François Thil

Iron sources and dissolved‐particulate interactions in the seawater of the Western Equatorial Pacific, iron isotope perspectives

This work presents iron isotope data in the western equatorial Pacific. Marine aerosols and top core margin sediments display a slightly heavy Fe isotopic composition (δ 56 Fe) of 0.33 ± 0.11‰ (2SD) and 0.14 ± 0.07‰, respectively. Samples reflecting the influence of Papua New Guinea runoff (Sepik River and Rabaul volcano water) are characterized by crustal values. In seawater, Fe is mainly supplied in the particulate form and is found with a δ 56 Fe between A0.49 and 0.34 ± 0.07‰. The particulate Fe seems to be brought mainly by runoff and transported across continental shelves and slopes. Aerosols are suspected to enrich the surface Vitiaz Strait waters, while hydrothermal activity likely enriched New Ireland waters. Dissolved Fe isotopic ratios are found between A0.03 and 0.53 ± 0.07‰. They are almost systematically heavier than the corresponding particulate Fe, and the difference between the signature of both phases is similar for most samples with Δ 56 Fe DFe – PFe = +0.27 ± 0.25‰ (2SD). This is interpreted as an equilibrium isotopic fractionation revealing exchange fluxes between both phases. The dissolved phase being heavier than the particles suggests that the exchanges result in a net nonreductive release of dissolved Fe. This process seems to be locally significantly more intense than Fe reductive dissolution documented along reducing margins. It may therefore constitute a very significant iron source to the ocean, thereby influencing the actual estimation of the iron residence time and sinks. The underlying processes could also apply to other elements.

Atlantic Ocean circulation changes preceded millennial tropical South America rainfall events during the last glacial

During the last glacial period, Greenlands climate shifted between cold (stadial) and warm (interstadial) phases that were accompanied by ocean circulation changes characterized by reduced Atlantic Meridional Overturning Circulation (AMOC) during stadials. Here we present new data from the western tropical Atlantic demonstrating that AMOC slowdowns preceded some of the large South American rainfall events that took place during stadials. Based on 231Pa/230Th and Ti/Ca measurements in the same sediment core, we determine that the AMOC started to slowdown 1420 ± 250 and 690 ± 180 (1σ) years before the onset of two large precipitation events associated with Heinrich stadials. Our results bring unprecedented evidence that AMOC changes could be at the origin of the large precipitation events observed in tropical South America during Heinrich stadials. In addition, we propose a mechanism explaining the differences in the extent and timing of AMOC slowdowns associated with shorter and longer stadials.

Neodymium isotopic composition in foraminifera and authigenic phases of the South China Sea sediments: Implications for the hydrology of the North Pacific Ocean over the past 25 kyr

epsilon Nd and normalized Rare Earth Elements (REE) patterns of benthic and planktonic foraminifera and Fe-Mn coatings precipitated on sediments have been investigated for the South China Sea (SCS) to (1) assess the reliability of the extraction of past seawater epsilon Nd in the SCS and to (2) reconstruct past hydrological changes during the last 25 kyr. Reductively cleaned mono-specific planktonic foraminifera (Globigerinoides ruber) and mixed benthic foraminifera in core-top sediments from 1500 to 2400 m display similar epsilon Nd values to those of the modern Pacific Deep Water (PDW) (epsilon Nd of -3.9 to -4.4). Furthermore, the epsilon Nd of the reductive cleaning solutions shows similar epsilon Nd values to ones obtained on cleaned foraminifera. Combined with PAAS-normalized REE patterns, these results confirm that the oxidative and reductive cleaning procedure applied to foraminifera does not totally remove all of the Fe-Mn coatings and that epsilon Nd values yielded by cleaned planktonic foraminifera retain the epsilon Nd imprint of the bottom and/or pore water. epsilon Nd values obtained from a leaching procedure carried out on the bulk non-decarbonated sediments are comparable to the epsilon Nd values of the modern PDW, whereas a similar leaching procedure applied to decarbonated sediments reveals a bias due to contamination with Nd deriving from lithogenic particles. In core MD052904, seawater epsilon Nd, reconstructed from planktonic foraminifera, indicates that the last glacial period is characterized by lower epsilon Nd (-5.2 +/- 0.2 to -6.4 +/- 0.3) than the late Holocene (-4.1 +/- 0.2). Assuming that Nd input from river does not change strongly the epsilon Nd of the PDW of the northern SCS, these epsilon Nd variations suggest a higher relative proportions of southern-sourced water in the deep water of the western subtropical Pacific Ocean during the last glacial period.

Quantifying sediment sources in a lowland agricultural catchment pond using (137)Cs activities and radiogenic (87)Sr/(86)Sr ratios.

Soil erosion often supplies high sediment loads to rivers, degrading water quality and contributing to the siltation of reservoirs and lowland river channels. These impacts are exacerbated in agricultural catchments where modifications in land management and agricultural practices were shown to accelerate sediment supply. In this study, sediment sources were identified with a novel tracing approach combining cesium ((137)Cs) and strontium isotopes ((87)Sr/(86)Sr) in the Louroux pond, at the outlet of a lowland cultivated catchment (24km(2), Loire River basin, France) representative of drained agricultural areas of Northwestern Europe. Surface soil (n=36) and subsurface channel bank (n=17) samples were collected to characterize potential sources. Deposited sediment (n=41) was sampled across the entire surface of the pond to examine spatial variation in sediment deposits. In addition, a 1.10m sediment core was sampled in the middle of the pond to reconstruct source variations throughout time. (137)Cs was used to discriminate between surface and subsurface sources, whereas (87)Sr/(86)Sr ratios discriminated between lithological sources. A distribution modeling approach quantified the relative contribution of these sources to the sampled sediment. Results indicate that surface sources contributed to the majority of pond (μ 82%, σ 1%) and core (μ 88%, σ 2%) sediment with elevated subsurface contributions modeled near specific sites close to the banks of the Louroux pond. Contributions of the lithological sources were well mixed in surface sediment across the pond (i.e., carbonate sediment contribution, μ 48%, σ 1% and non-carbonate sediment contribution, μ 52%, σ 3%) although there were significant variations of these source contributions modeled for the sediment core between 1955 and 2013. These fluctuations reflect both the progressive implementation of land consolidation schemes in the catchment and the eutrophication of the pond. This original sediment fingerprinting study demonstrates the potential of combining radionuclide and strontium isotopic geochemistry measurements to quantify sediment sources in cultivated catchments.

GEOTRACES Intercalibration of the Stable Silicon Isotope Composition of Dissolved Silicic Acid in Seawater

The first inter-calibration study of the stable silicon isotope composition of dissolved silicic acid in seawater, δ30Si(OH)4, is presented as a contribution to the international GEOTRACES program. Eleven laboratories from seven countries analyzed two seawater samples from the north Pacific subtropical gyre (Station ALOHA) collected at 300 m and at 1000 m water depth. Sampling depths were chosen to obtain samples with a relatively low (9 μmol L-1, 300 m) and a relatively high (113 μmol L-1, 1000 m) silicic acid concentration as sample preparation differs for low- and high- concentration samples. Data for the 1000m water sample were not normally distributed so the median is used to represent the central tendency for the two samples. Median δ30Si(OH)4 values of +1.66 ‰ for the low-concentration sample and +1.25 ‰ for the high-concentration sample were obtained. Agreement among laboratories is overall considered very good; however, small but statistically significant differences among the mean isotope values obtained by different laboratories were detected likely reflecting interlaboratory differences in chemical preparation including pre-concentration and purification methods together with different volumes of seawater volume analyzed, and the use of different mass spectrometers including the Neptune MC-ICP-MS (Thermo Fisher™, Germany), the Nu Plasma MC-ICP-MS (Nu Instruments™, Wrexham, UK), and the Finnigan™ (now Thermo Fisher™, Germany) MAT 252 IRMS. Future studies analyzing δ30Si(OH)4 in seawater should also analyze and report values for these same two reference waters in order to facilitate comparison of data generated among and within laboratories over time.

Seasonal variations, origin, and fate of settling diatoms in the Southern Ocean tracked by silicon isotope records in deep sediment traps

The Southern Ocean plays a pivotal role in the control of atmospheric CO 2 levels, via both physical and biological sequestration processes. The biological carbon transfer to the ocean interior is tightly coupled to the availability of other elements, especially iron as a trace-limiting nutrient and dissolved silicon as the mineral substrate that allows diatoms to dominate primary production. Importantly, variations in the silicon cycling are large but not well understood. Here we use δ 30 Si measurements to track seasonal flows of silica to the deep sea, as captured by sediment trap time series, for the three major zones (Antarctic, AZ; Polar Frontal, PFZ; and Sub-Antarctic, SAZ) of the open Southern Ocean. Variations in the exported flux of biogenic silica (BSi) and its δ 30 Si composition reveal a range of insights, including that (i) the sinking rate of BSi exceeds 200 m d −1 in summer in the AZ yet decreases to very low values in winter that allow particles to remain in the water column through to the following spring, (ii) occasional vertical mixing events affect the δ 30 Si composition of exported BSi in both the SAZ and AZ, and (iii) the δ 30 Si signature of diatoms is well conserved through the water column despite strong BSi and particulate organic carbon (POC) attenuation at depth and is closely linked to the Si consumption in surface waters. With the strong coupling observed between BSi and POC fluxes in PFZ and AZ, these data provide new constraints for application to biogeochemical models of seasonal controls on production and export.

Development of laser ablation multi-collector inductively coupled plasma mass spectrometry for boron isotopic measurement in marine biocarbonates: new improvements and application to a modern Porites coral†

RATIONALE Laser Ablation coupled to Multi-Collector Inductively Coupled Plasma Mass Spectrometry (LA-MC-ICPMS) is a powerful tool for the high-precision measurement of the isotopic ratios of many elements in geological samples, with the isotope ratio ((11) B/(10) B) of boron being used as an indicator of the pH of oceanic waters. Most geological samples or standards are polished and ablation occurs on flat surfaces. However, the shape and the irregularities of marine biocarbonates (e.g., corals, foraminifera) can make precise isotopic measurements of boron difficult. Even after polishing, the porosity properties and the presence of holes or micro-fractures affect the signal and the isotopic ratio when ablating the material, especially in raster mode. METHODS The effect of porosity and of the crater itself on the (11) B signal and the isotopic ratio acquired by LA-MC-ICPMS in both raster and spot mode was studied. Characterization of the craters was then performed with an optical profilometer to determine their shapes and depths. Surface state effects were examined by analyzing the isotopic fractionation of boron in silicate (NIST-SRM 612 and 610 standards) and in carbonate (corals). RESULTS Surface irregularities led to a considerable loss of signal when the crater depth exceeded 20 µm. The stability and precision were degraded when ablation occurred in a deep cavity. The effect of laser focusing and of blank correction was also highlighted and our observations indicate that the accuracy of the boron isotopic ratio does not depend on the shape of the surface. After validation of the analytical protocol for boron isotopes, a raster application on a Porites coral, which grew for 18 months in an aquarium after field sampling, was carried out. CONCLUSIONS This original LA-MC-ICPMS study revealed a well-marked boron isotope ratio temporal variability, probably related to growth rate and density changes, irrespective of the pH of the surrounding seawater. Copyright

Primary Life Stage Boron Isotope and Trace Elements Incorporation in Aposymbiotic Acropora millepora Coral under Ocean Acidification and Warming

Early-life stages of reef-building corals are vital to coral existence and reef maintenance. It is therefore crucial to study juvenile coral response to future climate change pressures. Moreover, corals are known to be reliable recorders of environmental conditions in their skeletal materials. Aposymbiotic Acropora millepora larvae were cultured in different seawater temperature (27 and 29oC) and pCO2 (390 and 750 µatm) conditions to understand the impacts of ‘end of century’ ocean acidification (OA) and ocean warming (OW) conditions on skeletal morphology and geochemistry. The experimental conditions impacted primary polyp juvenile coral skeletal morphology and growth resulting in asymmetric translucent appearances with brittle skeleton features. The impact of OA resulted in microstructure differences with decreased precipitation or lengthening of fasciculi and disorganized aragonite crystals that led to more concentrations of centers of calcifications. The coral skeletal δ11B composition measured by laser ablation MC-ICP-MS was significantly affected by pCO2 (p = 0.0024) and water temperature (p = 1.46 x 10-5). Reconstructed pH of the primary polyp skeleton using the δ11B proxy suggests a difference in coral calcification site and seawater pH consistent with previously observed coral pH up-regulation. Similarly, trace element results measured by laser ablation ICP-MS indicate the impact of pCO2. Primary polyp juvenile Sr/Ca ratio indicates a bias in reconstructed sea surface temperature (SST) under higher pCO2 conditions. Coral microstructure content changes (center of calcification and fasciculi) due to OA possibly contributed to the variability in B/Ca ratios. Our results imply that increasing OA and OW may lead to coral acclimation issues and species-specific inaccuracies of the commonly used Sr/Ca-SST proxy.