B44C-01 INVITED
New Insights to Interplay of Factors that Influence Biomineral Signatures: Intrigue of the Solvation Environment
For many years, studies of compositional signatures contained in the skeletons of certain calcifying species have suggested that Mg levels correlate with temperature of formation. With the growing urgency to decipher how earth environments have changed over geologic time, this long-standing proxy for estimating temperature has developed and also encouraged a frenzy of new measurements that claim additional composition and isotope- based proxy relations. Because organisms employ mineralization strategies that create highly controlled chemical and spatial microenvironments for the nucleation and growth of biominerals, it would seem probable that local chemical/biochemical factors could enforce controls on impurity contents that are equal to or greater than the relatively small thermal differences found in earth surface conditions. Over the last seven years, our group has conducted molecular scale studies using in situ atomic force microscopy and computational modeling to understand the physical basis for shape and signature controls on biomineral formation. By linking direct measurements of growth modification with the underlying chemical interactions, our findings have provided new insights into relative importance of factors that produce observed compositional signatures. Several principles are emerging to give a mechanism-based understanding of growth processes and are also allowing us to establish the interplay of factors that produce mineral shapes and measured compositional signatures: 1) Changes in dominant growth processes as determined by chemical driving force; 2) Roles of growth dynamics and corresponding kink site density at step edges; 3) First-order influence of minor changes in local solvation environment induced by simple biomolecules and salinity; 4) Comparatively small influence of temperature; 5) Realization that classical models of crystal growth, developed for kink-rich highly soluble salts, need to be extended to explain most earth materials which are only sparingly soluble (kink limited). By developing these insights, the possibilities for a new mechanism-based understanding are significant. New findings showing the importance of solvation environment give reason to suspect that signatures also have a significant sensitivity to local alkalinity, salinity, and the types of electrolytes present in solution but this remains to be tested.
B44C-02
Experimental Study of Sr, Mg, and Ba Partitioning Into Calcite at Various Linear Growth Rates Using Elemental and Isotopic Spikes: Preliminary Results.
The ratios of Mg/Ca and Sr/Ca in calcium carbonates are widely used as paleothermometers. However, inconsistencies in the relationship between temperature and composition amongst inorganic and biogenic calcites suggest the influence of factors other than temperature, including precipitation rate. We examine the influence of growth rate on Mg, Sr, and Ba partition coefficients between calcite and aqueous fluid (Kdi= XiCalcite*XCaFluid/XCaCalcite*XiFluid, where i is the cation of interest and X is the mole fraction). Previous experimental studies of growth-rate-dependent partitioning were based on bulk analysis of carbonate precipitates; here we develop an alternative approach based on in situ measurements of single crystals having known linear growth rates. Experiments were performed by continuously adding two titrates (Na2CO3 and CaCl2) into growth solution of NaCl doped with trace elements. Linear growth rates of calcite crystals are determined by adding of elemental and isotopic spikes into the growth medium (episodically for elemental spikes; continuously for isotopic spikes). Harvested crystals were randomly mounted into high vacuum epoxy and examined with NanoSIMS and SEM techniques. Our NanoSIMS measurements of product calcite crystals document approximately concentric growth, the rate of which is marked by the sequential appearance of the various elemental and isotopic spikes we added to the parental fluid during growth. For example, ratios of 55Mn/42Ca in carbonate are 1e-5 in crystal cores and sharply increase up to 1.6e-3 in crystal rims, in response to our addition of Mn to the parental liquid at a known time. Similarly, the 26Mg/24Mg ratio is close to natural abundances in crystal cores and monotonically increases toward the crystal edges, reflecting monotonic addition of 26Mg spike throughout the experiment. Such data, combined with known timing of elemental spike additions and rates of increase of the 26Mg/24Mg ratio of the solution, permitted us to quantitatively reconstruct the growth histories of individual calcite crystals. Calcite growth rates were calculated as the ratio of the distance between two adjacent analytical spots and the corresponding time interval required to explain their difference in composition. This is straightforward for the portions of the experiments when 26Mg/24Mg changed monotonically at a known rate, but generally provides only limits to the growth rate during the period when elemental spikes were added episodically. The linear growth rates estimated in this way varied from 0.6 to 6 microns per hour, with uncertainties as good as 6 percents, relative. The lower edge of our growth rate interval is similar to those of benthonic foraminifera, suggesting that partitioning data based on our experiments will constrain growth-rate effects relevant to common paleothermometry applications. Note that the growth rates we derive could be under- estimates because analyzed sections of the crystals do no necessarily expose the center of calcite crystallization. We anticipate determining Sr/Ca, Mg/Ca, and Ba/Ca ratios in the fluid, after which we shall report the partition coefficients for Sr, Mg, and Ba as function of growth rate of the calcite crystals.
B44C-03
Deriving Paleotemperatures From Coral Skeleton Using a Rayleigh Fractionation Model for Coral Biomineralization
Paleotemperature proxy records are typically derived from coral skeleton using empirical relationships between elemental ratios and water temperature calibrated using the skeletons of living organisms grown under known conditions. While this approach has produced significant advances in our understanding of Earth's climate system, its accuracy is limited by the influence of physiological processes ("vital effects") on compositional variability within the skeleton. "Vital effects" are evident as differences in the composition of carbonates precipitated experimentally and accreted by organisms at the same conditions, as well as differences in composition amongst skeletons of the same group of organisms, or even the same species, accreted under identical environmental conditions. Several recent studies have identified the importance of Rayleigh fractionation in producing "vital effects" in coral skeleton [1-3]. On the basis of this advance in our understanding of coral biomineralization, we have developed a new approach to deriving paleotemperature estimates from coral skeletons through knowledge of their combined Mg/Ca, Sr/Ca and Ba/Ca ratios. Using experimentally determined partition coefficients for Mg, Ca, Sr, and Ba between abiogenic aragonite and seawater [1] combined with a Rayleigh fractionation model for the precipitation of aragonite from a calcifying fluid, the temperature at which the aragonite was precipitated can be accurately determined. To test the accuracy and precision of this approach, which relies only on experimentally determined partition coefficients for abiogenic aragonite and the Rayleigh equation (i.e. contrary to conventional paleothermometer calibrations, no prior knowledge of water temperature is required), aragonite skeletons from 2 different coral species were analyzed for Mg/Ca, Sr/Ca and Ba/Ca ratios using the Cameca 3f ion microprobe at WHOI: (1) tropical coral Diploria labyrinthiformis collected from Bermuda and (2) deep water coral Lophelia pertusa collected from Tisler Reef, NE Skagerrak. Using only the elemental ratios and experimentally determined partition coefficients, the model predicts seasonally-resolved ocean temperatures to within 0.5 °C of the recorded temperatures at the sites the corals were collected. The predictive capability of our approach is outstanding, and is completely independent of any empirical calibration of coral skeleton from these environments. From these results it is clear that our new approach to coral skeleton paleothermometry yields ocean temperatures that are both accurate and precise to within a few tenths of a degree, and that the approach is applicable both across species and to corals growing in vastly different environments. References: [1]Gaetani, GA and Cohen, AL (2006) Geochim Cosmochim Acta 70, 4617-4634; [2] Cohen, AL, Gaetani, GA, Lundälv, T, Corliss, BH and George, RY, (2006), Geochem, Geophys, Geosys 7, Q12004, doi:10.1029/2006GC001354.; [3] Gagnon, AC, Adkins, JF, Fernandez, DP, Robinson, LF (2007) Earth Planet Sci Lett 261, 280-295.
B44C-04
Atmospheric Sulphur Archives in Tree Rings: a First Comparison With Speleothems
The high atmospheric loading of sulphur pollutants during the 20th century is recorded in archives such as ice cores, archived soils, and some peats. In karstic settings, both trees and speleothems offer proxy records capable of encoding environmental information and our work demonstrates how sulphur is encoded in these archives. Recent work by others (Kawamura et al., 2005, Environmental Science and Technology) on two Japanese tree species has demonstrated that they preserve S concentration and isotopic archives broadly consistent with 20th century atmospheric variations. Frisia et al. (2005, Earth and Planetary Science Letters) first demonstrated the preservation of secular atmospheric sulphur records as carbonate-associated sulphate (CAS) in speleothem calcite, but at the forested alpine Ernesto site (NE Italy) it was noted that there was a lagged response in the speleothem compared with the atmosphere. Our new study of sulphur and oxygen isotopes in sulphate in atmospheric precipitation and cave dripwaters demonstrates that this lag is associated with the mineralization of sulphate in organic form followed by re-oxidation to sulphate leading to a resetting of the sulphate oxygen signal. The speleothems themselves show pronounced annual cycles in sulphate, despite the constant level of sulphate in dripwater. New experiments on sulphate incorporation in calcite precipitated at rates comparable to those in the cave system demonstrates that the main control is the pH of the system, which is modulated by seasonal variations in the PCO2 of the cave air. A study of S distribution and concentration within two tree species Picea abies and Abies alba from close to the Ernesto site has been made by a combination of synchrotron micro-XRF and NEXAFS and high mass-resolution ICP-MS analysis. Resin was removed from the tree samples before analysis since it is likely to be associated with element mobility. XRF mapping clearly demonstrates enhanced levels of S in the primary cell wall, where it is likely to be associated with proteins, and hence that it is likely to remain fixed over time, since this layer is overlain by a thicker secondary layer of carbohydrates (cellulose and lignins). Abies alba has the higher S concentrations (50-100 ppm in bulk wood) and these increase during the 20th century. Pronounced annual variations observed in synchrotron analyses largely reflect changes in wood density since bulk ICP-MS analyses show relatively little difference between early and late wood of the annual cycle. NEXAFS analyses demonstrate the occurrence of S in a mixture of oxidation states, primarily 0 to +1, +2, +5 and +6. We predict that S isotope analysis of the tree rings will preserve that of the soil S pool at the time of tree growth.
B44C-05
Uncovering of High Strontium Content in Biogenic Calcite: New Data From a Murolith- producing Coccolithophore
Biogeochemical studies of modern coccolithophores have demonstrated a link between growth rate and the incorporation of strontium into coccolith calcite. Values for the Sr content of coccolithophores have been quoted as high as 30 mmol/mol. A paleochemical study of the Early Jurassic deposits of the Paris Basin reveals a biogeochemical anomaly. It appears indeed that a coccolithophorid genus, Crepidolithus, contains unexpected amounts of strontium (ca. 12-20% weight) in their calcite lattice that corresponds to a Sr/Ca ratio around 130 mmol/mol. These results have been obtained using a innovative protocol which enables the concentration of Crepidolithus from a polyspecific assemblage. Chemical measurements have then been performed by ICP-GAAS after a weak acid leaching. SEM spot analyses confirm the reproducibility of high values. At the same time, other calcareous particles (such as Schizosphaerella, coccoliths and euhedral monocrystals) do not show such parallel enrichments in strontium. This observation enables us to preclude diagenesis as the cause of the high Sr values. Moreover, further investigations (SEM, XRD, selective leaching) rule out an additional contaminant phase such as sulfates. Finally, in view of the good preservation of Crepidolithus in the sediment, an aragonite precursor for this coccolith genus can also be excluded. If true, these results raise the question about the mode of incorporation of strontium into the calcite lattice because these values require huge KSr (>> 0.13) and/or unrealistic Sr content in the seawater (ca. 150 ppm !). We discuss here several explanations to explain this bioincorporation of strontium into these coccoliths. 1. A mineralization of (Crepidolithus from a distinct water mass extraordinarily enriched in Sr. For that, we present the physico-chemical structure of the water column during the Early Toarcian OAE and the geographic repartition of this anomaly. 2. A differential vital effect explained by physiological pathways (cross-membrane transport). 3. A kinetic effect due to the growth mode of this murolith where vertical V-units are dominant. In order to verify that last hypothesis, we also investigate the Sr content from others same-structured coccoliths (e.g., Crucirhabdus).
B44C-06
Environmental versus biological controls on multi-proxy records from bivalve shell carbonate
Field experiments on different bivalve species have been conducted in the Dutch Wadden Sea near the Netherlands Institute for Sea Research (NIOZ) to investigate the environmental and biological controls on the fractionation pattern of different stable isotope systems (18O/16O, 26Mg/24Mg, 44Ca/42Ca) and elemental/calcium ratios, and their reliability as geochemical proxies for high- resolution palaeoclimate reconstructions. Our study includes various individuals of the blue mussel, Mytilus edulis and juveniles of the ocean quahog, Arctica islandica. Our study is completed using samples from growth experiments on young Arcitca's which were also performed at the NIOZ laboratory. These individuals were cultured in five temperature controlled basins, ranging from 1 to 12° Celcius. The third sample set analysed was collected at several sites in the Baltic Sea characterised by salinities lower than 15 permil. Our data shows that both species deposit their shells in isotopic equilibrium with ambient seawater. Trace element ratios in Mytilus edulis further indicate that Mg/Ca may indeed provide a reliable palaeothermometer. In contrast, Calcium and magnesium isotope fractionation is obviously controlled by biological processes. However, environmental signals (e.g. temperature, salinity) are well recorded in the shells. Therefore, the combined approach could give new insights into biomineralization processes. This is a contribution to EuroCLIMATE project 04 ECLIM FP08 CASIOPEIA.
B44C-07
Mg isotopes in biocarbonates: new insight into vital effects associated to echinoderms and bivalves calcification
Mg isotopes can be helpful tracers to reveal the fundamental pathways of Mg incorporation during biomineralisation. We report in this study a detailed characterisation of the Mg isotopic signatures of different biominerals: high magnesium calcitic skeletons of selected echinoderms (sea urchins and starfish) and low magnesium aragonitic shells of a bivalve species (clam). State of the art analytical procedures were applied including sample purification step followed by high precision measurements using MC-ICP-MS (Nu instrument) in dry plasma conditions. 26Mg/24Mg and 25Mg/24Mg are expressed as per mil deviations from the DSM3 (Dead Sea Metal 3) reference standard in delta notation (d26Mg and d25Mg). For echinoderms, we considered: (a) adult specimens of six starfish species (Asteria r., Marthasterias g., Anseropoda p., Asterina g., Echinaster s. and Henricia o.), sampled in Brittany (France); (b) a sea urchin species (Paracentrotus lividus) with field samples (Mediterranean Sea, Marseille, France) and culture specimen under T and S controlled conditions. In vivo endoskeletons display negative, but different d26Mg values of -3.06 for starfish (with uniform interspecies signatures) and -2.65 for sea urchin. Relative to seawater signature (-0.82), all echinoderms favour the incorporation of light isotopes during biocalcification. The d26Mg depletion is lower than theoretically expected from a inorganic calcite precipitation from seawater (at -3.5). These differences suggest that on its route from seawater to the shell, Mg isotopes are partly biologically fractionationated through "vital effects" leaving heavier Mg isotopic signatures. Taken into account that calcification in echinoderms is an intra- cellular process involving transient amorphous calcium carbonate (ACC) phase, the observed bio-fractionation factors can be related to: (1) changes in the isotopic composition of the precipitating intracellular fluids due to active pumping in and out of the cell; (2) a specific control of isotopic exchange by ACC. For bivalves, we considered one clam species (Ruditapes Philippinarum) from two different sampling sites in the gulf of Morbihan (Brittany, France). One site (Locmariaquer, Loc) is coastal and marine while the other (Le bono, BO) located upstream in the Auray river. For each site and specimen, we considered all reservoirs involved in the shell build-up: seawater, internal fluids with hemolymph (H) and extrapaleal fluid (EPF), soft tissues with mantle, muscle and remaining part (R) and finally aragonitic shells. Water d26Mg are -0.82 at Loc and -0.79 at BO and appear to be highly similar to internal fluids values (EPF and H), -0.6 for Loc and -0.8 for BO. The soft tissues with mean values of -2.7 (Mantle), -2.1 (Muscle), -2.8 (R) at both sites, show pronounced enrichments in the light isotopic fraction of Mg. The shells display the widest range of Mg isotopic signatures with -1.9 at Loc and -4.2 at BO suggesting that different routes of fractionation are acting. The shell signature at Loc similar to aragonitic coral (-1.9) suggests that moderate biological effects have influenced the shell composition at this site. In contrast, at BO, the very light signature of the shell suggests that a significant fraction of Mg has been internally recycled by the clam and used for the building of the shell.