(Table 2) Oxygen isotope composition of authigenic and detrital phosphates from ODL Leg 201 sites

Many (bio)geochemical processes that bring about changes in sediment chemistry normally begin at the sediment-water interface, continue at depth within the sediment column and may persist throughout the lifetime of sediments. Because of the differential reactivity of sedimentary phosphate phases in response to diagenesis, dissolution/precipitation and biological cycling, the oxygen isotope ratios of phosphate (d18OP) can carry a distinct signature of these processes, as well as inform on the origin of specific P phases. Here, we present results of sequential sediment extraction (SEDEX) analyses combined with d18OP measurements, aimed at characterizing authigenic and detrital phosphate phases in continental margin sediments from three sites (Sites 1227, 1228 and 1229) along the Peru Margin collected during ODP Leg 201. Our results show that the amount of P in different reservoirs varies significantly in the upper 50 m of the sediment column, but with a consistent pattern, for example, detrital P is highest in siliciclastic-rich layers. The d18OP values of authigenic phosphate vary between 20.2 per mil and 24.8 per mil and can be classified into at least two major groups: authigenic phosphate precipitated at/near the sediment-water interface in equilibrium with paleo-water oxygen isotope ratios (d18Ow) and temperature, and phosphate derived from hydrolysis of organic matter (Porg) with subsequent incomplete to complete re-equlibration and precipitated deeper in the sediments column. The d18OP values of detrital phosphate, which vary from 7.7-15.4 per mil, suggest two possible terrigenous sources and their mixtures in different proportions: phosphate from igneous/metamorphic rocks and phosphate precipitated in source regions in equilibrium with d18Ow of meteoric water. More importantly, original isotopic compositions of at least one phase of authigenic phosphates and all detrital phosphates are not altered by diagenesis and other biogeochemical changes within the sediment column. These findings help to understand the origin and provenance of P phases and paleoenvironmental conditions at/near the sediment-water interface, and to infer post-depositional activities within the sediment column.

Stable oxygen, hydrogene and chlorine isotopes of mid-Atlantic abyssal peridotites

Serpentinized abyssal peridotites sampled by the Ocean Drilling Program Leg 209 along the mid-Atlantic Ridge near the 15°20'N Fracture Zone have been analyzed for oxygen, hydrogen, and chlorine isotope compositions in order to determine isotopic behavior under a wide range of serpentinization conditions and place constraints on fluid history. Oxygen and hydrogen thermometry suggests peak serpentinization temperatures of 300-500°C. Serpentine separates have low deltaD values possibly due to a magmatic fluid component or low-temperature exchange during seafloor weathering. Chlorine geochemistry focused on three holes: 1274A and 1272A (serpentinized peridotites) and 1268A (serpentinite locally altered to talc). Concentrations of both, water-soluble chloride (WSC) and structurally bound chloride (SBC) are significantly lower at Hole 1268A compared to Holes 1274A and 1272A. The delta37Cl values for WSC and SBC of serpentinites in Holes 1274A and 1272A are slightly positive (avg. WSC = 0.20 per mil, n = 22 and avg. SBC = 0.35 per mil, n = 22), representing typical seawater-hydration conditions commonly determined for abyssal peridotite. The SBC of serpentinites from Hole 1268A are also positive (avg. = 0.63 per mil); whereas, the SBC in talc-dominated samples is negative (avg. = -1.22 per mil). The WSC of both talc- and serpentine-dominated samples are also negative (avg. = -0.15 per mil). We interpret the chlorine isotope data to preserve a record of multiple fluid events. As seawater hydrated the peridotite, 37Cl was preferentially incorporated into the forming serpentine and water-soluble salts, yielding similar delta37Cl values on a regional scale as sampled by Holes 1268A, 1274A and 1272A. The resultant pore fluid was left depleted in 37Cl. Locally (Hole 1268A), this evolved fluid was remobilized possibly due to the initiation of hydrothermal circulation in response to emplacement of a mafic magma body. The low delta37Cl pore fluids attained elevated SiO2 and sulfur concentrations due to interaction with the gabbroic intrusion and, when ascending through the surrounding serpentinite, caused formation of isotopically negative talc. This secondary fluid also flushed the preserved serpentinite of its previously formed salts, resulting in negative delta37Cl WSC values. The delta37Cl SBC values of the serpentinite samples remained unmodified by reaction with the secondary fluid.

Stable oxygen isotope ratios of Globigerinoides sacculifer from sediment core RC09-150 from the southeastern Indian Ocean

Oxygen isotopic and microfaunal analyses and shell size variations of Orbulina universa in two Indian Ocean cores indicate that the position of the Subtropical Convergence has fluctuated between a northern limit north of 31°S during glacial stages and its present, maximum southern limit. The northward displacement of the Subtropical Convergence to a position off Durban, South Africa, reflects the general weakness of the Agulhas Current during glacial stages and parts of interglacial stages, representing about 65 percent of the past 540,000 years.

Stable carbon and oxygen isotope analyses of Neogloboquadrina pachyderma

Detailed records of the carbon and oxygen isotopic ratios of Neogloboquadrina pachyderma are compared between nine high-latitude sediment cores, from the Northern and Southern Hemispheres, covering the last 140000 yrs. The strong analogies between the delta13C records permit to define a delta13C stratigraphic scale, with three clear cut transitions simultaneous with the oxygen isotopic transitions 6/5 (125 kyrs.), 5/4 (65 kyrs.), and 2/1 (13 kyrs.). The delta13C records of N. pachyderma in the high-latitude cores, which follow the changes in delta13C of the surface water TCO2 near areas of deep water formation present trends similar to the benthic foraminifera delta13C records in cores V19-30 and M12-392, although amplitudes of the isotopic shifts are different. This implies that a large part of the observed variations represents global changes in the carbon distribution between biosphere and ocean. The 13C/12C ratios of N. pachyderma in the North Atlantic cores display larger regional variations at 18 kyrs. B.P. than at present. To explain these differences, we have plotted the 18 kyrs. B.P. delta13C values of N. pachyderma from 17 cores distributed N of 40°N. Comparison with published surface water temperature distribution at 18 kyrs. B.P. indicates that a strong divergent cyclonic cell, centered approximatively 55°N and 15°W, was active during most of the last ice-age maximum. This hydrology, analogous to the present Weddell Sea, explains the published evidences of bottom water formation, if located on the northern flank of the gyre, and the strong polar front on the southern flank, probable location of intermediate water formation.

Stable isotopic composition and carbonate concentrations of sediments of the Cretaceous/Paleogene boundary in the Antarctic

Stable isotopic records across the Cretaceous/Paleogene (K/P) boundary in Maud Rise Holes 689B and 690C indicate that significant climatic changes occurred during the latest Cretaceous, beginning approximately 500 k.y. prior to the mass extinction event and the enrichment of iridium at the K/P boundary (66.4 Ma). An oxygen isotopic decrease of ~0.7 per mil - ~1.0 per mil is recorded in the Late Cretaceous planktonic and benthic foraminifers between 66.9 and 66.6 Ma. The negative isotope excursion was followed by a positive excursion of similar magnitude between 66.6 Ma (latest Cretaceous) and ~66.3 Ma (earliest Paleocene). No other isotopic excursions of this magnitude are recorded in the planktonic and benthic microfossil records 1.0 m.y prior to, and for 2.0 m.y following the mass extinction event at the K/P boundary. The magnitude and duration of these isotopic excursions were similar to those at the Paleocene/Eocene and Eocene/Oligocene boundaries. A major d13C excursion occurred 200 k.y. prior to the boundary, involving a positive shift in planktonic and benthic d13C of ~0.5 per mil - 0.75 per mil. Similar changes observed in other deep-sea sequences indicate that this reflected a global change in d13C of the oceanic total dissolved carbon (TDC) reservoir. The magnitude of this inferred carbon reservoir change and its association with high latitude surface-water temperature changes recorded in the d18O records implies that it was linked to global climate change through feedback loops in the carbon cycle. At the K/P boundary, the surface-to-deep water d13C gradient is reduced by approximately 0.6 per mil - ~0.2 per mil. However, unlike sequences elsewhere, the planktonic-benthic d13C gradient (Delta d13C) was not eliminated in the Antarctic. The surface-to-deep water gradient was re-established gradually during the 400 k.y. following the mass extinction. Full recovery of the Delta d13C occurred by ~60.0 Ma. In addition to the reduced vertical d13C gradient across the K/P boundary, there was a negative excursion in both planktonic and benthic d13C beginning approximately 100 k.y. after the boundary (66.3 Ma). This excursion resulted in benthic d13C values in the early Paleogene that were similar to those in the pre-K/P boundary intervals. This negative shift appears to reflect a change in the d13C of the oceanic TDC reservoir shift that may have resulted from reduced carbon burial and/or increased carbon flux to the oceans. Any model that attempts to explain the demise of the oceanic plankton at the end of the Cretaceous should consider the oceanic environmental changes that were occurring prior to the massive extinction event.

Stable carbon and oxygen isotope ratios of foraminifera from North Atlantic sediments

Oxygen-18 records of benthic foraminifera from the Atlantic Ocean are significantly different from those of the Pacific and Indian Oceans indicating that the Glacial North Atlantic Deep Water was about 1.3°C cooler than today because different deep water sources appeared in the North Atlantic Ocean during glacial times. The present study seeks to interprete carbon-13 records of planktonic and benthic foraminifera as a tracer of the cycle of the CO2 dissolved in surface and deep water of the ocean during the last climatic cycle. Carbon-13 records of planktonic foraminifera indicate that the delta13C of atmospheric CO2 and total CO2 dissolved in surface water did not vary noticeably (-0.2 +/- 0.3 per mil) during glacial times. Carbon-13 records of benthic foraminifera indicate that the eastern North Atlantic Ocean was an area of deep water formation dying isotopic stage 2, but not during most of stage 3. Moreover, large delta13C differences in the NADW between 20°N and 50°N show that the residence time of the glacial NADW was about 4 times that of today.

Oxygen and carbon isotope ratios for six planktonic foraminifer species from different water depth, Ontong-Java Plateau, Pacific (Table 2)

Stable isotope analyses and scanning electron micrographs have been carried out on six planktonic forminifera species, Pulleniatina obliquiloculata, Globorotalia tumida, Sphaeroidinella dehiscens, Globigerinoides ruber, Globigerinoides sacculifer and Globigerinoides quadrilobatus from eleven box-cores taken at increasing depths in the equatorial Ontong-Java Plateau (Pacific). This allows us to describe the way dissolution affects the microstructures of the tests of the different species and to quantify the changes of isotopic composition. We may conclude that: 1) dissolution effects on test morphology and stable isotope compositions are species dependent, species with a similar habitat showing a similar trend; 2) the shallow water, thin-shelled species are the first to disappear: scanning electron microscope (SEM) work shows alteration of outer layers. Deep water, thick-shelled species are present in all samples: SEM work shows breakdown and disparition of inner layers; 3) for all species there is a similar trend towards increasing delta18O values with increasing water depths and increasing dissolution. This effect may be as high as 0.6 ? per thousand meters for Globorotalia tumida; 4) below the lysocline, around 3500 m, it appears that 13C/12C ratios slightly increase towards equilibrium values for thick shelled species: G. tumida, P. obliquiloculata and S. dehiscens. 14C dates and isotope stratigraphy of two box-cores show that all samples are recent in age, and exclude upward mixing of glacial deposits as an important factor.

Age dating and isotopic analyses of sediment core RC11-120

Using the concept of 'orbital tuning', a continuous, high-resolution deep-sea chronostratigraphy has been developed spanning the last 300,000 yr. The chronology is developed using a stacked oxygen-isotope stratigraphy and four different orbital tuning approaches, each of which is based upon a different assumption concerning the response of the orbital signal recorded in the data. Each approach yields a separate chronology. The error measured by the standard deviation about the average of these four results (which represents the 'best' chronology) has an average magnitude of only 2500 yr. This small value indicates that the chronology produced is insensitive to the specific orbital tuning technique used. Excellent convergence between chronologies developed using each of five different paleoclimatological indicators (from a single core) is also obtained. The resultant chronology is also insensitive to the specific indicator used. The error associated with each tuning approach is estimated independently and propagated through to the average result. The resulting error estimate is independent of that associated with the degree of convergence and has an average magnitude of 3500 yr, in excellent agreement with the 2500-yr estimate. Transfer of the final chronology to the stacked record leads to an estimated error of +/-1500 yr. Thus the final chronology has an average error of +/-5000 yr.

(Table S1) Stable carbon and oxygen isotope ratios of benthic foraminifera of the western Atlantic Ocean

Oxygen and carbon isotopic data were produced on the benthic foraminiferal taxa Cibicidoides and Planulina from 25 new piston cores, gravity cores, and multicores from the Brazil margin. The cores span water depths from about 400 to 3000 m and intersect the major water masses in this region. These new data fill a critical gap in the South Atlantic Ocean and provide the motivation for updating the classic glacial western Atlantic d13C transect of Duplessy et al. (1988). The distribution of 13C of SumCO2 requires the presence of three distinct water masses in the glacial Atlantic Ocean: a shallow (~1000 m), southern source water mass with an end-member d13C value of about 0.3-0.5 per mil VPDB, a middepth (~1500 m), northern source water mass with an end-member value of about 1.5 per mil, and a deep (>2000 m), southern source water with an end-member value of less than -0.2 per mil, and perhaps as low as the -0.9 per mil values observed in the South Atlantic sector of the Southern Ocean (Ninnemann and Charles, 2002, doi:10.1016/S0012-821X(02)00708-2). The origins of the water masses are supported by the meridional gradients in benthic foraminiferal d18O. A revised glacial section of deep water d13C documents the positions and gradients among these end-member intermediate and deep water masses. The large property gradients in the presence of strong vertical mixing can only be maintained by a vigorous overturning circulation.

Radiogenic neodymium, strontium and lead isotopes of authigenic and detrital sedimentary phases and stable oxygen and carbon isotopes of benthic foraminifera from gravity cores SO213-59-2 and SO213-60-1 (central South Pacific).

The South Pacific is a sensitive location for the variability of the global oceanic thermohaline circulation given that deep waters from the Atlantic Ocean, the Southern Ocean, and the Pacific Basin are exchanged. Here we reconstruct the deep water circulation of the central South Pacific for the last two glacial cycles (from 240,000 years ago to the Holocene) based on radiogenic neodymium (Nd) and lead (Pb) isotope records complemented by benthic stable carbon data obtained from two sediment cores located on the flanks of the East Pacific Rise. The records show small but consistent glacial/interglacial changes in all three isotopic systems with interglacial average values of -5.8 and 18.757 for epsilon Nd and 206Pb/204Pb, respectively, whereas glacial averages are -5.3 and 18.744. Comparison of this variability of Circumpolar Deep Water (CDW) to previously published records along the pathway of the global thermohaline circulation is consistent with reduced admixture of North Atlantic Deep Water to CDW during cold stages. The absolute values and amplitudes of the benthic delta13C variations are essentially indistinguishable from other records of the Southern Hemisphere and confirm that the low central South Pacific sedimentation rates did not result in a significant reduction of the amplitude of any of the measured proxies. In addition, the combined detrital Nd and strontium (87Sr/86Sr) isotope signatures imply that Australian and New Zealand dust has remained the principal contributor of lithogenic material to the central South Pacific.

Stable carbon and oxygen isotope composition of Recent, Tertiary and Cretaceous foraminifera (Table III)

Oxygen isotope analyses of Tertiary and Cretaceous planktic foraminifera indicate that species have been stratified with respect to depth in the water column at least since Albian time. There is a relationship between morphology and depth habitat. Species with globigerine morphology have consistently occupied shallower depths than have species with globorotalid morphology. Biserially arranged species occupied both shallow and deep levels in the water column. On the average, it appears that ancient species with shallow habitats have been more susceptible to dissolution and have been preserved less well than species dwelling in deeper habitats. This relationship is similar to that observed for Recent planktic foraminifera. Comparison of carbon isotope ratios of adult and juvenile forms indicates that either the source of the carbon found in the shell or the carbon isotopic fractionations which occur during calcite secretion change during the development of individual foraminifera. The carbon isotopic ratios do not provide a reliable means for reconstructing the depth habitats of ancient species. Temperature-depth profiles for tropical Tertiary oceans have been reconstructed from the isotopic temperatures of planktic and benthic foraminifera. The vertical thermal structure of Oligocene oceans resembled that of modern oceans most closely. Those of Paleocene and Maastrichtian times differed most from that of modern oceans.

Stable carbon and oxygen isotope ratios of benthic foraminifera

Carbon isotopic measurements on the benthic foraminiferal genus Cibicidoides document that mean deep ocean delta13C values were 0.46 per mil lower during the last glacial maximum than during the Late Holocene. The geographic distribution of delta13C was altered by changes in the production rate of nutrient-depleted deep water in the North Atlantic. During the Late Holocene, North Atlantic Deep Water, with high delta13C values and low nutrient values, can be found throughout the Atlantic Ocean, and its effects can be traced into the southern ocean where it mixes with recirculated Pacific deep water. During the glaciation, decreased production of North Atlantic Deep Water allowed southern ocean deep water to penetrate farther into the North Atlantic and across low-latitude fracture zones into the eastern Atlantic. Mean southern ocean delta13C values during the glaciation are lower than both North Atlantic and Pacific delta13C values, suggesting that production of nutrient-depleted water occurred in both oceans during the glaciation. Enriched 13C values in shallow cores within the Atlantic Ocean indicate the existence of a nutrient-depleted water mass above 2000 m in this ocean.

Biogenic opal, carbonat concentrations and stable oxygen isotope ratios of foraminifera from sediment cores of the Southern Ocean

We present records of biogenic opal percentage and burial rate in 12 piston cores from the Atlantic and Indian sectors of the Southern Ocean. These records provide a detailed, quantitative description of changing patterns of opal deposition over the last 450 kyr. The striking regional coherence of these records suggests that dissolution in the deep sea and sediment pore waters does not obscure the surface productivity signal, and therefore these opal time series can be used in combination with other surface water tracers to make inferences about the chemistry and circulation of the Southern Ocean under different global climate conditions. Three broad depositional patterns can be distinguished. Northernmost records (39°-42°S latitude) are characterized by enhanced opal burial during glacial periods and strong 41 kyr periodicity. Records from cores just north of the present Antarctic Polar Front (46°-49°S) show even larger increases in opal burial rate during glacial intervals, but have variance concentrated in the 100 and 23 kyr bands. Southernmost records (51°-55°S) are completely out of phase with those to the north, with greatly reduced opal burial rates during glacial periods. Taken as a whole, the opal records show no evidence for the increased total Antarctic productivity predicted by recent geochemical models of atmospheric CO2 variability. The areal expansion of Southern Ocean sea ice over the present zone of high siliceous productivity provides one plausible explanation for the glacial-interglacial opal patterns. The excess silica not taken up in this zone during glacial periods would contribute to greater nutrient availability and thus higher productivity in the subantarctic region. However, local circulation changes may act to modify this basic signal, possibly accounting for the observed differences in the opal variance spectra.

(Table 1) Planktonic foraminifera counts and stable oxygen isotope ratios of Pacific Ocean surface sediments

Core-top samples from the eastern tropical Pacific (10°N to 20°S) were used to test whether the ratio between Globorotalia menardii cultrata and Neogloboquadrina dutertrei abundance (Rc/d) and the oxygen isotope composition (?18O) of planktonic foraminifera can be used as proxies for the latitudinal position of the Equatorial Front. Specifically, this study compares the ?18O values of eight species of planktonic foraminifera (Globigerinoides ruber sensu stricto (ss) and sensu lato (sl), Globigerinoides sacculifer, Globigerinoides triloba, Pulleniatina obliquiloculata, Neogloboquadrina dutertrei, Globorotalia menardii menardii, Globorotalia menardii cultrata and Globorotalia tumida) with the seasonal hydrography of the region, and evaluates the application of each species or combination of species for paleoceanographic reconstructions. The results are consistent with sea surface temperature and water column stratification patterns. We found that in samples north of 1°N, the Rc/d values tend to be higher and d18O values of G. ruber, G. sacculifer, G. triloba, P. obliquiloculata, N. dutertrei, and G. menardii cultrata tend to be lower than those from samples located south of 1°N. We suggest that the combined use of Rc/d and the d18O difference between G. ruber and G. tumida or between P. obliquiloculata and G. tumida are the most suitable tools for reconstructing changes in the latitudinal position of the Equatorial Front and changes in the thermal stratification of the upper water column in the eastern tropical Pacific.