Description: Temporal variations in Fe isotope compositions at three locations in the Pacific Ocean over the last 10 Ma are inferred from high-resolution analyses of three hydrogenetic ferromanganese crusts. Iron pathways to the central deep Pacific Ocean appear to have remained constant over the past 10 Ma, reflected by a remarkably constant Fe isotope composition, despite large changes in the Fe delivery rates to the surface ocean via dust. These results suggest that the Fe cycle in the deep ocean is decoupled from that in surface waters. By contrast, one ferromanganese crust from the Izu-Bonin (IB) back-arc/marginal basin of the W. Pacific exhibits large δ56Fe variations. In that crust, decreases in δ56Fe values correlate with increases in Mn, Mg, Ni, Cu, Zn, Mo, and V contents, and consistent with periods of intense hydrothermal input and increased growth rates. A second crust located within 100 km of the first IB sample does not record any of these periods of enhanced hydrothermal input. This probably reflects the isolated pathways by which hydrothermally sourced Fe may have migrated in the back arc, highlighting the high degree of provinciality that Fe isotopes may have in the modern (oxic) oceans. Our results demonstrate that despite efficient removal at the source, hydrothermal Fe injected into the deep ocean could account for a significant fraction of the dissolved Fe pool in the deep ocean, and that hydrothermally sourced Fe fluxes to the open ocean may have lower δ56Fe values than those measured so far in situ at hydrothermal vents. Correlation between δ56Fe values and elements enriched in hydrothermal fluids may provide a means for distinguishing hydrothermal Fe from other low-δ56Fe sources to the oceans such as dissolved riverine Fe or porewaters in continental shelf sediments.

Description: The Atlantis Massif (Mid-Atlantic Ridge, 30°N) is an oceanic core complex marked by distinct variations in crustal architecture, deformation and metamorphism over distances of at least 5 km. We report Sr and Nd isotope data and Rare Earth Element (REE) concentrations of gabbroic and ultramafic rocks drilled at the central dome (IODP Hole 1309D) and recovered by submersible from the southern ridge of the massif that underlie the peridotite-hosted Lost City Hydrothermal Field. Systematic variations between the two areas document variations in seawater penetration and degree of fluid–rock interaction during uplift and emplacement of the massif and hydrothermal activity associated with the formation of Lost City. Homogeneous Sr and Nd isotope compositions of the gabbroic rocks from the two areas (87Sr/86Sr: 0.70261–0.70429 and εNd: + 9.1 to + 12.1) indicate an origin from a depleted mantle. At the central dome, serpentinized peridotites are rare and show elevated seawater-like Sr isotope compositions related to serpentinization at shallow crustal levels, whereas unaltered mantle isotopic compositions preserved in the gabbroic rocks attest to limited seawater interaction at depth. This portion of the massif remained relatively unaffected by Lost City hydrothermal activity. In contrast, pervasive alteration and seawater-like Sr and Nd isotope compositions of serpentinites at the southern wall (87Sr/86Sr: 0.70885–0.70918; εNd: − 4.7 to + 11.3) indicate very high fluid–rock ratios (~ 20 and up to 106) and enhanced fluid fluxes during hydrothermal circulation. Our studies show that Nd isotopes are most sensitive to high fluid fluxes and are thus an important geochemical tracer for quantification of water–rock ratios in hydrothermal systems. Our results suggest that high fluxes and long-lived serpentinization processes may be critical to the formation of Lost City-type systems and that normal faulting and mass wasting in the south facilitate seawater penetration necessary to sustain hydrothermal activity.

Description: During the Last Glacial Maximum much of North America was covered by the Laurentide ice sheet. Its melting during termination 1 led to systematic changes in proglacial lake formation, continental runoff, and possibly North Atlantic Meridional Overturning Circulation. The accompanying change in chemical weathering rates in the interior of North America throughout the deglaciation resulted in a pronounced change in seawater Pb isotope composition in the western North Atlantic Ocean. Here we present the first high-resolution records of seawater Pb isotope variations of North Atlantic Deep Water extracted from authigenic Fe–Mn oxyhydroxides in three sediment cores (51GGC, 1790 m depth; 31GGC, 3410 m depth; 12JPC, 4250 m depth) from the Blake Ridge off Florida. These data reveal a striking excursion from relatively unradiogenic 206Pb/204Pb as low as 18.93 towards highly radiogenic Pb isotope compositions that was initiated during the Bølling–Allerød interstadial and was most pronounced in both intermediate and deep waters during and after the Younger Dryas (206Pb/204Pb as high as 19.38 at 8.8 ka in 4250 m). This pattern is interpreted to be a direct function of increased inflow of continent-derived radiogenic Pb into the western North Atlantic, supplied through chemical weathering of North American rocks that had been eroded and freshly exposed during the preceding glacial cycle. These sediment-derived data are complemented by new laser ablation Pb isotope data from a ferromanganese crust from the Blake Plateau at 850 m water depth, which show only small glacial–interglacial Pb isotope variations of the Florida Current (206Pb/204Pb between 19.07 and 19.16). The lack of change in the Blake Plateau record at the same time as the radiogenic excursion in the deeper sediments supports a northern origin of the pulse of radiogenic Pb. After the Younger Dryas, the deep western North Atlantic has experienced a persistent highly radiogenic Pb supply that was most pronounced during the first half of the Holocene and still lasts until today.

Description: Neogene marine sediments can be dated via decay of the cosmogenic radionuclide Be-10. Two cores from the Alpha and Mendeleev Ridges in the Arctic Ocean have been analyzed for seawater-derived beryllium (Be) isotopes in order to date the sediments and to calculate sedimentation rates. The decrease of Be-10 concentration in the cores was used to calculate first order sedimentation rates. To eliminate the dilution effect of beryllium caused by short-term changes in sedimentation rate and grain size, the Be-10 concentrations were normalized to the terrigenous stable isotope Be-9 determined in the same sample aliquot. The measured Be-10 concentrations yield low average sedimentation rates for the Alpha and Mendeleev Ridges of 2.3 mm ka(-1) and 2.7 mm ka(-1), respectively. Sedimentation rates calculated from the Be-10/Be-9 ratios result in similarly low values, ranging from 0.2 to 6.8 mm ka(-1) for the Alpha Ridge core and from 1.9 to 6.9 mm ka(-1) for the Mendeleev Ridge core. However, amino acid racemization dating for the past 150 ka of a core adjacent to the Mendeleev Ridge core studied here indicates significantly higher sedimentation rates than calculated from the downcore decrease of Be-10 and Be-10/Be-9. If such higher rates also prevailed at the locations of our cores, for which there is biostratigraphic evidence, either the supply of Be-10 was much lower than assumed or that of Be-9 was much higher. This could imply that the signature of the deep waters in this part of the Arctic Ocean compared to today was largely different for most of the past approximately one million years with a significantly lower Be-10/Be-9 ratio. Our study also addresses the variability of beryllium isotopes in sediment cores across the Arctic Ocean through a comparison of previously published results. Calculated Be-10 fluxes reveal low values in the Amerasian Basin and highest values in the Eurasian Basin, near the Fram. Strait. The decrease of Be isotopes in the two studied Amerasian Basin cores may thus have been caused by environmental factors such as significantly reduced inflow of Atlantic waters in the past, reduced input of Be-10 and/or increased input of Be-9 from the shelves, combined with a more efficient sea ice shielding in this part of the Arctic Ocean. (C) 2009 Elsevier B.V. All rights reserved.

Description: Ferromanganese crusts provide records of long term change in ocean circulation and continental weathering. However, calibrating their age prior to 10 Ma has been entirely based on empirical growth rate models using Co concentrations, which have inherently large uncertainties and fail to detect hiatuses and erosional events. We present a new method for dating these crusts by measuring their osmium (Os) isotope record and matching it to the well-known marine Os isotope evolution of the past 80 Ma. The well-characterised crust CD29-2 from the central Pacific, was believed to define a record of paleooceanographic change from 50 Ma. Previous growth rate estimates based on the Co method are consistent with the new Os isotope stratigraphy but the dating was grossly inaccurate due to long hiatuses that are now detectable. The new chronology shows that it in fact started growing prior to 70 Ma in the late Cretaceous and stopped growing or was eroded between 13.5 and 47 Ma. With this new technique it is now possible to exploit the full potential of the oceanographic and climatic records stored in Fe–Mn crusts.

Description: Global Nd–Hf isotope systematics can be mainly described with two linear arrays, the global silicate Earth array (“the terrestrial array”) and the global ferromanganese crust and nodule array (”the seawater array”). The offset between these two arrays provides evidence for the sources and mechanisms by which these elements are added to ocean water. However, the reason for this offset is under debate, with the two preferred hypotheses being (i) incongruent release of Hf during continental weathering and (ii) hydrothermal contribution of Hf to the seawater budget. Here we present new Nd and Hf isotope data on glacio-marine core-top sediments from around the perimeter of the Antarctic continent. The results range from εHf = − 30.0 to εHf = + 3.9 and εNd = − 21.3 to εNd = + 0.9, reflecting the large range of basement ages and lithologies around the Antarctic continent. In Nd–Hf isotope space, they confirm the systematic correlations found in rocks from other parts around the world and provide valuable insights into the previously underrepresented group of sediments with very old provenance. In this paper we revisit the cause for the offset of the seawater array from the terrestrial array using simple mass balance considerations. We use these calculations to test to what degree the seawater array could be a product of preferential weathering of “non-zircon portions” of the upper continental crust, implying retention of zircons in the solid residue of weathering. Lutetium–Hf and Sm–Nd evolution and mixing calculations show that the global seawater array can be generated with continental sources only. On the other hand, a predominantly hydrothermal origin of Hf in the ocean is not possible because the seawater Hf isotopic composition is significantly less radiogenic than hydrothermal sources, and requires a minimum fraction of 50% continental Hf. While hydrothermal sources may contribute some Hf to seawater, continental contributions are required to balance the budget.

Description: A high-resolution authigenic Nd isotope record has been extracted from the Fe–Mn oxyhydroxide fraction of drift sediments along the Blake Ridge in the North Atlantic. These sediments facilitate reconstruction of the timing and extent of major hydrographic changes in the western North Atlantic since the Last Glacial Maximum (LGM). This is one of the few locations where sediments were deposited in the major flow path of the Western Boundary Undercurrent (WBUC), which transports North Atlantic Deep Water (NADW) southward at the present day. The hydrodynamic setting, however, also causes problems. Authigenic Nd isotope compositions similar to the typical present-day NADW εNd value of − 13.5 ± 0.5 were only extracted from sediments located within the main water body of the WBUC coinciding with the highest along slope current velocity below 3200 m water depth. Above this depth the authigenic Nd-isotopic composition is more radiogenic than measured in a nearby seawater profile and appears to be influenced by downslope and lateral sediment redistribution. Our data suggest that these radiogenic signals were formed at shallow depths in Florida current waters, compromising the recorded ambient deep water Nd isotope signal in the Blake Ridge Fe–Mn oxyhydroxide coatings from intermediate depths during the Holocene and the deglaciation. The unradiogenic Nd-isotopic composition typical of present-day NADW is not detectable along the Blake Ridge for any water depth during the LGM. Unlike the deglacial and Holocene sections, the intermediate core from 1790 m water depth did not experience significant sediment focusing during the LGM, in accord with the higher current velocities at this depth, suggesting that at this site an ambient LGM bottom water Nd isotope signal was recorded. Assuming this to be correct, our results indicate that the εNd of the shallower glacial equivalent of NADW, the Glacial North Atlantic Intermediate Water (GNAIW) may have been as radiogenic as − 9.7 ± 0.4. Since the authigenic Nd isotope compositions of the Holocene and the deglacial sections of the intermediate depth sediment core were biased towards a shallow water signal, this first determination of a GNAIW εNd for the LGM will have to be corroborated by results from other locations and archives. The LGM and deglacial sediments below 3400 m water depth bear no evidence of an ambient deep water εNd as unradiogenic as − 13.5. Although the deep core sites also experienced enhanced degrees of sediment focusing before the Younger Dryas, the εNd values of between − 11 and − 10 are more readily explained in terms of increased presence of Southern Source Waters. If this is the case, the change to Nd-isotopic compositions that reflect a modern circulation pattern, including the presence of Lower NADW, only occurred after the Younger Dryas.

Description: Nd concentration and isotope data have been obtained for the Canada, Amundsen, and Makarov Basins of the Arctic Ocean. A pattern of high Nd concentrations (up to 58 pM) at shallow depths is seen throughout the Arctic, and is distinct from that generally seen in other oceans where surface waters are relatively depleted. A range of isotopic variations across the Arctic and within individual depth profiles reflects the different sources of waters. The dominant source of water, and so Nd, is the Atlantic Ocean, with lesser contributions from the Pacific and Arctic Rivers. Radiogenic isotope Nd signatures (up to epsilon(Nd) = -6.5) can be traced in Pacific water flowing into the Canada Basin. Waters from rivers draining older terrains provide very unradiogenic Nd (down to epsilon(Nd) = -14.2) that can be traced in surface waters across much of the Eurasian Basin. A distinct feature of the Arctic is the general influence of the shelves on the Nd concentrations of waters flowing into the basins, either from the Pacific across the Chukchi Sea, or from across the extensive Siberian shelves. Water-shelf interaction results in an increase in Nd concentration without significant changes in salinity in essentially all waters in the Arctic, through processes that are not yet well understood. In estuarine regions other processes modify the Nd signal of freshwater components supplied into the Arctic Basin, and possibly also contribute to sedimentary Nd that may be subsequently involved in sediment-water interactions. Mixing relationships indicate that in estuaries, Nd is removed from major river waters to different degrees. Deep waters in the Arctic are higher in Nd than the inflowing Atlantic waters, apparently through enrichments of waters on the shelves that are involved in ventilating the deep basins. These enrichments generally have not resulted in major shifts in the isotopic compositions of the deep waters in the Makarov Basin (epsilon(Nd) similar to -10.5), but have created distinctive Nd isotope signatures that were found near the margin of the Canada Basin (with epsilon(Nd) similar to -9.0). The deep waters of the Amundsen Basin are also distinct from the Atlantic waters (with epsilon(Nd) = -12.3), indicating that there has been limited inflow from the adjacent Makarov Basin through the Lomonosov Ridge. (C) 2009 Elsevier Ltd. All rights reserved.

Description: Techniques for the purification of Si for the determination of its natural stable isotopic composition have in the past been based on the requirements for gas-source mass-spectrometry, rather than MC-ICPMS. For high precision analyses by MC-ICPMS it is essential to have very pure solutions and in this paper a new technique is presented for the separation and purification of Si from natural samples to improve the determination of isotope ratios. A method has been optimised based on alkaline fusion followed by ion-exchange chromatography. The application to natural samples, such as river water samples and silicate mineral/rock samples is demonstrated. Alkali fusion avoids the use of hydrofluoric acid (HF), which introduces difficulties for the determination of Si isotope ratios using MC-ICPMS. By eliminating HF a 30–40% increase in sensitivity is achieved as well as a marked enhancement of mass bias stability leading to a factor of 2 improvement in reproducibility. The cation-exchange method enables processing of very small samples (3.6 μg Si) and a rapid and effective separation of Si from other cationic species. The overall recovery of Si during the entire procedure is better than 98% and no Si isotope fractionation is generated. Matrix tests demonstrate that this method is suitable for silicates, and that typical sulphate and nitrate abundances of river waters have no effect on measured Si isotope composition. The latter aspect is vital for analysis of river waters since the technique does not separate dissolved Si (silicic acid) from ambient anionic species. Overall, the new method presents a faster, safer and more reliable way to measure Si isotopes via MC-ICPMS.