Tracing patterns of erosion and drainage in the Paleogene Himalaya through ion probe Pb isotope analysis of detrital K-feldspars in the Indus Molasse, India more

Earth and Planetary Science Letters 188 (2001) 475^491 www.elsevier.com/locate/epsl Tracing patterns of erosion and drainage in the Paleogene Himalaya through ion probe Pb isotope analysis of detrital K-feldspars in the Indus Molasse, India Peter D. Clift *, Nobumichi Shimizu, Graham D. Layne, Jerzy Blusztajn Department of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA Received 23 October 2000; received in revised form 2 March 2001; accepted 12 April 2001 Abstract The Indus Molasse is a pre- and syn-tectonic sedimentary sequence situated in the Indus Suture Zone of the western Himalaya. Spanning in time the collision of India and Asia, this deposit is well placed to record the evolving uplift and erosion history of the early Himalayan orogen. Nd isotope analyses from clay extracted from shales interbedded within the dominantly alluvial sequence indicate a low negative ONd (31.64 to 0.72), in the basal Paleocene Chogdo Formation, slightly more negative than measured values from the Transhimalaya and Kohistan/Dras Arc. Up-section ONd becomes more negative, as low as 310.05, indicating influence of a different, more enriched source. Ion microprobe Pb isotopic analyses of single K-feldspars help constrain this source as being either the Lhasa or Karakoram Blocks, with westward paleo-current flow favoring the former. 207 Pb/204 Pb ratios are too low to be consistent with known Indian plate sources, a conclusion supported by the lack of muscovite or garnet that would be indicative of a High Himalayan contribution. Given the known age of rapid cooling of the High Himalaya at V20 Ma, and the lack of exposure of suitable lithologies prior to that time, an age of sedimentation prior to V20 Ma is inferred. The post-collisional change in paleo-flow and provenance is suggested to reflect the initiation of the Indus River during the Early Eocene. This study demonstrates the power of combined bulk sediment and single grain analyses in resolving provenance in tectonically complex settings. ß 2001 Elsevier Science B.V. All rights reserved. Keywords: India; Asia; Himalayas; provenance; ion probe data; isotopes 1. Introduction The Early Cenozoic collision of India with Eurasia and the consequent uplift of the Himalaya and Tibetan Plateau have created the most dra- * Corresponding author. Tel.: +1-508-457-2000/3437; Fax: +1-508-457-2187; E-mail: pclift@whoi.edu matic relief on earth. Understanding the growth of this system is not only important to understanding the processes of orogeny, but is also crucial to testing models of climate^tectonic coupling in South Asia. The height and extent of the Tibetan Plateau and High Himalaya disrupts atmospheric circulation on a global scale [1] and hence the Cenozoic growth of the plateau, coupled with chemical weathering in the Himalaya, may ultimately be responsible for the global cooling that led to the Plio^Pleistocene Ice Age [2,3]. More- 0012-821X / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 0 1 2 - 8 2 1 X ( 0 1 ) 0 0 3 4 6 - 6 476 P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 over, the Tibetan Plateau appears to play an important role in driving a strong summer monsoon [4^6]. Determining the uplift history of the system is crucial to testing such models. Knowledge of the Miocene to Recent uplift and erosion history is now constrained in outline, due to work on sediments from the foreland basins [7], the Bengal Fan [8,9], and from direct measurements on the crystalline basement of the High Himalaya (e.g. [10]). However, the early development of the system has remained obscure. This is because those rocks exposed at or close to the surface at that time have been eroded away leaving only the relatively insensitive, high temperature paleothermometers to record the cooling history at that time. Peak metamorphism in the Pakistan Himalaya postdates collision by V10 Myr [11] so that no record of Himalayan orogenesis preceding V45 Ma can be found in this tectonic unit. Further east, in the High Himalaya of Zanskar and Lahaul, rapid cooling dates from 20^25 Ma [12,13], limiting the record of earlier orogenesis even further. Uplift history is therefore best charted through study of the detrital sedimentary record that spans this period. Unfortunately access to Eocene^Oligocene sediments is limited due to the di¤culty in drilling the great thicknesses of the Indus or Bengal Fans, and due to the fragmentary nature of the sedimentary record of this age in the Indian foreland [14], which in any case was located far from the zone of active collision during the Eocene. In this study we present data on the erosion history of the early Himalaya recorded in the Indus Molasse Basin, located in the Indus Suture Zone in the Indian Himalaya, which help constrain the nature of early Cenozoic erosion and the development of a post-collisional drainage system. To do this we investigate the source of the detrital minerals that comprise this succession using a combination of bulk sediment and single grain isotopic analyses. Fig. 1. (A) Schematic tectonic map of the Indus Suture showing the main tectonic units discussed in this paper, as well as principle localities. Insert shows the location of Ladakh with the Himalaya^Tibet system. (B) Detailed map of the Zanskar Gorge showing the relationships of the Indus Molasse with the underlying units of the Asian and Indian margins. P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 477 2. Geologic setting The Indus Molasse is a folded and thrusted sequence of dominantly clastic formations which is observed locally to rest unconformably over the Ladakh Batholith, where this contact is not rethrusted. The Indus Molasse also unconformably overlies Indian passive margin units (Lamayuru ¨ Group) south of Upsi [11], ophiolitic melanges west of Chilling [15], and the Cretaceous forearc of the Kohistan/Dras Arc (Nindam Formation), also west of Chilling [15,16]. However, the Indus Molasse is observed to conformably overlie the Upper Cretaceous^Paleogene Jurutze Formation [16,17] within the Zanskar Gorge, south of Sumda-Do (Fig. 1). The Jurutze Formation represents the forearc basin to the Asian margin prior to Indian collision. The Ladakh Batholith represents the precollisional active margin of Asia, and is thus correlative to the Gandese Batholith and other parts of the Transhimalayan igneous belt exposed further east. The Ladakh Batholith is dated as young as 60 Ma near Leh by U^Pb methods [18]. Most recently di¡erent phases of intrusion spanning 49 to 61 Ma have been mapped by Weinberg [19] through ion probe dating of single zircon crystals. The position of the Indus Molasse within the suture between Indian and Asian plates led earlier workers to postulate a forearc position for its deposition prior to India^Asia collision [16,17,20]. In this scenario sedimentation in what was a continental arc forearc basin continued in an early intra-montane setting after collision [17]. Garzanti and van Haver [17] subdivided the molasse section into a series of formations, the older of which are dated by their marine fauna, but which become continental and barren of datable fauna above the Nummulitic Limestone on the south side of the basin (Fig. 2). Recent re-dating of the fauna in the Nummulitic Limestone [21] as latest Paleocene, and the recognition that the molasse oversteps both Indian and Asian tectonic units [16] allows India^ Asia collision in Ladakh to be constrained as pre-Eocene. Since that time the entire sequence has been deformed by Neogene north-vergent tectonism, related to motion along the main Zanskar backthrust, located just south of the molasse outcrop [22]. 3. Stratigraphy We choose to follow the de¢ned stratigraphy of Searle et al. [22], based on the Indus Molasse section exposed in the Zanskar Gorge (Figs. 1 and 2), because this section is the focus of this study. In this scheme the base of the section is marked by a well-cleaved, light, bu¡-colored shalely carbonate sequence, the Sumda Formation, in practice the upper part of the Jurutze Formation [17,23]. The Sumda Formation is clearly marine, having a foraminifer fauna that constrains the Maastrichtian age [21]. It is conformably overlain by the Chogdo Formation, a dark, red-purple weathering series of interbedded sandstones, siltstones and mudstones, as well as minor conglomerates. This formation is sharply overlain by dark, thick-bedded, Nummulite-bearing limestone of latest Paleocene age [21]. Searle et al. [22] map the Chogdo Formation as unconformably overlying deformed turbidites close to Chilling Village, which they assigned to the Nindam Formation, part of the Kohistan/Dras intraoceanic volcanic arc. However, these have been re-assigned to the enigmatic, but presumably Asian, Khalsi Flysch of Brook¢eld and Andrews-Speed [23] by Clift et al. [16] (Fig. 1B). The Chogdo Formation further overlies Lamayuru Group Indian slope turbidites and oceanic serpentinized harzburgite in the same area. Consequently the Chogdo Formation is the ¢rst unit to overstep both terrains of Indian and Asian origin and so constrains the timing of collision to being prior to its deposition, i.e. s 54.6 Ma [16]. Above the Nummulitic Limestone the section is dominated by coarse-grained siliciclastic sediments of typical braided alluvial facies. Three upper formations are recognized (Fig. 2). The lowermost Nurla Formation is topped by a shalely lacustrine section containing plant fossils, and is overlain by thick conglomerates of the Choksti Conglomerate, in turn succeeded by coarse alluvial sediments of the Choksti Formation (Nimmu Formation of Garzanti and van Ha- 478 P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 ver [17]). The Choksti Formation onlaps the edge of the Ladakh Batholith, although the original depositional contact is often rethrust. The Hemis Conglomerate, exposed at Hemis, located east of the Zanskar Gorge (Fig. 1), is a particularly thick sequence seen on both sides of the basin, and may re£ect the ¢rst in£ux of coarse material following collision. In this study we equate the Choksti Conglomerate with the Hemis Conglomerate because along the Hemis and Zanskar sections both units represent the oldest thick conglomerate bed in the stratigraphy. Although it is conceivable that a second higher conglomerate in the section at Hemis might be the equivalent of the Choksti Conglomerates [17], this is not possible to demonstrate without precise dating or long-distance stratigraphic correlation, neither of which is presently possible. We therefore de¢ne the Choksti Conglomerate as the ¢rst major granite-bearing conglomerate to be derived into the basin, as seen in the Zanskar Gorge. The Choksti Conglomerate contains palm leaves that have been tentatively dated as early Oligocene [24], but otherwise the Molasse is undated above the Nummulitic Limestone. Clift et al. [16] report apatite ¢ssion track ages of 35 Ma for the Ladakh Batholith underlying the upper part of the Choksti Formation just north of Khalsi (Fig. 1). This date constrains the top of the formation to be younger than 35 Ma. As discussed below, the lack of minerals derived from the High Himalayan crystalline units means that sedimentation of the Indus Molasse pre-dates exposure of these terrains at V20 Ma [12,13]. It seems likely that the northward thrusting of the Zanskar Himalayas synchronous with the unroofing of the High Himalaya at V20^25 Ma inverted the basin and halted sedimentation [22]. If so then sedimentation may have continued until V20^ 25 Ma. Only modest data presently exist to constrain the provenance of the Indus Molasse clastic sediments. Sandstones below the Choksti Conglomerate on the northern side of the Indus Molasse Basin (Temesgam Formation of van Haver [24], Nurla Formation of Searle et al. [22]) comprise quartzose alluvial and deltaic sandstones and shales, with intercalations of marine facies. Gar- zanti and van Haver [17] interpreted these to have been derived by erosion of the Ladakh Batholith during active Andean-type subduction prior to India collision, based on petrographic work. Similarly other studies [22,23] have interpreted the Molasse as principally being derived from erosion of the Ladakh Batholith, in part because of the unconformable relationship between the Molasse and the Ladakh Batholith noted along the northern boundary of the basin. Along this boundary the similar composition of boulders in the basal conglomerate and the basement is clear. 4. Detrital mineral compositions The provenance of the Indus Molasse can be constrained in part through its mineralogy. While epidote is common throughout the section, hornblende is only found in signi¢cant amounts in the upper parts of the Choksti Formation (Fig. 2). There is also apparent decrease in the dominance of biotite up-section. What is most noticeable in backscattered electron probe images is how sandstone located above the Nummulitic Limestone contains grains of large, single K-feldspar and quartz minerals, whereas the proportion of microcrystalline granites is high in the Chogdo Formation. Alteration is also higher for feldspars in the Chogdo Formation, as shown by electron probe analytical totals signi¢cantly less than 100% in this unit. Chogdo Formation feldspar crystals are often cloudy in thin section, due to their conversion to clays. However, it is not clear whether this change from more altered to less weathered feldspar re£ects a change in the source character, or if this might be in part due to the greater burial temperatures experienced by the Chogdo Formation. Illite crystallinity data indicate that the Chogdo Formation approached anchizonal conditions (V250³C, [16]) compared to low diagenetic grade for the northernmost Choksti Formation (110^200³C). The most remarkable aspect of the detrital mineralogy is its apparent lack of minerals characteristic of the High Himalaya, i.e., muscovite and garnets (cf., [25]). At outcrop there are clear di¡erences between the Chogdo and younger formations, aside from P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 479 the weathering color. Speci¢cally, in the Chogdo Formation the proportion of granitic clasts is less, whereas degraded volcanic rocks, serpentinized harzburgite and red chert are more dominant. In the upper part of the Indus Molasse granite boulders are the greatest single contributors, but unlike the altered appearance of those boulders in the Chogdo Formation, these are often quite fresh and show well developed biotite and hornblende phenocrysts, in addition to quartz, plagioclase and potassium feldspars. Fig. 3. Measured paleo-current indicators for the Indus Molasse showing evolving provenance during India^Eurasia collision and onset of Indus River £ow. 5. Paleo-current indicators Paleo-£ow directions can be useful in constraining possible sediment sources. Since most of the sediments that comprise the Indus Molasse are braided river facies sandstones [24], this means that cross bedding is the most common form of recognizable paleo-current indicator. Brief turbidite intervals in the Choksti Formation, presumably of lacustrine origin, show scour structures within channel complexes. The anastamosing nature of braided streams means that there is an inherent spread of current indicators, due to the lateral growth of sand or gravel bars. In addition, because the Indus Molasse was deposited in a tectonically active setting, one might expect the paleo-rivers that carried the sediment to have negotiated irregular topography and not to have a perfectly linear geometry. Further disruption to a simple current pattern may be introduced during the rethrusting of the Indus Molasse towards the north during the Neogene, when individual thrust sheets may rotate relative to one another along the strike of the suture zone. Despite these complications the paleo-current indicators from the Indus Molasse show a clear evolution from a pre-collisional trend towards the SW in the Jurutze Formation, followed by dominant NNE-directed £ow in the Chogdo Formation, switching to £ow from NE to SW above the Nummulitic Limestone in the Nurla and Choksti Formations (Fig. 3). This latter trend appears to be constant from east to west across the studied area and matches earlier observations [22]. The SW-directed £ow in the Jurutze Formation is consistent with these sediments being eroded from the Transhimalayan Arc, located to the Fig. 2. Simpli¢ed stratigraphic log of the Indus Molasse showing relative abundance of di¡erent mineral groups upsection. 480 P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 north. The ophiolitic character of many of the grains (red cherts, basalts, with rarer gabbros and peridotites) and the NNE-directed £ow in the Chogdo Formation argues in favor of a marked change of provenance during the earliest stages of collision, dominated by erosion from ophiolites and associated units on the Zanskar Platform [17]. The presence of some granitic clasts in the Chogdo Formation is more di¤cult to reconcile with an origin from the SW, because there are no appropriate sources known of that age in that area. Although a proto-arc complex has been identi¢ed associated with the Spontang Ophiolite (Spong Arc [26]) this comprises only volcaniclastic material and cannot be considered a source of the granite cobbles. Most likely these clasts were derived from the Ladakh Batholith and were mixed Table 1 Nd isotopic values measured from the Indus Molasse Sample LA98-27 LA98-29 LA98-59 LA98-41 LA98-34 LA98-61 LA-00-1 LA-00-5 LA-00-3 LA-00-4 LA-00-6 LA98-38 LA-00-11 LA-00-13 LA-00-10 LA-00-15 LA-00-16 LA-00-17 LA-00-9 LA98-15 LA98-19 LA98-23 LA98-24 LA-00-18 LA-00-21 LA-00-23 LA-00-25 LA-00-27 LA-00-22 LA-00-20 143 with the ophiolitic and platform material in an intermontane basin setting. The correlation of a change to a westward £ow with a more granitic clast population in the Nurla and Choksti Formations is noteworthy. It is not clear what the source of this in£ux is, although the Transhimalaya and Lhasa Block are the most likely candidates. 6. Nd isotopes The source of the Indus Molasse can be further constrained using the Sm^Nd isotopic system. The technique is based on the assumption that the ¢nest fraction of detrital material represents a good average composition of the source area drained. Since weathering and the sediment trans- Nd/144 Nd ONd 36.83 0.82 310.05 34.56 39.75 36.32 1.60 33.37 36.41 38.40 35.71 35.69 33.39 3-4.43 33.19 32.30 31.44 33.18 32.42 30.18 31.64 31.31 31.38 30.21 30.72 30.98 0.02 30.64 0.72 0.19 Stratigraphic position Choksti Formation Choksti Formation Choksti Formation Choksti Formation Choksti Formation Choksti Formation Choksti Formation Choksti Formation Choksti Formation Choksti Formation Choksti Formation Choksti Conglomerate Choksti Conglomerate Nurla Formation Nurla Formation Nurla Formation Nurla Formation Nurla Formation Nurla Formation Chogdo Formation Chogdo Formation Chogdo Formation Chogdo Formation Chogdo Formation Chogdo Formation Chogdo Formation Chogdo Formation Chogdo Formation Chogdo Formation Chogdo Formation Location Stok Stok Likir Upsi Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar Zanskar 0.512290 þ 5 0.512682 þ 5 0.512125 þ 5 0.512406 þ 5 0.512140 þ 5 0.512316 þ 5 0.512722 þ 5 0.512467 þ 5 0.512311 þ 5 0.512209 þ 5 0.512347 þ 5 0.512348 þ 4 0.512466 þ 6 0.512413 þ 5 0.512476 þ 5 0.512522 þ 5 0.512566 þ 8 0.512477 þ 5 0.512516 þ 5 0.512631 þ 5 0.512556 þ 5 0.512573 þ 5 0.512569 þ 6 0.512629 þ 4 0.512603 þ 5 0.512590 þ 5 0.512641 þ 8 0.512607 þ 4 0.512677 þ 5 0.512650 þ 6 Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Gorge Results corrected for La Jolla standard = 0.511847. P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 481 port process are not expected to result in isotopic fractionation, the measured isotopic signature of the shale fraction should re£ect the bulk composition of the source. We compare modern Nd isotopic character of the sedimentary rocks and the sources with no age correction. Because the concentration of the parent 146 Sm is di¡erent in the sources and sediments, 146 Nd/144 Nd values in the two will gradually diverge after erosion. However, the half life of 146 Sm is 1.03U108 yr, while the sediments were all deposited after 55 Ma, insu¤cient time for an isotopic di¡erence of signi¢cant magnitude compared to the isotopic di¡erences between sources to emerge. Twenty nine samples were analyzed, 25 taken along the Zanskar River section, and four from the Choksti Formation along strike at Upsi, Stok and Likir (Fig. 1). The clay fraction was separated from shales by simple crushing, sieving and then centrifuging to concentrate the ¢nest fraction ( 6 2 Wm). The clay was then dissolved and the Nd separated using standard column extraction techniques. Nd isotopic compositions were determined on VG354 mass spectrometer at Woods Hole Oceanographic Institution (WHOI). 143 Nd/ 144 Nd values are normalized to 146 Nd/144 Nd = 0.7219 and are relative to 0.511847 for the La Jolla standard. The results are shown in Table 1. We calculate the parameter ONd [27] using a 143 Nd/144 Nd value of 0.512638 for the Chondritic Uniform Reservoir (CHUR ; [28]). On the Bengal Fan comparison of clay mineral Nd characteristics with those of High Himalayan metamorphic rocks [29] provided a good ¢t and indicated that these were the principal sediment sources to the fan [9,30^32]. Despite the suggestion that Nd isotopes are sensitive to input from oceanic sources [33], values measured from Bengal Fan clays are consistent with other measurements of £uvial and aeolian particles, and are similar to those measured in modern sediments from the Ganges River [34]. This probably re£ects the very low concentration of rare earth elements in seawater compared to the source terrains. In the case of the £uvial Indus Molasse contamination from the oceans is not a factor and therefore our con¢dence in the Nd characteristics re£ecting source is even greater. Fig. 4 shows the range and frequency of ONd values noted in the source and Indus Molasse and compares these with published values from Fig. 4. Comparison of ONd values measured (A) in the Indus Molasse, and (B) from the potential source terranes of the Himalaya and Tibet. 482 P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 Table 2 Pb isotopic values measured on detrital feldspars from the Indus Molasse Sample LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA-98-45 LA98-63 LA98-63 LA98-63 LA98-63 LA98-63 LA98-63 LA98-36 LA98-36 LA98-36 LA98-36 LA98-36 LA98-36 LA98-36 LA98-36 LA98-36 LA98-36 LA98-36 LA98-22 LA98-22 LA98-25 LA-98-12 LA-98-12 LA-98-12 LA-98-12 LA-98-12 LA-98-12 LA-98-12 LA-98-12 LA-98-12 LA-98-12 Formation Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Choksti Chogdo Chogdo Chogdo Chogdo Chogdo Chogdo Chogdo Chogdo Chogdo Chogdo Chogdo Chogdo Chogdo Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Conglomerate Conglomerate Conglomerate Conglomerate Conglomerate Conglomerate Conglomerate Conglomerate Conglomerate Conglomerate Conglomerate Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation Formation 206 Pb/204 Pb Pb/204 Pb c (%) 0.28 0.21 0.55 0.48 0.21 0.22 0.21 0.42 0.30 0.49 0.23 0.51 0.33 0.35 0.31 0.42 0.61 0.25 0.22 0.22 0.35 0.33 0.76 0.50 0.22 0.61 0.41 0.27 0.28 0.56 0.74 0.29 0.47 2.70 0.26 0.78 0.22 0.29 0.18 0.58 0.28 5.00 0.36 0.45 0.39 0.34 0.36 0.33 0.29 0.22 0.32 0.35 0.28 206 207 Pb/206 Pb Pb/206 Pb c (%) 0.13 0.11 0.23 0.25 0.14 0.10 0.10 0.32 0.15 0.34 0.23 0.28 0.16 0.18 0.14 0.27 0.26 0.15 0.11 0.11 0.17 0.17 0.31 0.35 0.13 0.50 0.29 0.14 0.13 0.10 0.25 0.16 0.23 2.60 0.18 0.46 0.12 0.15 0.09 0.48 0.11 1.00 0.16 0.20 0.17 0.17 0.15 0.17 0.15 0.13 0.15 0.18 0.17 207 208 Pb/206 Pb 208 Pb/206 Pb c (%) 0.12 0.12 0.28 0.26 0.21 0.09 0.11 0.41 0.19 0.48 0.29 0.33 0.15 0.20 0.15 0.32 0.27 0.19 0.12 0.14 0.19 0.16 0.35 0.39 0.12 0.86 0.44 0.14 0.19 0.11 0.26 0.23 0.24 3.30 0.20 0.64 0.11 0.17 0.10 0.79 0.12 0.94 0.21 0.18 0.14 0.13 0.14 0.16 0.12 0.11 0.13 0.15 0.17 18.7336 19.2864 18.4128 18.4638 19.2976 18.7614 18.7238 18.6668 18.9380 18.4094 18.7210 18.6518 18.6150 18.7913 18.3318 18.8437 18.4379 18.8126 18.5401 19.0042 18.7981 18.5995 17.6907 18.4366 18.6150 18.4809 18.4638 18.6289 18.4997 18.6289 18.7935 18.7441 18.5151 17.3883 18.7021 18.3925 18.6220 18.5908 18.6741 19.0186 18.3352 16.8067 18.3150 18.3918 18.4807 18.5551 18.4673 18.4679 18.4981 18.4720 18.4550 18.5901 18.5154 0.8363 0.8081 0.8380 0.8397 0.7968 0.8343 0.8344 0.8326 0.8227 0.8332 0.8364 0.8386 0.8367 0.8302 0.8406 0.8346 0.8359 0.8333 0.8341 0.8214 0.8275 0.8338 0.8360 0.8356 0.8399 0.8390 0.8365 0.8396 0.8419 0.8433 0.8399 0.8315 0.8335 0.8491 0.8332 0.8372 0.8354 0.8324 0.8385 0.8201 0.8471 0.8464 0.8389 0.8393 0.8406 0.8412 0.8404 0.8428 0.8405 0.8407 0.8402 0.8413 0.8403 2.0830 1.9990 2.0710 2.0720 2.0544 2.0844 2.0865 2.0824 2.0548 2.0663 2.0913 2.0759 2.0744 2.0768 2.0865 2.0832 2.0755 2.0862 2.0856 2.0558 2.0671 2.0782 2.0867 2.0790 2.0850 2.0843 2.0870 2.0870 2.1010 2.1070 2.0770 2.0730 2.0820 2.1052 2.0740 2.0740 2.0840 2.0750 2.0850 2.0380 2.0860 2.1060 2.0708 2.0765 2.0835 2.0808 2.0870 2.0913 2.0883 2.0847 2.8699 2.0891 2.0903 P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 Table 2 (continued) Sample LA-98-12 LA-98-12 LA-00-94 LA-00-94 LA-00-94 LA-00-94 Formation Chogdo Formation Chogdo Formation Ladakh Granite Ladakh Granite Ladakh Granite Ladakh Granite 206 483 Pb/204 Pb 206 Pb/204 Pb c (%) 207 Pb/206 Pb 207 Pb/206 Pb c (%) 208 Pb/206 Pb 208 Pb/206 Pb c (%) 0.14 0.10 0.36 0.72 0.21 0.25 18.5727 18.3785 18.6162 18.3877 18.4810 18.3036 0.34 0.32 0.63 0.59 0.38 0.48 0.8380 0.8410 0.8398 0.8435 0.8442 0.8435 0.17 0.13 0.37 0.28 0.27 0.26 2.0774 2.0883 2.0832 2.0934 2.0880 2.0936 the High Himalaya [32,35], Lesser Himalaya [9,30,36], the Transhimalaya [37], Kohistan/Dras Arcs [23,38,39], the Karakoram [40] and Lhasa Block [41]. It is noteworthy that the range of values from the Transhimalayan and Kohistan/Dras Arcs are similar, as are those from the Karakoram and Lhasa Block, meaning that this isotopic system cannot resolve between these sources. The similarity of the Karakoram and Lhasa Block re£ects the fact that they both represent parts of the Asian margin on which the Transhimalaya Arc was built. Di¡erences between these terranes today principally re£ect greater degree of exhumation in the Karakoram [42]. The Chogdo Formation has the most positive values (0.72 to 31.64), the Nurla Formation is intermediate (31.44 to 34.43), and the Choksti Formation has the least radiogenic values (1.60 to 310.05). The range of ONd values from the Chogdo Formation requires that this unit be dominated by sources with positive ONd values, i.e., the Transhimalaya and Kohistan/Dras Arc. Although no Nd data are known from the Spontang Ophiolite, its similar age and oceanic origin would imply that it too would have a strongly positive ONd value. Some values from the Lhasa Block also overlap the measured range. The Nd data are thus consistent with the Spontang Ophiolite, or some equivalent Cretaceous ophiolite unit, together with the Ladakh Batholith being the dominant sources to the Chogdo Formation. These sources are evidenced by the south to north paleo-current data (indicating the Spontang Ophiolite) and the presence of granitic clasts (indicating the Ladakh Batholith). The ONd values permit additional limited mixing with an isotopically more negative source. In the Nurla and Choksti Formations ONd becomes progressively more negative up-section. The range of ONd values is similar to measured values from the Lhasa Block, but could represent mixing of both more isotopically positive and negative sources. Two samples from the Choksti Formation di¡er from the general pattern in showing positive ONd values, more similar to the Chogdo Formation. Both these samples come from exposures directly adjoining the Ladakh Batholith and represent local derivation from that block in the form of alluvial fans. This observation con¢rms a relatively positive ONd value for the Transhimalaya in Ladakh. On the basis of the Nd data alone it is not possible to resolve which of the possible end members may be contributing to the values seen in the Nurla and Choksti Formations, although the petrographic data does not favor involvement of the High Himalaya as the isotopically negative source. Consequently we employ single grain analyses to resolve which sources were mixing with Transhimalayan and/or ophiolitic sources in providing sediment to the basin in post-Chogdo times. 7. Pb isotopes of detrital feldspars Although the mineral assemblage observed in the Indus Molasse and the positive ONd values argue against the modern High Himalaya as a plausible source, further discrimination is di¤cult because the mineralogy is insu¤cient to distinguish between possible Transhimalayan, Lhasa Block and Kohistani sources. It is also possible that the isotopically negative Nd source needed 484 P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 to account for the evolving Nd compositions was in part from the Indian Zanskar Shelf, if not the High Himalaya. We employ the Pb isotope system applied to detrital K-feldspars in order to address these problems. This is done with the understanding that the Nd isotopic character is measured on very ¢ne clay, while the single grain Pb isotope work requires sand grade material. It is possible for the two grain sizes to have di¡erent sources, and this needs to be considered in the case of con£icting results. The Pb isotope character of detrital K-feldspars has previously been used as a provenance tool using conventional mass spectrometry methods [43]. In this study our data were obtained from individual K-feldspar grains and is compared with the distinct isotopic character of the High Himalaya, Transhimalaya, Kohistan/Dras Arc, Karakoram and Lhasa Block sources (e.g., [40,41,44^ 46]). In addition, we analyzed four grains from the Ladakh Batholith at Hunder (Fig. 1) in order to check that the range of Pb isotopic values measured further east on the Gandese Batholith was applicable to this region. For reference we also consider the composition of the ocean mantle using modern values for the Paci¢c and Indian Ocean mid ocean ridge basalt (MORB) as a proxy for Tethyan mantle [47,48]. We employ the newly developed technique of measuring Pb in situ [49] using a high-resolution Cameca 1270 ion microprobe of the Northeast National Ion Microprobe Facility (NENIMF) at WHOI. In order to exploit the potential of this method to characterize heterogeneous feldspar populations several analyses were run on di¡erent feldspars from a small number of samples, three from the Chogdo Formation, one from the Choksti Conglomerate and three from the upper Choksti Formation, including 23 analyses from a single sandstone (Table 2; Fig. 5). The sandstones were disaggregated and sieved, after which the size fraction 1 mm to 200 Wm, was mounted in epoxy and polished using aluminum oxide abrasives. The K-feldspar grains were then identi¢ed by area mapping of Al2 O3 and K2 O using the JEOL Superprobe electron microprobe at the Massachusetts Institute of Technology. This allowed the K-feldspars to be identi¢ed for isotopic analysis. After gold coating the grains were analyzed using a beam of negatively charged oxygen ions (O3 ) focused to a spot as small as 15^20 Wm. Analytical uncertainties are principally a re£ection of the counting statistics, typically averaging 2c = 1%. The analytical results are shown in Table 2. Analysis of K-feldspar standards verify that there is no signi¢cant mass fractionation e¡ect in analyzing Pb isotopes using the ion microprobe methodology compared to conventional mass spectrometry. In order to minimize the risk of secondary Pb contamination from sources outside the feldspar, analyses were made in the center of the grain, away from cracks, inclusions or alteration zones. In order to avoid any contamination that might have occurred during preparation of the grain mounts the beam was trained on the spot to be analyzed for 5 min before analysis began, so that any surface Pb contamination was removed. Through probing grain centers and allowing the beam to remove surface coating of the sectioned grains we avoid analysis of excess secondary Pb that is normally removed by leaching procedures prior to conventional mass spectrometry [44]. Fig. 5 shows the spread of measured isotopic ratios compared with those previously recorded from the central Himalaya^Tibet area [40,41,44^ 46]. The ¢elds de¢ned for the Indian Plate, Transhimalaya, Karakoram and Lhasa Block are also K-feldspar analyses, while those from the Kohistan/Dras Arc represent whole rock analyses from Pakistani exposures [39], since no K-feldspar data are available from this unit. The Spontang Ophiolite is not considered because there are no K-feldspar-rich lithologies in this unit. Although analytical uncertainties are higher for 204 Pb than other Pb isotopes, use of this isotope is crucial because it provides signi¢cant separation between the different sources considered, without which no discrimination can be achieved. Despite the di¤culty in analyzing the feldspars within the dominant microcrystalline granite clasts of the Chogdo Formation several reliable measurements were made. Several analyses were possible in samples from the Choksti Conglomerate and Choksti Formation, where abundant, large K-feldspar grains were identi¢ed. Some P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 485 Fig. 5. Pb isotopic discrimination diagrams showing the current measured range of Pb compositions for K-feldspars from the Lhasa Block, Transhimalaya and Indian Plate [44^46], and from whole rock values from the Kohistan Arc section of Pakistan [39]. Indian and Paci¢c MORB ¢elds from Sun [47] and Ben Othman et al. [48]. All analyses are shown with 1c uncertainties. 486 P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 K-feldspars were too Pb-poor to allow analysis. It is possible that our analyses provide a biased image of the feldspar population by only considering those with high Pb content. Alteration of the grains is not considered to be an important source of error, since the major element analytical totals were typically high ( s 98%), suggesting little alteration. 7.1. Interpretation of Pb data One noteworthy aspect to the data is that most of the analyses do not plot in any of the prede¢ned ¢elds, even when the 1c uncertainty is accounted for. Given the depositional setting and the constraints derived from the Nd isotope work it seems most likely that these data represent parts of the compositional range of the di¡erent sources that have not yet been identi¢ed by earlier studies. This is perhaps not surprising given that much of the earlier work was located in the central part of the Himalaya, remote from the Molasse basin in Ladakh. In addition, those studies only analyzed a modest number of grains from each block. The de¢ned range for the Karakoram is based on only four analyses [40]. Similarly, the plutonic parts of the Kohistan/Dras Arc are only exposed in Pakistan, and are only analyzed as whole rock data. In any case, given the paleocurrent data these exposures are most unlikely to be the source of the Indus Molasse sediments. The analyzed grains from the Ladakh Batholith at Hunder (Figs. 1 and 5D) show a range of values overlapping with, but trending to lower isotopic ratios than, the previously known Transhimalayan values [44]. These data demonstrate isotopic heterogeneity within the Transhimalaya and also shows that a local source is possible for those grains with low Pb isotopic ratios. The low Pb isotopic ratios seen in the Chogdo Formation rule out Indian plate rocks as important sources for these sandstones. The most likely source of the Chogdo Formation K-feldspars is the Ladakh Batholith. Although the paleo-current information and abundance of ophiolitic lithologies within the Chogdo Formation argues for sediment derivation from the SW, this does not rule out further input from the batholith. The lack of sources towards the SW that contain abundant K-feldspars also makes sole derivation of sediment from that direction unlikely. With the exception of a few grains, it is clear that the Indian Plate ¢eld is remote from nearly all the measured values from all formations, especially with respect to 207 Pb/204 Pb. This con¢rms other lines of evidence that suggest that this was not a major source of K-feldspar. Only a few of the highest ratios from the Choksti Formation and Conglomerate may be of Indian Plate origin, and even those are within error of the Karakoram and Lhasa Block. This conclusion is in accord with the relatively positive ONd values seen throughout the Indus Molasse. The Choksti Formation and Conglomerate appear to be mixtures of material from the Transhimalaya and a source with a more negative range of ONd , which cannot be Indian. The only possible such sources known are the Karakoram or Lhasa Block. The dominant NE to SW paleo-current directions favor involvement of the Lhasa Block rather than the Karakoram. Consequently, we conclude that those Pb analyses lying close to the Lhasa Block ¢eld de¢ned by Gariepy et al. [44] were derived from this terrane, albeit from areas with slightly di¡erent Pb isotopic values than those previously analyzed. Likewise, those feldspar analyses within the Choksti Formation that trend to lower 207 Pb/204 Pb values are taken to represent erosion from the Ladakh Batholith or other parts of the Transhimalaya exposed towards the east. For those grains whose analyses fall between the higher 207 Pb/204 Pb (Lhasa Block) and lower 207 Pb/204 Pb (Transhimalaya) grains the sources are less clear. Indeed, the overlap between measured basement values from these areas probably means that a unique de¢nition of the source may well be impossible for every grain even if truly comprehensive source compositional ranges could be de¢ned. Despite these shortcomings the ion probe data provide evidence that above the Chogdo Formation the clastic sediments entering the basin formed a mixture of material from the Transhimalaya and the Lhasa Block. Paleo-current data argue against signi¢cant involvement from the Karakoram or the Kohistan/Dras Arc. This means that while the lower Indus Molasse P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 487 represents local drainage in an intermontane setting, the upper part of the section is in£uenced by sediment transport over a regional scale, because the Lhasa Block is not exposed close to the study region. 8. Discussion The ion probe isotopic data combined with the geologic constraints and the clay Nd data provide an image of the erosion history of the early Himalaya. It is important to remember that erosion is not the same as tectonic uplift, since this can be triggered by other factors, such as climate change. For the detrital history to be used to track rates of denudation thermochronological work would need to be performed on individual grains. This approach is most e¡ective when depositional age can be compared with cooling age (e.g., [50]) to provide a minimum estimate of cooling rate, and this in turn can be used to constrain erosion rates. In the poorly dated Indus Molasse this approach may only be applicable in the Paleocene Chogdo Formation. 8.1. Uplift of the Lhasa Block It is noteworthy that there is a change in sediment provenance from the Paleocene Chogdo Formation to the upper parts of the section. We infer that this re£ects increased input from the Lhasa Block. The pre-collisional location of the Indus Molasse Basin on the active margin of Asia means that if the Lhasa Block had been a signi¢cant sediment source prior to collision this should have been recorded in the Chogdo Formation. Instead the continental and accreted arcs that lay along the active margin dominate the record. We thus tentatively suggest that the reason there was no early erosion of Lhasa Block was that it was not a signi¢cant topographic feature until after the collision, i.e. at least the Eocene. This conclusion ¢ts reconstructions that have proposed much of Tibet close to sealevel during the Cretaceous but with an elevated continental volcanic arc (Transhimalaya) along its south margin [51,52]. The Lhasa Block would then be progres- sively uplifted as crustal thickening propagated northwards after collision. It is important to realize that topography does not automatically imply rapid erosion, although it is rare to ¢nd rapid erosion without elevated topography. However, a close correspondence between exhumation rates derived from thermochronology and estimates of modern surface uplift derived from stream incision rates indicate a strong relationship between uplift and erosion in the modern Himalayas and Karakoram [53]. In the absence of other data we follow this logic in interpreting the erosion data presented here. The provenance data require increasing erosion of the Lhasa Block and the £ow of this material into the Indus Molasse Basin during the Eocene. This conclusion is at odds with alternative reconstructions that consider the Lhasa Block to have been highly elevated since the Late Cretaceous [54] because these do not provide a tectonic mechanism to preferentially increase the erosion of the Lhasa Block after collision. Provenance work on Cretaceous sequences from central southern Tibet, interpreted as forearc basin deposits (Xigaze Group), has identi¢ed the northern portion of the Lhasa Block as their source, implying that the Transhimalaya arc was subdued topographically at this time [55]. Although our study cannot clearly resolve this argument for areas located much further east, the new data can eliminate this possibility in Ladakh. There is strong evidence to support continuous erosion of the Transhimalaya throughout the collision, and with increasing in£uence up-section. This scenario is supported by reconstructions of the cooling history of the Ladakh Batholith based on thermochronology work [16,56], that indicate accelerated cooling during the Eocene (V45^ 52 Ma). This is presumably driven in part by tectonic uplift and erosion following collision. A similar accelerated uplift and unroo¢ng for the Lhasa Block would be consistent with this pattern. The implications of the Lhasa Block being uplifted early in the collision process are signi¢cant for models of strain accommodation and for climate^tectonic interactions. Our result is consistent with crustal thickening, i.e., horizontal short- 488 P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 ening, being the ¢rst reaction of the Asia margin following collision with India. Unless large-scale regional £exure or magmatic underplating can be invoked, crustal thickening and the following isostatic reaction is the only common mechanism for generating elevated topography of the type inferred. Although early uplift need not preclude tectonic extrusion of Indochina or Tibet [57], many reconstructions that emphasize extrusion as an important method of strain accommodation have suggested that little crustal thickening occurs in Tibet and the Himalaya during the Eocene and Oligocene (e.g., [58]). Such models do not predict the increasing uplift and erosion of the Lhasa Block recorded by the Indus Molasse. 8.2. Origin of the Indus River The evolving provenance and paleo-current data suggest that the Indus River may have initiated soon after collision, in practice during Nurla Formation times (post-latest Paleocene). The change in paleo-£ow towards the west, and importantly the in£ux of detritus from the Lhasa Block is suggestive of an axial river, draining the western part of the plateau. In practice this is what the modern Indus does today. Our result is incompatible with the model of Sinclair and Ja¡ey [59], who suggested an internally drained Indus Molasse Basin, pre-dating Indus River initiation. However, the absence of an appropriate source with the necessary isotopic characteristics and the observed change in paleo-£ow within the basin now make this model untenable. We suggest that the change between Chogdo and Nurla Formations represents the change from a basin dominated by local drainage to one of regional drainage. In this scenario the Indus has remained stationary since at least the Middle Eocene, despite the subsequent horizontal compression that inverted the Indus Molasse Basin. Such a timing for Indus River initiation is consistent with the apparent start of Indus deltaic sedimentation in the Katawaz Basin of Pakistan close to the Paleocene^Eocene boundary [60], as well as the evidence of Indus Suture Zone material reaching the Arabian Sea at least by the Middle Eocene [61]. The presence of such grains in the earliest parts of the Indus Fan requires a long river system, analogous to the modern Indus. 9. Conclusions This study demonstrates that despite signi¢cant analytical uncertainties the in situ analysis of Pb isotopes within single K-feldspar grains is e¡ective at constraining provenance in tectonically complex areas when used in conjunction with bulk mineral analyses, such as the Nd work on clays presented here. The approach allows end members to mixed sedimentary sequences to be constrained. Whole rock analyses only provide an average measurement of source composition. Without such single grain work it is not possible to know how many sources are involved, and in settings like the Indus Molasse it may not be possible to single out which source is acting as an end member if data from only one isotopic system is used. This type of data are important to reconstructing erosion patterns and as an important precursor before attempting thermochronology work on detrital minerals. Only single grain work is able to provide the resolution required to provide unambiguous identi¢cation of sediment sources. In the case of the Indus Molasse the isotopic provenance work shows the lack of input from the Indian Plate and demonstrates an evolution from a Paleocene (Chogdo Formation) basin dominated by erosion of Tethyan ophiolites (including Spontang) and the Transhimalaya to one incorporating signi¢cant detritus from the Lhasa Block (Nurla and Choksti Formations). The onset of erosion of the Lhasa Block is suggestive of its uplift following India^Asia collision and that this area was not signi¢cantly elevated before the Eocene. The uplift accompanies, or may even cause, the initiation of the Indus River, following the suture, a location it continues to occupy to the present day. Acknowledgements P.C. thanks JOI/USSAC and WHOI for ¢nan- P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 489 cial support to perform ¢eldwork in the Indus Suture and for some analytical support. P.C. is indebted to M.P. Searle for his introduction to the geology of the suture zone and to Fida Hussein Mittoo of Leh and Rockland Tour and Trek for all their help. M.P. Searle and Y.M.R. Najman are thanked for their advice on Himalayan erosion. J.P. Burg, C. France-Lanord and an anonymous reviewer provided helpful reviews that improved the quality of the work. The NENIMF at WHOI is supported by Grant EAR-9904400 from the National Science Foundation. This is WHOI contribution 10457.[AH] [11] [12] [13] [14] References [1] I.M. Held, Climate models and the astronomical theory of the ice ages, Icarus 50 (1982) 449^461. [2] W.F. Ruddiman, M.E. Raymo, H.H. Lamb, J.T. Andrews, Northern Hemisphere climate regimes during the past 3 Ma; possible tectonic connections; discussion, in: N.J. Shackleton, R.G. West, D.Q. Bowen (Eds.), Cambridge University Press, Cambridge, 1988, pp. 1^20. [3] M.E. Raymo, W.F. Ruddiman, Tectonic forcing of the late Cenozoic climate, Nature 359 (1992) 117^122. [4] S. Manabe, T.B. Terpstra, The e¡ects of mountains on the general circulation of the atmosphere as identi¢ed by numerical experiments, J. Atmos. Sci. 31 (1974) 3^42. [5] W.L. Prell, J.E. Kutzbach, Sensitivity of the Indian monsoon to forcing parameters and implications for its evolution, Nature 360 (1992) 647^651. [6] F. Fluteau, G. Ramstein, J. Besse, Simulating the evolution of the Asian and African monsoons during the past 30 Myr using an atmospheric general circulation model, J. Geophys. Res. 104 (1999) 11995^12018. [7] D.W. Burbank, R.A. Beck, T. Mulder, The Himalayan foreland basin, in: A. Yin, T.M. Harrison (Eds.), The Tectonic Evolution of Asia, Cambridge University Press, Cambridge, 1996, pp. 149^188. [8] P. Copeland, T.M. Harrison, M.T. Heizler, 40 Ar/39 Ar single-crystal dating of detrital muscovite and K-feldspar from leg 116, southern Bengal fan: implications for the uplift and erosion of the Himalayas, Proc. Ocean Drill. Prog. Sci. Res. 116 (1990) 93^114. [9] C. France-Lanord, L. Derry, A. Michard, Evolution of the Himalaya since Miocene time: isotopic and sedimentological evidence from the Bengal Fan, in: P.J. Treloar, M.P. Searle (Eds.), Himalayan Tectonics, Geol. Soc. Lond. Spec. Publ. 74, 1993, pp. 603^622. [10] M.P. Searle, Cooling history, erosion, exhumation and kinematics of the Himalaya^Karakorum^Tibet orogenic belt, in: A. Yin, T.M. Harrison (Eds.), The tectonic evo[15] [16] [17] [18] [19] [20] [21] [22] [23] lution of Asia, Cambridge University Press, Cambridge, 1996, pp. 110^137. S. Tonarini, I.M. Villa, F. Oberli, Eocene age of eclogite metamorphism in Pakistan Himalaya; implications for India^Eurasia collision, Terra Nova 5 (1993) 13^20. P.J. Treloar, D.C. Rex, P.G. Guise, M.P. Coward, M.P. Searle, B.F. Windley, M.G. Petterson, M.Q. Jan, I.W. Lu¡, K^Ar and Ar^Ar geochronology of the Himalayan collision in NW Pakistan constraints on the timing of suturing, deformation, metamorphism and uplift, Tectonics 8 (1989) 881^909. J.D. Walker, M.W. Martin, S.A. Bowring, M.P. Searle, D.J. Waters, K.V. Hodges, Metamorphism, melting, and extension; age constraints from the High Himalayan slab of Southeast Zanskar and Northwest Lahaul, J. Geol. 107 (1999) 473^495. Y.M.R. Najman, P.D. Clift, M.R.W. Johnson, A.H.F. Robertson, Early stages of foreland basin evolution in the Lesser Himalaya, N. India, in: P.J. Treloar, M.P. Searle (Eds.), Himalayan Tectonics, Geol. Soc. Lond. Spec. Publ. 74, 1993, pp. 541^558. ¨ A.H.F. Robertson, Formation of melanges in the Indus Suture Zone, Ladakh Himalaya by successive subductionrelated, collisional and post-collisional processes during Late Mesozoic-Late Tertiary time, in: M.A. Khan, P.J. Treloar, M.P. Searle, M.Q. Jan (Eds.), Tectonics of the Nanga Parbat Syntaxis and the Western Himalaya, Geol. Soc. Lond. Spec. Publ. 170, 2000, pp. 333^374. P.D. Clift, A. Carter, M. Krol, E. Kirby, Evidence for ophiolite obduction and continental collision within the Indus Molasse Group, Ladakh Himalayas, India, J. Geol. Soc. Lond., in press. E. Garzanti, T. van Haver, The Indus clastics: forearc basin sedimentation in the Ladakh Himalaya (India), Sediment. Geol. 59 (1988) 237^249. © U. Scharer, R.H. Xu, C.J. Allegre, U^(Th)^Pb systematics « and ages of Himalayan leucogranites, South Tibet, Earth Planet. Sci. Lett. 77 (1986) 35^48. R. Weinberg, The disruption of a diorite magma pool by intruding granite; the Sobu Body, Ladakh Batholith, Indian Himalayas, J. Geol. 105 (1997) 87^98. A. Steck, L. Spring, J.-C. Vannay, H. Masson, H. Bucher, E. Stutz, R. Marchant, J.-C. Tieche, The tectonic evolution of the Northwestern Himalaya in eastern Ladakh and Lahaul, India, in: P.J. Treloar, M.P. Searle (Eds.), Himalayan Tectonics, Geol. Soc. Lond. Spec. Publ. 74, 1993, pp. 265^276. P.D. Clift, J. Blusztajn, M. Krol, A. Carter, E. Kirby, O. Green, Timing of Collision and Patterns of Early Uplift along the Indus Suture, Ladakh Himalaya, India, Eos, Am. Geophys. Union Abstr., 1999. M.P. Searle, K.T. Pickering, D.J.W. Cooper, Restoration and evolution of the intermontane Indus molasse basin, Ladakh Himalaya, India, Tectonophysics 174 (1990) 301^ 314. M.E. Brook¢eld, C.P. Andrews-Speed, Sedimentology, petrography and tectonic signi¢cance of the shelf, £ysch 490 P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 and molasse clastic deposits across the Indus suture zone, Ladakh, NW India, Sediment. Geol. 40 (1984) 249^286. T. van Haver, Etude stratigraphique, sedimentologique et structurale d'un bassin d'avant arc: exemple du bassin de l'Indus, Ladakh, Himalaya, Ph.D. Thesis, Univ. Grenoble, Grenoble, 1984, pp. 1^204. M.P. Searle, R.R. Parrish, K.V. Hodges, A. Hurford, M.W. Ayers, M.J. Whitehouse, Shisha Pangma leucogranite, South Tibetan Himalaya: Field relations, geochemistry, age, origin, and emplacement, J. Geol. 105 (1997) 295^317. R.I. Cor¢eld, M.P. Searle, O.R. Green, Photang thrust sheet; an accretionary complex structurally below the Spontang Ophiolite constraining timing and tectonic environment of ophiolite obduction, Ladakh Himalaya, NW India, J. Geol. Soc. Lond. 156 (1999) 1031^1044. D.J. DePaolo, G.J. Wasserburg, Nd isotopic variations and petrogenetic models, Geophys. Res. Lett. 3 (1976) 249^252. P.J. Hamilton, R.K. O'Nions, D. Bridgewater, A.P. Nutman, Sm^Nd studies of Archean metasediments and metavolcanics from west Greenland and their implication for the earth's early history, Earth Planet. Sci. Letts. 62 (1983) 263^272. C. Deniel, P. Vidal, A. Fernandez, P. Le Fort, Isotopic study of the Manaslu Granite (Himalaya, Nepal): inferences of the age and source of Himalayan Leucogranites, Contrib. Mineral. Petrol. 96 (1987) 78^92. A. Bouquillon, C. France-Lanord, A. Michard, J. Tiercelin, Sedimentology and isotopic chemistry of the Bengal Fan sediments: the Denudation of the Himalaya, Proc. Ocean Drill. Prog. Sci. Res. 116 (1990) 43^58. A. Galy, C. France-Lanord, L.A. Derry, The late Oligocene^early Miocene Himalayan belt; constraints deduced from isotopic compositions of early Miocene turbidites in the Bengal Fan, Tectonophysics 260 (1996) 109^118. L.A. Derry, C. France-Lanord, Neogene Himalayan weathering history and river 87 Sr/86 Sr; impact on the marine Sr record, Earth Planet. Sci. Lett. 142 (1996) 59^74. S.M. McLennan, M.T. McCulloch, S.R. Taylor, J.B. Maynard, E¡ect of sedimentary sorting on Neodymium isotopes in deep-sea turbidites, Nature 33 (1989) 547^549. S.L. Goldstein, R.K. O'Nions, P.J. Hamilton, A Sm^Nd isotopic study of atmospheric dusts and particulates from major river systems, Earth Planet. Sci. Lett. 70 (1984) 221^236. C. France-Lanord, P. Le Fort, Crustal melting and granite genesis during the Himalayan collision orogenesis, Trans. R. Soc. Edinb. 79 (1988) 183^195. A.C. Pierson-Wickmann, L. Reisberg, C. France-Lanord, The Os isotopic composition of Himalayan river bedloads and bedrocks; importance of black shales, Earth Planet. Sci. Lett. 176 (2000) 203^218. © C.J. Allegre, D. Ben Othman, Nd^Sr isotopic relationship in granitoid rocks and continental crust development a chemical approach to orogenesis, Nature 286 (1980) 325^342. [38] M.G. Petterson, M.B. Crawford, B.F. Windley, Petrogenetic implications of neodymium isotope data from the Kohistan Batholith, North Pakistan, J. Geol. Soc. Lond. 150 (1993) 125^129. [39] M.A. Khan, R.J. Stern, R.F. Gribble, B.F. Windley, Geochemical and isotopic constraints on subduction polarity, magma sources and palaeogeography of the Kohistan intra-oceanic arc, northern Pakistan Himalayas, J. Geol. Soc. Lond. 154 (1997) 935^946. [40] U. Scharer, P. Copeland, T.M. Harrison, M.P. Searle, « Age, cooling history and origin of post-collisional leucogranites in the Karakoram batholith: a multi-system isotope study, J. Geol. 98 (1990) 233^251. [41] F. Debon, P. Le Fort, S.M.F. Sheppard, J. Sonet, The four plutonic belts of the Transhimalaya^Himalaya; a chemical, mineralogical, isotopic, and chronological synthesis along a Tibet^Nepal section, J. Petrol. 27 (1986) 219^250. [42] M.P. Searle, A.J. Rex, R. Tirrul, D.C. Rex, A.C. Barnicoat, B.F. Windley, Metamorphic, magmatic, and tectonic evolution of the central Karakoram in the Biafo^Baltoro^ Hushe regions of northern Pakistan, in: L.L. Malinconico, R.J. Lillie (Eds.), Tectonics of the Western Himalayas, Geol. Soc. Am., Spec. Pap. 232, 1989, pp. 47^73. [43] D.K. McDaniel, S.R. Hemming, S.M. McLennan, G.N. Hanson, Petrographic, geochemical, and isotopic constraints on the provenance of the early Proterozoic Chelmsford Formation, Sudbury Basin, Ontario, J. Sediment. Res. 64 (1994) 362^372. © [44] C. Gariepy, C.J. Allegre, R.H. Xu, The Pb-isotope geochemistry of granitoids from the Himalaya^Tibet collision zone: implications for crustal evolution, Earth Planet. Sci. Lett. 74 (1985) 220^234. [45] P. Vidal, A. Cocherie, P. Le Fort, Geochemical investigations of the origin of the Manaslu leucogranite (Himalaya, Nepal), Geochim. Cosmochim. Acta 46 (1982) 2279^2292. [46] R.R. Parrish, K.V. Hodges, A. MacFarlane, U^Pb geochronology of igneous and metamorphic rocks near the Main Central Thrust in the Langtang area, central Nepal Himalaya, Abstr. 7th Himalaya^Karakorum^Tibet workshop, Oxford Univ., 1992, pp. 67^68. [47] S.S. Sun, Lead isotopic study of young volcanic rocks from mid-ocean ridges, ocean islands and island arcs, Phil. Trans. R. Soc. London Ser. A 297 (1980) 409^ 445. [48] D. Ben Othman, W.M. White, J. Patchett, The geochemistry of marine sediments, island arc magma genesis, and crust^mantle recycling, Earth Planet. Sci. Lett. 94 (1989) 1^21. [49] G.D. Layne, N. Shimizu, Measurement of lead isotope ratios in common silicate and sul¢de phases using the Cameca 1270 Ion Microprobe, in: G. Gillen et al. (Eds.), Secondary Ion Mass Spectrometry SIMS XI, John Wiley, New York, 1998, pp. 63^65. [50] P. Copeland, T.M. Harrison, Episodic rapid uplift in the Himalaya revealed by 40 Ar/39 Ar analysis of detrital K- [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] P.D. Clift et al. / Earth and Planetary Science Letters 188 (2001) 475^491 feldspar and muscovite, Bengal fan, Geology 18 (1990) 354^357. P. England, G. Houseman, Finite strain calculations of continental deformation 2/Comparison with the India^ Asia collision zone, J. Geophys. Res. 94 (1986) 17,561^ 17,569. P. England, M.P. Searle, The Cretaceous^Tertiary deformation of the Lhasa Block and its implications for crustal thickening in Tibet, Tectonics 5 (1986) 1^14. D. Burbank, J. Leland, M. Reid, R. Anderson, M. Ca¡ee, R. Finkel, Bedrock incision and uplift rates along the Indus River, N. Pakistan, Abstr. Geol. Soc. Am. 26 (1994) 208. M.A. Murphy, A. Yin, T.M. Harrison, S.B. Durr, Z. « Chen, F.J. Ryerson, W.S.F. Kidd, X. Wang, X. Zhou, Did the Indo-Asian collision alone create the Tibetan Plateau, Geology 25 (1997) 719^722. S.B. Durr, Provenance of Xigaze fore-arc basin clastic « rocks (Cretaceous, south Tibet), Geol. Soc. Am. Bull. 108 (1996) 669^691. R.B. Sorkhabi, A.K. Jain, S. Nishimura, T. Itaya, N. Lal, R.M. Manickavasagam, T. Tagami, New age constraints 491 [51] [57] [52] [53] [58] [54] [59] [55] [56] [60] [61] on the cooling and unroo¢ng history of the Trans-Himalayan Ladakh Batholith (Kargil area), N.W. India, Proc. Indian Acad. Sci. 103 (1994) 83^97. P. Tapponnier, G. Peltzer, R. Armijo, On the mechanics of the collision between India and Asia, in: M.P. Coward, A.C. Ries (Eds.), Collision Tectonics, Geol. Soc. London Spec. Publ. 19, 1986, pp. 115^157. T.M. Harrison, A. Yin, F.J. Ryerson, Orographic evolution of the Himalaya and Tibetan Plateau, in: T.J. Crowley, K.C. Burke (Eds.), Tectonic Boundary Conditions for Climate Reconstructions, Oxford University Press, Oxford, 1998, pp. 39^72. H.D. Sinclair, N. Ja¡ey, Sedimentology of the Indus Group, Ladakh, northern India; implications for the timing of initiation of the palaeo-Indus River, J. Geol. Soc. Lond. 158 (2001) 151^162. M. Qayyum, R.D. Lawrence, A.R. Niem, Discovery of the palaeoIndus delta-fan complex, J. Geol. Soc. Lond. 154 (1997) 753^756. P.D. Clift, C. Gaedicke, N. Shimizu, G. Layne, M. Clark, 55 million years of Tibetan and Karakoram evolution recorded in the Indus Fan, EOS 81 (2000) 277^282.