. 2015). The well-bedded hyaloclastite suggests quiet depositional conditions below wave base. The structureless and chaotic parts of the hyaloclastite indicate deposition by mass movements due to gravitational instabilities (rock falls and slides) or possibly even earthquakes, which disturbed any original bedding and brought large blocks of basalt (centimeters to over a meter in size) from shallower depths to greater depths. Hyaloclastite formation can occur at various depths, from shallow submarine conditions at emergent islands to deep submarine conditions in island arcs, mid-ocean ridges and seamounts (Batiza and White 2000).

Radiolarian chert

Radiolarian chert is a siliceous marine sedimentary rock commonly found overlying oceanic crust in ophiolites, and is often cited as indicating deep (~ 5 km) deposition (e.g. Holmes and Holmes 1978; Emiliani 1992; Tucker 2011). However, this aspect is complex and related to the concept of carbonate compensation depth (CCD). Briefly put, seawater contains increased amounts of dissolved CO2 with depth as a combined effect of increasing pressure and decreasing temperature. Organic tests sinking through seawater, if composed of carbonate, begin getting dissolved as they cross the lysocline and are completely dissolved at the CCD, whereas siliceous tests (of radiolarians and diatoms) sink through the CCD unaffected. The CCD is thus the boundary between calcareous deposits above and non-calcareous deposits below and can be a guide to the palaeodepth of siliceous oceanic sediment such as that overlying the Nidar oceanic crust.

However, the CCD shows significant differences across the different oceans and with latitude even today, and has varied greatly (several kilometers) through space and time (van Andel 1975; Thiede et al. 1980). It was shallow (around 3600 m) between the Late Jurassic and the early Late Cretaceous in the Pacific and Indian oceans (van Andel 1975). According to Sclater et al. (1977, cited in Thiede et al. 1980), the CCD in the Indian Ocean rose from an ~ 3500 m palaeodepth in the Late Jurassic to as shallow as 2.0–2.5 km palaeodepth by the mid-Cretaceous. This means that a CCD palaeodepth of 2.5–3 km may well apply to the 132–112 Ma Nidar radiolarian cherts deposited in the Neo-Tethys Ocean, implying their relatively shallow depth of deposition. The abundance of this Early Cretaceous radiolarian chert may also reflect the fact that carbonate-secreting pelagic microorganisms had evolved only shortly before, in mid-Mesozoic or Late Jurassic time (Tappan and Loeblich 1973; Saraswati and Srinivasan 2016). We therefore consider the bedded radiolarian cherts (Fig. 9a, b) overlying the Early Cretaceous Nidar oceanic crust as formed in a shallow submarine setting.

Peperite

We interpret the deposit containing globular basalt clasts in an impure sandstone matrix as a peperite. Peperite is a composite rock formed essentially in situ by the disintegration of magma which intrudes and mingles with unconsolidated or poorly consolidated, typically wet, sediment (Busby-Spera and White 1987; White et al. 2000; Skilling et al. 2002; El Desoky and Shahin 2020). The Nidar peperite thus suggests contemporaneous sedimentation and volcanism (or shallow-level intrusion).

Magma–sediment mingling is favoured when the intruding magma and the host wet sediment have similar densities and viscosities (Zimanowski and Büttner 2002). The mingling is aided by vapourisation of the pore waters by the heat of the magma and fluidisation of the sediment (Kokelaar 1982), and magma quenching and fragmentation of both rock types in a rheologically liquid-like or plastic condition (see Skilling et al. 2002 for a discussion of several other mechanisms). Cooling of the magma in a peperite occurs orders of magnitude slower than with water (White et al. 2015), but the poorly vesiculated basalt clasts in the Nidar peperite (Fig. 10c–f) suggest quenching before vesiculation could occur. The small glassy fragments in the sandstone matrix (Fig. 11a–c) reflect such quenching of juvenile basalt magma.

Magma–sediment mingling at Nidar also destroyed any former bedding. Peperites around the world are characteristically unstratified and ungraded and often highly discordant to bedding (White et al. 2000; Skilling et al. 2002). The Nidar peperite can be described as globular rather than blocky (terminology of Busby-Spera and White 1987; Skilling et al. 2002) and as dispersed rather than closely packed (terminology of Hanson and Wilson 1993). The exposed section of the peperite near Nidar village (Fig. 10a–f) does not show either a lava flow or an intrusion, but the intrusion was likely small and broke up entirely at the level observed.

Like the hyaloclastite, peperite has not been described in the Nidar ophiolite so far, but it is reported from other ophiolites such as the ~ 95 Ma Oman ophiolite (Jordan et al. 2008). Note that the 132–112 Ma age of the Nidar radiolarian cherts implies that the spatially associated dispersed peperite reflects small-scale, shallow-level, mafic intrusive activity continuing several million years after the formation of the underlying basaltic oceanic crust. The silicic ash layers found within the chert beds (Fig. 9b, c) indicate ongoing explosive eruptions at the same time. Below, we discuss the tectonic implications of all our combined observations and volcanological interpretations.

Discussion

Intermediate- and fast-spreading mid-ocean ridges, which have high magma supply rates, most commonly contain lobate and sheet flows, and hyaloclastite is rare or absent on fast-spreading ridges like the East Pacific Rise (Batiza and White 2000; White et al. 2015; Perfit and Soule 2016). The lack of abundant sediment also characterises fast-spreading ridges. On the other hand, pillow lavas are the dominant lava type at slow-spreading ridges (which have low magma supply and lava extrusion rates) and on flanks of seamounts. These considerations rule out an origin of the Nidar ophiolite at an intermediate- to fast-spreading ridge. A slow-spreading ridge setting is also excluded by the fact that such ridges (such as the Mid-Atlantic Ridge) are deep-seated (Winter 2011, p. 264), whereas the abundantly vesicular pillow lavas require a shallow (≤ 2 km) emplacement depth, an explanation also favoured by the peperite which requires unconsolidated sediment to form.

Besides, the layers of soft silicic ash found in the chert cover (Fig. 9b, c) reflect explosive eruptions from distant or nearby volcanoes, and the abundant grains of quartz, feldspar and volcanic rocks in the sandstone matrix of the Nidar peperite suggest the input of terrigenous sediment from emergent land nearby. Several composite clasts (Fig. 11c, d) indicate that this sediment was reworked at least in part.

The simplest tectonic interpretation of the combined field and petrographic observations and volcanological interpretations is a mafic–silicic island arc, exemplified by modern intra-oceanic arcs such as the Kermadec, Mariana, or Andaman arcs (e.g. Kano 2003; Pal and Bhattacharya 2011; Saha et al. 2019). We suggest that the thick deposits of radiolarian chert in the Nidar ophiolite may themselves be a consequence of such a compositionally evolved island arc setting: erosion of the emergent silicic arc volcanoes would supply abundant SiO2 to the sea, leading to proliferating biological (radiolarian) activity and ultimately extensive radiolarian chert deposition. Peperites are frequently associated with syn-volcanic intrusions in submarine sedimentary successions including those in island arcs (Hanson and Hargrove 1999; Gifkins et al. 2002).

Mahéo et al. (2004) and Ahmad et al. (2008) invoked an intra-oceanic arc setting for the Nidar ophiolite based on geochemical and isotopic data, which show no continental crustal signature but show the involvement of subduction-related fluids. That interpretation is consistent with the volcanological observations and interpretations we have made in the present study. The geochemical and volcanological interpretations complement and corroborate each other well in the case of the Nidar ophiolite. There can, however, be differences in detail. Ahmad et al. (2008) suggested that various arc-related magmatic rocks from the Indus suture zone contain a record of increasing arc maturity with time. According to them, the arc at 140 ± 32 Ma was immature and composed of mafic rocks, intermediate rocks (andesites and granodiorites) began forming at ~ 100 Ma, and finally rhyolites and diorites formed when India-Asia collision occurred at 50–45 Ma. However, the silicic ash layers in the 132–112 Ma radiolarian chert at Nidar village show that shallow-depth mafic intrusive activity (seen in the peperite) and explosive silicic eruptive activity were already occurring early on. Evolving arc maturity from mafic through intermediate to silicic compositions, over ~ 100 million years (Ahmad et al. 2008), is not in evidence.

It is known that ophiolite sections commonly become tectonically dismembered during obduction on land (Coleman 1977), and such dismemberment has also affected the Nidar ophiolite. Around Nidar village, strong tectonic deformation is evident in the steeply to vertically dipping to tightly isoclinally folded radiolarian chert beds (Fig. 9b). It is curious therefore that a number of diverse field observations we have made in the Mahe section, such as the direction of convexity of the tumulus (Fig. 6b), the direction of convexity of the pillows (Fig. 6d, e) and the subhorizontal nature of the transition from pillowed lava to sheet lava (Fig. 7a, b), all suggest that this section is not only right side up, but the present horizontal is also roughly the palaeohorizontal. This, in turn, indicates only minor tectonic tilting of the obducted volcanic section. How this might have happened is a question for structural geologists, but it is possible that the Mahe section is the locally subhorizontal limb of a regional-scale fold whose steep to subvertical limb forms the Nidar village section.

Conclusions

The Early Cretaceous Nidar ophiolite in Ladakh, Indian Trans-Himalaya, is one of the important Tethyan ophiolites of the world. The basaltic volcanic facies in the Nidar ophiolite include abundant pillow lavas associated with hyaloclastite. Layers of silicic ash occur in the overlying radiolarian chert, and peperite is also found in the sedimentary cover. We interpret this assemblage to have formed in a shallow (2–2.5 km) submarine environment in a compositionally bimodal intra-oceanic island arc in the Neo-Tethys Ocean, consistent with past interpretations based on geochemical (including Sr–Nd isotopic) data. Geochemical data are very useful, in fact essential, in understanding the petrogenesis (mantle vs. crustal sources and magmatic processes) of ophiolites. However, palaeotectonic interpretations of ophiolites based on geochemical data should be backed by independent lines of evidence. As shown here with the case study of the Nidar ophiolite, volcanic facies, seriously underutilised in the ophiolite literature, are a valuable guide to the palaeodepths and palaeotectonic settings of ophiolites, and constitute an independent check on their geochemistry-based tectonic interpretations.