Late Neogene carbonate sedimentation and stable isotope record of ODP Site 121-758

The evolution of oceanic and climatic conditions the northeast Indian Ocean during the last 7 m.y. is revealed in the sediments from Site 758. We present detailed and continuous records of d18O and d13C from planktonic foraminifers, weight percent calcium carbonate, weight percent coarse fraction, magnetic susceptibility, and geomagnetic reversals. Sample spacing of the records ranges from 3 to 10 cm and is equivalent to an average time interval of 2000 to 6000 yr. Despite the fact that core recovery ranged between 100% and 105%, recovery gaps as large as 2.7 m occurred at nearly every break between advanced hydraulic piston cores. Approximately 12% of the late Neogene sequence was not recovered in each of the two holes drilled at Site 758. To circumvent the discontinuity introduced by the gaps, a composite depth section was constructed from multiple cores taken from offset holes at Site 758. The resulting composite depth section extends continuously from 0 to 116 mbsf, from the Holocene to the upper Miocene. A detailed chronostratigraphy is based on geomagnetic reversals which extend from the Brunhes Chron to Chron 6, and on d18O stages 1 through 105, which span from 0 to 2.5 Ma. The d18O record is dominated by a ~40-k.y. cycle in the late Pliocene and early Pleistocene, and is followed by a change to a ~100-k.y. cycle in the late Pleistocene. The mid-Pleistocene transition between these two modes of variability occurs between d18O stages 25 and 22 (between 860 and 800 Ka). Thirteen major volcanic ash horizons from the Indonesian arc are observed throughout the sedimentary section and are dated by their relative position within the geomagnetic reversals and the d18O chronostratigraphy. Since 5 Ma, there has been a long-term decline in weight percent CaCO3 and CaCO3 mass accumulation rates, and an associated rise in non-CaCO3 mass accumulation rates. We attribute these changes to a decrease in CaCO3 productivity and an increase in terrigenous sedimentation through enhanced riverine input. Such input may be linked to rapid tectonic uplift of the Himalayas and the Tibetan Plateau via mechanisms such as the intensification of the monsoonal rains, increased fluvial erosion, and regional glaciation. The long-term increase in percent coarse fraction since 5 Ma suggests a gradual increase in CaCO3 preservation. Higher frequency fluctuations in CaCO3 preservation are superimposed on the long-term trend and are related to climate fluctuations. The abrupt drop (-50%) in CaCO3 accumulation at 3.4 Ma signals a dramatic decrease in CaCO3 production that occurred over much of the Indian Ocean.

Geochemistry of ODP Sites 121-758 and 121-757

Nd isotopes are useful tracers for paleoceanography due to the short Nd residence time in seawater and the large differences between the isotopic signatures of various geological reservoirs. Therefore, ?Nd variations reflect the geological history of individual oceanic basins. Using a differential dissolution technique, which extracts Nd isotopes of seawater trapped in MnO2 coatings and carbonates in marine sediment, we measured almost two hundred samples from ODP Sites 758 and 757 in the Northern Bay of Bengal covering the last 4 Ma. For the first time, we have shown a covariation between epsilon-Nd and d18O over at least the last 800 ka. We also show that from 4 Ma to 2.6 Ma, epsilon-Nd is almost constant and starts to fluctuate at 2.6 Ma when northern glaciations increased. From 2.6 Ma to 1 Ma the fluctuation period is close to 40 ka while from 1 Ma to present it is dominantly 100 ka. We attribute these findings to mixing between Himalayan river water (that ultimately originates as Indian summer monsoon rain) and normal Bay of Bengal seawater. Previous studies on seawater, using epsilon-Nd, d18O analyzed on planktonic foraminifera and sedimentary data, can be integrated into this model. A simple quantitative binary mixing model suggests that the summer monsoon rain was more intense during interglacial than glacial periods. During last glacial episode, the monsoon trajectory was deviated to the east. At a large scale, the Indian monsoon is fully controlled by the variations in Northern Hemisphere climate but with a complex response function to this forcing. Our study clearly establishes the large potential of Nd isotope data to evaluate the hydrological river regime during the Quaternary and its relationship with climate fluctuations, particularly when the sediment archive is sampled close to sediment sources.

Stable isotope record and age determination of ODP Site 121-758 in the northeastern Indian Ocean

Numerous studies have shown that delta18O records from benthic and planktonic foraminifera, primarily a proxy of global ice volume variations, reflect Milankovitch periodicities. To study climatic response to orbital forcing at Ocean Drilling Program site 758, we have generated continuous delta18O and delta13C records from a single benthic foraminiferal species Cibicides wuellerstorfi for the last 3.6 m.y. and extended the planktonic foraminiferal isotope records of Farrell and Janecek (1991, doi:10.2973/odp.proc.sr.121.124.1991) (0-2.5 Ma, based on Globigerinoides sacculifer) to 3.6 Ma (Chen, 1994). We then constructed an age model by matching, correlating and tuning the benthic delta18O record to a model simulation of ice volume (Imbrie and Imbrie, 1980, doi:10.1126/science.207.4434.943). The filtered 41- and 23-kyr signals based on the resultant astronomically tuned age model are highly correlated to obliquity (r=0.83) and precession (r=0.75), respectively. Although derived with methodology different from Shackleton et al. (1990) and Hilgen (1991, doi:10.1016/0012-821X(91)90206-W, 1991, doi:10.1016/0012-821X(91)90082-S), our results generally agree with their published astronomical timescales for the time interval from 0 to 3.0 Ma, providing additional support for the newly emerging chronology based on orbital tuning. Slight discrepancies exist in the time interval from 3.0 to 3.6 Ma, suggesting several possibilities, including differences in the approaches of orbital tuning and the relatively low amplitude of delta18O variations in our record. However, even if the discrepancies are due to the relatively low amplitude of the isotope signals in our record at 3.0-3.6 Ma, our resultant timescale as a whole does not adversely affect our evaluation of the paleoclimatology and paleoceanography of the Indian Ocean, such as the evolution of the 100-, 41- and 23-kyr cycles, and variation of global ice volume and deepwater temperature during the past 3.6 m.y.

Occurrence and chemistry of microtektites of ODP Site 121-758

The Toba lake event, the Australasian microtektite event, and the Cretaceous/Paleogene boundary were analyzed on the basis of foraminifers, carbonate content, trace elements, and spherules (microtektites). The Toba ash event, recovered in Hole 758C, may have had minor influences on the foraminiferal populations. The Australasian tektite event has probably some influence on foraminiferal ecology, because the larger specimens become scarce just above the microtektite layer. Microtektites recovered from Hole 758B closely resemble spherules recovered from several Cretaceous/Paleogene boundary localities in North America. The Cretaceous/Paleogene spherules, however, are usually larger and are completely altered to goyazite in the terrestrial environment and to smectite in a marine environment. The Cretaceous/Paleogene boundary of Hole 752B does not show obvious anomalous trace-element concentrations, and iridium concentrations are below our detection limits. The trace-element pattern is dominated by the alternation of chalk with volcanic ash layers above the Cretaceous/Paleogene boundary.

Geochemistry of tephra layers in ODP Site 121-758

A tephrochronology of the past 5 Ma is constructed with ash layers recovered from Neogene sediments during drilling at ODP Leg 121 Site 758 on northern Ninetyeast Ridge. The several hundred tephra layers observed in the first 80 m of cores range in thickness from a few millimeters to 34 cm. Seventeen tephra layers, at least 1 cm thick, were sampled and analyzed for major elements. Relative ages for the ash layers are estimated from the paleomagnetic and d18O chronostratigraphy. The ash layers comprise about 1.7% by volume of the sediments recovered in the first 72 m. The median grain size of the ashes is about 75 ?m, with a maximum of 150 ?m. The ash consists of rhyolitic bubble junction and pumice glass shards. Blocky and platy shards are in even proportion (10%-30%) and are dominated by bubble wall shards (70%-90%). The crystal content of the layers is always less than 2%, with Plagioclase and alkali feldspar present in nearly every layer. Biotite was observed only in the thickest layers. The major element compositions of glass and feldspar reflect fractionation trends. Three groupings of ash layers suggest different provenances with distinct magmatic systems. Dating by d18O and paleomagnetic reversals suggests major marine ash-layer-producing eruptions (marine tephra layers > 1 cm in thickness) occur roughly every approximately 414,000 yr. This value correlates well with landbased studies and dates of Pleistocene Sumatran tuffs (average 375,000-yr eruptive interval). Residence times of the magmatic systems defined by geochemical trends are 1.583, 2.524, and 1.399 Ma. The longest time interval starts with the least differentiated magma. The Sunda Arc, specifically Sumatra, is inferred to be the source region for the ashes. Four of the youngest five ash layers recovered correlate in time and in major element chemistry to ashes observed on land at the Toba caldera.