Bengal Fan Research Articles

Fifty‐five million years of Tibetan evolution recorded in the Indus Fan

Although the Indus Fan is only about one‐third of the volume of its giant neighbor in the Bay of Bengal, it is one of the largest sediment bodies in the ocean basins, totaling ∼5×106 km3. Its detrital sedimentary record is an important repository of information on the uplift and erosion of the western Himalaya. New seismic and provenance data from the Pakistan margin now suggest that the Indus River and fan system was initiated shortly after the India‐Asia collision at ˜5 Ma.The modern Indus drainage basin is dominated by the high peaks of the Karakoram, Kohistan, and other tectonic units of the Indus Suture Zone rather than the High Himalaya. The Indus River, which rises in western Tibet near Mount Kailas, follows the Indus Suture Zone along strike before cutting orthogonally through the Himalaya to the Arabian Sea. The other tributaries to the Indus, such as the Chenab and Sutlej, do drain the crystalline High Himalayan range, but do so in an area where its topography is much reduced (Figure 1). In contrast, the Bengal Fans main feeder rivers, the Ganges and Brahmaputra, follow the High Himalaya along strike for much of the length of the orogen. In practice, this means that the Bengal Fan is swamped by the large volume of material derived from the rapidly unroofing High Himalaya [France‐Lanord et al, 1993], while the Indus Fan is dominated by tectonic units adjacent to the suture zone, including western Tibet. This allows their erosional signal to be more readily isolated in the Indus Fan compared to in the Bengal.

The significance of large sediment supply, active tectonism, and eustasy on margin sequence development: Late Quaternary stratigraphy and evolution of the Ganges–Brahmaputra delta

Abstract Borehole data from the Ganges–Brahmaputra delta system unveil an intriguing Late Quaternary history controlled by immense sediment discharge, tectonics, and eustasy. Sea level first intersected a major portion of the lowstand surface ∼10,000–11,000 cal yr BP, marking the onset of sediment trapping and delta growth. Despite rapid sea-level rise (>1 cm/yr), sediment load was sufficient to maintain relative shoreline stability during this time. By ∼7000 cal yr BP, accommodation was quickly infilled with slowed sea-level rise, leading to upstream channel migration and widespread dispersal of sands. This forced coastal progradation along the western delta front where the Ganges was situated, and much of the river's fine-grained discharge bypassed the subaerial delta and formed a prograding deltaic clinoform on the shelf. Concurrently, Brahmaputra sediments were sequestered to an inland tectonic basin, thus starving the adjacent shoreline and leading to transgression along the eastern delta. By ∼5000 cal yr BP, the two rivers had migrated or changed course to discharge along the eastern delta shoreline which began to prograde into what is now the modern river-mouth estuary. Present delta configuration appears to have developed by ∼3000 cal yr BP. Evolution of the tectonically active Ganges–Brahmaputra (G–B) system reveals important similarities and differences with other deltas. Overall facies succession follows that of basic models, progressing from an alluvial valley to coastal marine delta front to a prograding subaerial delta plain. However, the timing, thickness, and controls of these deposits differ. Immense sediment discharge from the Ganges and Brahmaputra rivers allowed initial delta growth 2000–3000 yr earlier than the global average, despite debouching onto a canyon-incised, high-energy margin. Subsequently, a thick (∼50 m) transgressive systems tract was formed during the early Holocene phase of delta aggradation, contrasting with extensive shoreline retreat documented along most margins. Highstand progradation of coastal and shelf sequences ensued by the middle Holocene, but subsiding inland basins also favored accumulation of a thick (to 40 m) highstand sequence in the lower floodplains. Unlike many other delta systems, subsidence in the subaerial G–B delta is not dominated by compaction, but rather by plate-driven tectonic processes that generate rates up to 4 mm/year. Overall, the huge sediment load, tectonic subsidence, major seismic events, and a nearshore canyon system have led to widespread sediment dispersal and sequence formation across the subaerial delta, shelf, and deep-sea Bengal Fan throughout the Late Quaternary.

The Os isotopic composition of Himalayan river bedloads and bedrocks: importance of black shales

Re–Os data are presented for bedrocks and bedloads of selected rivers from the Himalayan range, in Central Nepal and Bangladesh. These data demonstrate that the Os isotopic composition of most of the bedrocks is similar to that of the average continental crust. Notable exceptions to this are the black shales from the Lesser Himalaya (LH), which contain large quantities of radiogenic osmium. In the black shales, the Os is not concentrated in the sulfides, and may instead be associated with organic matter. Bedloads collected from rivers which include LH black shales in their catchment basins have notably high Os ratios ( 187Os/ 188Os≈3). This radiogenic signature persists into the Ganga river. In contrast, the Brahmaputra river is less radiogenic, suggesting an ophiolite input in the headwater region. Simple physical mixtures of LH black shales with eroded products from the high range require unreasonably high proportions of black shale to explain the high Os ratios observed at the outflow of the range. Assuming that our bedrock and bedload samples are representative of the source region and the eroded mixture, respectively, this observation suggests that erosion may not be occurring at steady-state. In addition, mass balance calculations indicate that about 20% of the Os and 60% of the Re are carried in dissolved form, implying that chemical erosion plays an important role in the osmium budget. Finally, the contrast in Os isotope ratio between Ganga bedloads and bulk sediments in the Bengal Fan suggests that radiogenic Os carried by particulate matter and later remobilized may significantly influence the Os composition of seawater.

Contribution of grain size and compositional data from the Bengal Fan sediment to the understanding of Toba volcanic event

Abstract Toba Tephra layers in cores from the Bengal Fan have been studied to characterize their granulometric, mineralogic and geochemical features. The layers are silt-sized sediments with no significant non-volcanic contamination. The samples are very homogeneous: glass shards have moderate hydration (about 6% water), a calc-alkaline character and a rhyolitic composition. The major and trace element compositions indicate an orogenic affinity and coincide with literature data for the youngest Toba Tephra level (YTT, 75 ka). X-ray diffraction (XRD) data indicate that the glass is the dominant phase, with minor feldspar, biotite and quartz. The grain size analyses of the beds show discontinuous vertical variation and no linear trends from south to north over about 1500 km. The vertical granulometric gap of the layers probably represents a single pyroclastic emplacement unit subjected to pre-depositional particle aggregation.

Abandoned Paleocene spreading center in the northeastern Indian Ocean: evidence from magnetic and seismic reflection data

Abstract Magnetic data from the northeastern Indian Ocean reveal east–west trending magnetic lineations 28 through 34, an abandoned spreading center (ASC) and part of the Cretaceous Magnetic Quiet Zone (CMQZ), which are used to reconstruct past tectonic processes involved in the evolution of the ocean. Offsets in the magnetic lineations outline the existence of four N–S trending fracture zones: 80°E, Indira, 84.5°E, and 86°E. The 84.5°E Fracture Zone (FZ) acts as a boundary separating oceanic crust with different magnetic isochron patterns. The pairs of magnetic lineations 30 through 32n.2 between the 86°E FZ and the Ninetyeast Ridge reveal an ASC of about 65 Ma age, parallel to ≈0.5°S latitude. The structure of the ASC is well-identified in seismic reflection data as an undulating basement topographic rise covered by up to 2 km of Bengal Fan sediments. The spreading center might have initiated its activity along with other spreading centers of the Wharton Ridge and India–Antarctica Ridge, after the first major plate reorganization of the Indian Ocean (about 95±5 Ma), but ceased shortly after formation of anomaly 30 (about 65 Ma). Then it jumped southward between anomalies 32n.2 and 33. The jump captured the oceanic crust of anomalies 30 through 32n.2 that initially formed on the Antarctica plate, but were later transferred to the Indian plate. The spreading centers in the vicinity of the Ninetyeast Ridge jumped several times southward, to maintain a close proximity to the Kerguelen hotspot.

Influence of sedimentation on enrichment of manganese and growth of ferromanganese micronodules, Bengal Fan, India

Based upon down core variations in O/Mn ratio in sediment and in micronodules, Mn content, organic carbon and abundance of micronodules in two box cores from high and low sedimentation regions of the Bay of Bengal, 0–20 and 0–40 cm levels were identified as “Mn oxidation zone” in the western and the eastern regions of the bay. Despite similarity in bottom water O 2 and organic carbon contents of the region, a higher abundance of micronodules down to the deeper levels was observed in the area of low sedimentation. But, in the upper surfaces increased enrichment of Mn was observed in core 1 located in the area of higher sediment accumulation. Rapid recycling of Mn (faster burial–transfer to Mn reduction zone–Mn dissolution and its upward diffusion–re-precipitation) increasing upward flux of Mn(II) and associated higher consumption of O 2 limits the width of Mn oxidation zone and appears to be responsible for enhanced enrichment of Mn in the area of high sedimentation. The results of the present study also suggest that though the availability of Mn controls the formation of micronodules, their resident time in Mn oxic zone regulates their growth and influences their morphology.

Surface wave tomography studies beneath the Indiansubcontinent

The Rayleigh wave group velocities at periods of 5 to 60 s across more than eightypropagation paths of the Indian subcontinent are used to construct estimated Rayleigh wavegroup velocity distribution maps of the region. Backus and Gilberts inversion theory is appliedfor the surface wave tomography studies. High values of Rayleigh wave group velocity areconcentrated at periods of 30 to 60 s near the region of the Precambrian shield of central India,the China–East India border, the Pakistan–Kashmir border, the Indo-Burmese arc and the westNepal–Tibet border regions. Low values of group velocity are located at periods of 10 to 20 s inthe northernmost part of India beneath the Bangladesh shelf and beneath the northernmost part ofthe Bay of Bengal Fan and the Arabian Sea, which may be due to the presence of thick sedimentsin the region. Higher group velocities at periods of 40 to 60 s around the region of the WestIndia–Pakistan border suggest thinning of the Indian lithosphere beneath the region. The natureof the high Rayleigh wave group velocities at periods of 30 to 60 s in the Peninsular shield andthe Himalaya indicates the continuation of shield type structure beneath the Himalaya. Thenature of the Rayleigh wave group velocities in the Bay of Bengal indicates the continuation ofthe Ninetyeast Ridge up to the Bangladesh shelf below a depth of 25 km.

Provenance of early Palaeozoic sandstones, southeastern Australia, Part 1: vertical changes through the Bengal fan-type deposit

Abstract Ordovician strata of the Lachlan Fold Belt in southeastern Australia are dominated by a turbidite succession that in its undeformed state was comparable in size to the modern-day Bengal Fan. The ancient fan was derived from the northern part of the Cambrian Ross–Delamerian orogenic belt that formed at a convergent boundary between plates containing Gondwanaland and part of the palaeo-Pacific Ocean, respectively. Substrate to the submarine fan consisted of mafic volcanic rocks with an immature island arc affinity similar to the Eocene Izu–Mariana–Bonin system. Sandstone compositions, determined from several stratigraphically constrained sections, demonstrate a uniform framework composition dominated by monocrystalline quartz with less abundant feldspar, polycrystalline quartz, metamorphic rock fragments, sedimentary rock fragments and mica. These indicate derivation from a low-grade metamorphic terrain. Up-sequence variation is limited to decreasing feldspar and lithic fragments and increasing quartz, indicating increased weathering and lowering of topography, in the source terrain over time. 25% of counted sandstones in the Late Ordovician interval contain a small component of mafic–intermediate volcanic rock fragments that accompanied a change in palaeocurrent direction from eastwards to northwards. In the Early Silurian, the Ordovician sequence was disrupted by the formation of several major deformation zones. Between these zones, developed several deep-marine sedimentary basins that contain a continuous succession from the Ordovician into the Early Silurian. Early Silurian sandstone is similar in composition to underlying Early Ordovician rocks except that the feldspar component is generally absent. Early Silurian sandstone was derived from uplifted portions of the Ordovician sequence and includes texturally inverted types with sandstone containing a bimodal mixture of coarse well-rounded quartz and fine angular quartz.

Evolution of fluvial style in the Siwalik Group in the foothills of the Nepal Himalaya

Abstract Middle Miocene to Pleistocene fluvial sediments of the Siwalik (Churia) Group are widely distributed in the southern frontal area of the Himalaya. The succession is about 6 km thick, and was derived from denudation of the Himalayan orogen. The Siwalik Group in Nepal is well exposed in the Surai Khola area (western Nepal), and the Hetauda–Bakiya Khola area (central Nepal). The group is separated into northern and southern belts by the Central Churia Thrust (CCT). Eight facies associations (FA1 to FA8) are recognized in the Siwalik Group in these areas. They are interpreted as the deposits of fine-grained meandering, flood-flow-dominated fine-grained meandering, sandy meandering, deep sandy braided, comparatively shallow sandy braided, anastomosed, gravelly braided, and debris-flow-dominated braided systems, respectively. FA6, FA7 and FA8 occur only in the southern belt. In each area, the sedimentary succession generally coarsens upwards. The accumulation of these facies associations, related to the paleomagnetic time frame, indicates that flooding increased dramatically from about 10.5 to 9.5 Ma, and fluvial style changed from meandering to braided between 9.0 and 6.5 Ma. A gravelly fluvial system prevailed after 3.0 to 2.5 Ma. The evolution of these fluvial styles is intimately related to the uplift of the Himalaya and associated thrust movements, and consequent effects on atmospheric circulation and precipitation. By comparison with the characteristics of the submarine Bengal Fan deposits, it is apparent that an increase in flood flow is strongly influenced by increased precipitation due to onset and intensification of monsoon climate. The sandy braided system was mainly induced by regional tectonic uplift, and the gravelly fluvial system may have been formed due to regional thrust movements along the Himalaya frontal area.

Subsea Channels and Incidence of Thermogenic Hydrocarbons in the Mid-proximal Bengal Fan, West of the Andaman-Nicobar Islands

Mid-fan areas in the Bay of Bengal to the west of the Andaman-Nicobar Islands and to the east of the 90 E ridge display remarkable channel-levee and interlevee characteristics over the seabed. These subsea channels at 3,000 m or more water depth have specific orientations along NNE or NNW. Most of these major and minor channels are at present inactive, except one running along the eastern part of the area following the reported subduction suture (Currey et al., 1979). The other channels in the western part retain the configuration of a floor-bank combination but are covered with a thin deposition of pelagic organisms without any significant terrigenous inputs to a depth of around 25 cm below the seabed. On the other hand, the active channel to the east is filled with continuous hemipelagic to terrigenous inputs up to at least 1.5 m bsf. The pelagic organisms of the upper 25 cm of seabed over the inactive channels as well as in their banks and interlevee areas are mainly represented by foraminiferal tests and thus constitute foraminiferal ooze. Most of the seafloor is covered by such ooze except in the easternmost part, where the active channel brings a terrigenous influx in the form of turbidites. The fan deposition has been restricted to the easternmost active channel since the early Holocene. The age of the pelagites that constitute the 25-cm top sediment slice is assumed to represent about 8,000 years, calculated at the rate of deposition of 3 cm 1,000 years (Kennet, 1982). Thus, most of the western seabed has not experienced any fan deposition during the last 8,000 years. However, below 25 cm, the entire area is seen to have experienced fan deposition as evidenced by occurrence of terrigenous silt and clay. Significantly, incidence of thermogenic hydrocarbons is proved in the levee-in terlevee complexes of the central part. Being closer to the subduction zone, gas seepages through pervasive fissures are perhaps more pronounced here.

A paleo-Brahmaputra? Subsurface lithofacies analysis of Miocene deltaic sediments in the Himalayan–Bengal system, Bangladesh

The Bengal foreland basin contains a succession of up to 16+ km of dominantly deltaic deposits, eroded from the eastern Himalayas and the Indo–Burman ranges and carried by major river systems similar to the present-day Ganges and Brahmaputra. Analysis of electric logs and core descriptions acquired during oil and gas exploration in Bangladesh allows construction of lithofacies maps, which constrain depositional patterns of Miocene strata. Compilations of sand thickness and sand/shale ratio of the Miocene Surma Group show that Lower to Middle Miocene strata of the Bhuban Formation accumulated in a large, elongate trough. Sand thickness and percentage both decrease markedly away from this depocenter, which describes a large-scale bend, running initially westward from Rashidpur (northeast Bengal basin) and curving southward toward the Bengal fan. Middle to Upper Miocene strata of the Boka Bil Formation show a similar geographic trend in deposition of coarsest and thickest sediment, but the major depocenter had shifted northward relative to that of the Bhuban Formation by some 30 km, passing near Beani Bazar. These trends suggest that deltaic deposits of the Surma Group filled the Sylhet trough of the northeast Bengal basin from the east. Published seismic data from western Bangladesh show that additional large channels also contributed materials to the Bengal basin from the northwest during the Late Miocene, but these channels resulted in very little accumulation in the northwestern part of the basin, probably due to restricted subsidence of underlying continental crust. This study suggests that there was a major drainage system similar to the modern Brahmaputra River during Miocene time, which carried orogenic sediments eroded from the uplifted terranes of the eastern Himalayas and Indo–Burman ranges to the eastern Bengal delta.

Holocene and modern sediment budgets for the Ganges-Brahmaputra river system: Evidence for highstand dispersal to flood-plain, shelf, and deep-sea depocenters

The partitioning of fluvial sediment load across continental margins is an important control on strata formation and sequence development; however, few quantitative sediment budgets that encompass entire dispersal systems exist. For the Ganges-Brahmaputra river system, sediment discharge is estimated to be 10 9 t/yr at gauging stations ~300 km inland of the coast, but little has been known of the downstream fate of this material. Geochronological, geophysical, and stratigraphic investigations of the lowland flood plain, delta plain, and shelf help to delineate the extent of Holocene fill and allow calculation of a first-order sediment budget. Results reveal that 1500 ◊ 10 9 m 3 of sediment fill has been sequestered within the flood plain and delta plain since ca. 7000 yr B.P., or about one-third of the annual discharge. The remaining load appears to be apportioned between the prograding subaqueous delta (1970 ◊ 10 9 m 3 ) and transport to the deep-sea Bengal fan via a nearshore canyon. Modern (<100 yr) budget estimates based on short-term accretion rates indicate a similar dispersal pattern and show that contemporaneous deposition continues within these disparate depocenters. The roughly equal partitioning of sediment among flood-plain, shelf, and deep-sea settings reflects the respective influence of an inland tectonic basin, a wide shelf, and a deeply incised canyon system. The findings also support new sequence stratigraphic models for these settings and indicate the important insight that modern river deltas can provide for ancient margin systems. Furthermore, results affirm that values of riverine sediment flux to the oceans may be considerably overestimated by not accounting for loss to the flood plains downstream of the gauging stations.

A new inaperturate pollen genus from the Late Miocene sediments of Site 218 of DSDP Leg 22 in the Central Bengal Fan, Indian Ocean

A new pollen genus Varisculptinaperturites has been described from the Late Miocene sediments of Site 218 of DSDP Leg 22 in the Central Bengal Fan, Indian Ocean. The genus accommodates inaperturate pollen ornamented with two types of sculpture, smaller spines and gemmae and larger clavae/baculae.

Evidence for long-term diffuse deformation of the lithosphere of the equatorial Indian Ocean

The presence of large earthquakes, east–west-striking folds and thrust faults in sediments, and east–west-striking undulations of wavelength 200 km in topography and gravity shows that the equatorial Indian Ocean is the locus of unusual deformation1,2,3,4,5,6,7,8. This deformation has been interpreted as a diffuse boundary between two tectonic plates9,10,11,12,13. Seismic stratigraphy and deep-sea drilling at two locations in the Bengal fan indicate that the deformation began 7.5–8.0 Myr ago3,14,15. Here, however, we show, using plate reconructions, that motion across this diffuse oceanic plate boundary began more than 10 Myr earlier than previously inferred and that the amount of north–south convergence across the boundary through the central Indian basin has been significantly greater than the convergence estimated from seismic profiles. The relative plate velocity accommodated across the central Indian basin has varied with time and has been as fast as ∼6 mm yr−1 — about half the separation rate of Earth's slowest-spreading mid-ocean ridge. The earliest interval of measurable motion, which began more than 18 Myr ago, may coincide with rapid denudation of the Tibetan plateau from ∼21 Myr to 15–17 Myr ( ref. 16). The present motion across the central Indian basin began no earlier than 11 Myr — following an earlier interval of slower motion from 18 to 11 Myr — and may have begun at ∼8 Myr, when the Tibetan plateau is thought to have attained its maximum elevation16,17.

The submarine delta of the Ganges–Brahmaputra: cyclone-dominated sedimentation patterns

The Ganges–Brahmaputra belongs to the world's largest rivers in terms of sediment discharge, but little is known yet about the processes controlling sediment distribution on the shelf and the documentation of these processes in large- and small-scale sedimentary structures. For this study the submarine delta of the river system was investigated by high-resolution seismic reflection profiles, and piston and gravity coring. Logged physical properties of the cores were combined with visual core description, grain-size analysis, and radiography. Sedimentation rates were estimated by 210Pb and 137Cs gamma spectrometry and cross-checked with evaluation of seismic profiles. The data reveal that the broad delta topset (<20 m waterdepth) is formed by a thick layer of sand + silt. Highest sedimentation rates (about 10 cm a −1) occur at the centre of clinoform delta foreset beds where about 20% of the total riverine sediment load is deposited and thus cause an annual progradation of the subaqueous delta front by about 15 m. The foreset beds consist of graded sand + silt layers which are thought to be deposited by sediment-laden flows generated during the surge of tropical cyclones (tempestites) interbedded with silty clay layers. The head of a canyon, the Swatch of No Ground, that deeply incises into the shelf, forms a sediment trap along the westward cyclonic transport path where sedimentation rate increases to about 50 cm a −1. Episodically earthquakes or storms destabilize especially the rapidly accumulated foreset beds in the eastern delta creating up to 8 m thick mass flows. On the outer shelf in water depth beyond 80 m no Holocene sediment is deposited.

Chemistry of surface sediment along a north–south transect across the equator in the Central Indian Basin: an assessment of biogenic and detrital influences on elemental burial on the seafloor

Nineteen surface sediment samples collected along a north–south transect across the equator in the Central Indian Basin (CIB) were analysed for their major, minor, and rare earth element (REE) composition. In absence of the element-flux data, we demonstrate here the association of various elements and their preferential sinks using inter-elemental associations in sediment exhibiting latitudinal dissimilarity. Biogenic sedimentation north of the equator (1° to 4°N latitudes) is dominated by carbonate ooze, whereas south of the equator (4° to 14°S latitudes) is dominated by siliceous ooze. Carbonate ooze deposition (>60% CaCO 3) above 4400 m and its drastic reduction (to <10%) below 4700-m water depths place the carbonate lysocline and the compensation depth (CCD) in a narrow depth zone of around 300-m thickness in the CIB. The siliceous ooze deposition is evident only below 5050-m water depth. This equatorial biogenic sediment domain is interrupted by a detrital clay zone (0° to 3°S latitudes) probably due to a combined effect of reduced dilution of distal Bengal fan sediment below CCD and focusing of suspended detritus drawn from the Bay of Bengal during vigorous southwest monsoon equatorial circulation. The latitudinal trends in the Element XS (detrital unsupported or non-terrigenous element) composition of the sediment and inter-elemental correlations yield four main carrier phases responsible for burial of several elements; (a) carbonate phase burying almost all Ca and Sr; (b) detrital phase burying the majority of Al, Fe, Mg and all of Ti and Sc; (c) oxide phase burying bulk of Mn,Co,Cu and Ba; and (d) biogenic phase burying the Al XS. Ba exhibits latitudinal similarity and strong positive association ( r=∼0.8 at 99.9% confidence level) with Mn, Co and Cu in both Element bulk and Element XS correlation matrices. The proportion of Ba XS, Mn XS, Co XS and Cu XS is >80% of their respective total content in all the three sediment zones. These observations suggest a close link between Ba and a Mn-oxide phase. On the other hand, nearly overlapping latitudinal trends of Al XS with carbonate in areas north of the equator and opal south of the equator; strong positive correlation of Al XS with opal ( r=+∼0.8), similar ratios between carbonate-free opal and Al XS (∼16) in both biogenic sediments suggest a close link between productivity and Al XS. The Al XS proportion in both biogenic sediment types is around 28% of the total Al suggesting nearly the same level of primary productivity along the transect (4°N to 14°S) which supports the estimates of Berger [Berger, W.H., 1989. Global maps of the ocean productivity. In: W.H. Berger, V.S. Smetacek, G. Wefer (Eds.), Productivity of the Ocean: Present and past. Dahlem workshop reports, Wiley, Chichester, pp. 429–455] and Kabanova [Kabanova, J.G., 1968. Primary production of the northern part of the Indian Ocean. Oceanology, 8: 214–255] (i.e., ∼70 g Cm −2 y −1). The REEs exhibit mutual coherence in the sediment ( r=>+0.7). However, they display, an increasing affinity towards Mn (` r' increases from +0.6 to +0.8) and decreasing affinity towards Ti (` r' decreases from +0.8 to +0.3) with increasing atomic number. Such a subtle but systematic variation within REE group suggests preferential burial of heavy-REE by Mn-oxide phase and light-REE by detrital phase.

Sources of Pb for Indian Ocean ferromanganese crusts: a record of Himalayan erosion?

A high resolution Pb isotope time-series for the last 26 Ma, dated by 10Be/ 9Be chronology, is reported for a north Indian Ocean ferromanganese crust. This record is compared with available Pb isotope time-series of six other crusts from the Atlantic, Indian and Pacific Oceans, each of which is based on 10Be/ 9Be chronology. The seven Pb isotope records reveal some remarkable features. In contrast to the Nd isotope time-series of these crusts which show a long-term (∼60 Ma) provinciality between the three main ocean basins, the Pb isotopes only show comparable provinciality over the last ∼5 Ma. Prior to about 15 Ma ago no distinct Indian Ocean Pb isotope signal existed. Within this established framework of Pb isotope distribution in the oceans the 208Pb/ 206Pb data for the north Indian Ocean crust reported here are anomalous. The 208Pb/ 206Pb ratio is particularly high and exceeds a value of 2.08 during the time interval from 20 to 8 Ma ago. Consideration of potential sources of Pb in the Indian Ocean which might provide such high 208Pb/ 206Pb ratios suggests that this crust most probably has recorded a time-varying erosional input of Pb from the Himalayas. The timing of the isotopic shift is in good agreement with maximum Himalayan exhumation rates deduced from crystallisation and cooling ages of synorogenic granites (20–14 Ma) and the sedimentation history of the Bengal Fan.

Unroofing history of the eastern Himalaya and the Indo-Burman ranges; heavy-mineral study of Cenozoic sediments from the Bengal Basin, Bangladesh

ABSTRACT Upper Eocene to Neogene fill of the Bengal basin provides an earlier unroofing history of the eastern Himalaya and Indo-Burman ranges than that provided by drilling of the Bengal fan, and analysis of heavy minerals in these sequences provides useful provenance constraints. Quartzose sandstones of the Eocene Kopili and Oligocene Barail Formations contain only 0.2% heavy minerals, comprising abundant opaque minerals and stable minerals such as tourmaline, garnet, rutile, and zircon. This assemblage suggests sediment sources dominated by low- to intermediate-grade metamorphic, silicic igneous, and metasedimentary rocks; the low heavy-mineral content and abundance of stable minerals suggest intense chemical weathering. These sediments may have been derived in part from incipient uplifts of the proto-Himalaya and/or the Indo-Burman ranges, but a more likely source is the Indian craton immediately to the west. Miocene sandstones of the Surma Group contain more abundant and more diverse heavy-mineral assemblages than Oligocene sandstones. These include abundant opaque minerals and garnet, mostly almandine, and moderate to minor amounts of tourmaline, kyanite, zircon, calcic amphibole, rutile, chlorite, zoisite, staurolite, epidote, sillimanite, and clinopyroxene, indicating a broad range of mostly metamorphic source rocks. Upper strata of the Surma Group also contain abundant blue-green amphibole, orthopyroxene, and sparse chromite, suggesting exhumation of arc and ophiolitic rocks. Sands and sandstones from the upper Miocene to Pliocene Tipam Group and the Pliocene to Pleistocene Dupi Tila Formation contain assemblages similar to those of the underlying Miocene sandstones, along with more abundant orthopyroxene and sillimanite in Dupi Tila sands. Temporal variations in heavy-mineral assemblages in the Bengal basin suggest that orogenic detritus first appears in the eastern Himalayan foredeep much later (early Miocene) than in the western part of the foredeep (Eocene). The appearance of high-pressure phases in upper Miocene strata reflects continued unroofing of deeper-crustal metamorphic and silicic to ultramafic plutonic rocks.

Cenozoic history of the Himalayan-Bengal system: Sand composition in the Bengal basin, Bangladesh

Stratigraphic sequences preserved in the Bengal basin provide detrital information that documents a significantly older history of the eastern Himalaya than that available from Ocean Drilling Program and Deep Sea Drilling Project cores recovered from the Bengal fan. The Bengal basin, formed as a result of the Himalayan collision, is located at the juncture of the Indian craton to the west, the Shillong massif and the Himalayan belt to the north, and the Indo-Burman ranges to the east. Modal analyses of Eocene and Oligocene sandstones of the Cherra, Kopili, and Barail Formations document compositions that are dominated by subangular monocrystalline quartz grains with scarce to no feldspar grains, and few lithic fragments (Cherra and Kopili: Qt 99 F 1 L 0 ; Barail: Qt 90 F 3 L 7 ; Qt—total quartz, F—feldspar, L—lithic fragments). These compositions are similar to sands derived from the Indian craton, suggesting that they underwent intense chemical weathering, likely in a source with low relief and considerable transport. Himalayan tectonism during this time was probably significantly more distant from the Bengal basin than at present. Sandstones of the Miocene Surma Group (Bhuban and Boka Bil Formations) are rich in feldspar grains, argillite, and very low-grade metamorphic lithic fragments (Bhuban: Qt 66 F 15 L 19 ; Boka Bil: Qt 57 F 23 L 20 ) relative to older sandstones, suggesting onset of uplift and erosional unroofing in the eastern Himalaya, and initiation of stream systems supplying orogenic detritus to the Bengal basin. Sands of the upper Miocene to Pliocene Tipam Group and the Pliocene–Pleistocene Dupi Tila Sandstone contain abundant argillitic and low- to medium-grade metamorphic lithic fragments and feldspar grains (Tipam: Qt 61 F 19 L 19 ; Dupi Tila: Qt 70 F 13 L 17 ), suggesting continued orogenic unroofing. These younger sands are rich in potassium feldspar (P/F=0.30, 0.20) relative to plagioclase (P)-rich Bhuban and Boka Bil sandstones (P/F=0.66 and 0.48), suggesting a granitic source, probably the Miocene leucogranites of the High Himalayan Crystalline terrane. These results document for the first time contrasts in orogenic history recorded in the Bengal system vs. western Himalayan foreland basins. Sands deposited in the Bengal basin are generally more quartzose and less lithic than those from the western Himalayan foreland basins, and pre-Miocene strata in the Bengal system show little to no evidence of orogenic activity. In part, this probably reflects west to east progression of the Himalayan collision, but it probably also results from sedimentary systems propagating southward, ahead of the advancing mountain belt as it has been consuming the remnant ocean basin trapped between the Indian craton and the Burmese block.

Secular variation of Nd and Pb isotopes in ferromanganese crusts from the Atlantic, Indian and Pacific Oceans

Two ferromanganese crusts from the Indian Ocean and one from the Atlantic Ocean have been analysed for 10Be/ 9Be, 143Nd/ 144Nd and 208,207,206Pb/ 204Pb ratios as a function of depth beneath their growth surfaces. 10Be/ 9Be ratios provide growth rate estimates for these crusts between 1.55 and 2.82 mm Ma −1 and further suggest that 87Sr/ 86Sr in crusts do not in any case examined so far provide reliable estimates for growth rates. A crust ALV-539 from 35°N in the western N. Atlantic has ϵ Nd and Pb-isotope variations indistinguishable from crust BM-1969.05 from 39°N in the N. Atlantic [K.W. Burton, H.-F. Ling, R.K. O'Nions, Closure of the central American isthmus and its impact on North Atlantic deepwater circulation, Nature (London) 386 (1997) 382–385] when 10Be/ 10Be ratios are used to estimate growth rates. Both crusts provide evidence for a marked change in deepwater composition in the western N. Atlantic with a reduction in ϵ Nd and an increase in 206Pb/ 204Pb from ∼8 Ma ago towards the present day. The two crusts from the Indian Ocean show comparatively small variations in ϵ Nd between −8.0 and −7.0 over the last 20 Ma and do not show the large shift in ϵ Nd seen in the Atlantic crusts. Comparison of ϵ Nd in the crusts analysed here with those published previously [H.-F. Ling, K.W. Burton, R.K. O'Nions, B.S. Kamber, F. von Blanckenburg, A.J. Gibb, J.R. Hein, Evolution of Nd and Pb isotopes in central Pacific seawater from ferromanganese crusts, Earth Planet. Sci. Lett. 146 (1997) 1–12; K.W. Burton, H-F. Ling, R.K. O'Nions, Closure of the central American isthmus and its impact on North Atlantic deepwater circulation, Nature (London) 386 (1997) 382–385] shows that provinciality in the present-day ϵ Nd structure of the Pacific, Atlantic and Indian Oceans has been maintained over ∼20 Ma or more despite the palaeogeographic changes that have occurred within this period. These include the closure of the Panama gateway and the uplift of the Himalayas. Superimposed on this broad inter-ocean structure are changes in ϵ Nd of the western N. Atlantic which may relate to the Panama gateway closure and shifts in the ϵ Nd of equatorial Pacific deepwater from 3–5 Ma ago. The absence of any such structure in ϵ Nd of the southwest and central Indian Ocean suggests that Himalayan erosion products such as preserved in the Bengal Fan sediments have not contributed significantly to Indian Ocean deepwater over the last 20 Ma. There is no straightforward relationship between records of 87Sr/ 86Sr in the global ocean and ϵ Nd in ocean deepwater as would be expected if inputs of radiogenic Sr and unradiogenic Nd were coupled.