Calculated sea surface temperatures of two sediment cores

The reconstruction of ocean history employs a large variety of methods with origins in the biological, chemical, and physical sciences, and uses modern statistical techniques for the interpretation of extensive and complex data sets. Various sediment properties deliver useful information for reconstructing environmental parameters. Those properties that have a close relationship to environmental parameters are called ''proxy variables'' (''proxies'' for short). Proxies are measurable descriptors for desired (but unobservable) variables. Surface water temperature is probably the most important parameter for describing the conditions of past oceans and is crucial for climate modelling. Proxies for temperature are: abundance of microfossils dwelling in surface waters, oxygen isotope composition of planktic foraminifers, the ratio of magnesium or strontium to calcium in calcareous shells or the ratio of certain organic molecules (e.g. alkenones produced by coccolithophorids). Surface water salinity, which is important in modelling of ocean circulation, is much more difficult to reconstruct. At present there is no established method for a direct determination of this parameter. Measurements associated with the paleochemistry of bottom waters to reconstruct bottom water age and flow are made on benthic foraminifers, ostracodes, and deep-sea corals. Important geochemical tracers are d13C and Cd/Ca ratios. When using benthic foraminifers, knowledge of the sediment depth habitat of species is crucial. Reconstructions of productivity patterns are of great interest because of important links to current patterns, mixing of water masses, wind, the global carbon cycle, and biogeography. Productivity is reflected in the flux of carbon into the sediment. There are a number of fluxes other than those of organic carbon that can be useful in assessing productivity fluctuations. Among others, carbonate and opal flux have been used, as well as particulate barite. Furthermore, microfossil assemblages contain clues to the intensity of production as some species occur preferentially in high-productivity regions while others avoid these. One marker for the fertility of sub-surface waters (that is, nutrient availability) is the carbon isotope ratio within that water (13C/12C, expressed as d13C). Carbon isotope ratios in today's ocean are negatively correlated with nitrate and phosphate contents. Another tracer of phosphate content in ocean waters is the Cd/Ca ratio. The correlation between this ratio and phosphate concentrations is quite well documented. A rather new development to obtain clues on ocean fertility (nitrate utilization) is the analysis of the 15N/14N ratio in organic matter. The fractionation dynamics are analogous to those of carbon isotopes. These various ratios are captured within the organisms growing within the tagged water. A number of reconstructions of the partial pressure of CO2 have been attempted using d13C differences between planktic and benthic foraminifers and d13C values of bulk organic material or individual organic components. To define the carbon system in sea water, two elements of the system have to be known in addition to temperature. These can be any combination of total CO2 , alkalinity, or pH. To reconstruct pH, the boron isotope composition of carbonates has been used. Ba patterns have been used to infer the distribution of alkalinity in past oceans. Information relating to atmospheric circulationand climate is transported to the ocean by wind or rivers, in the form of minerals or as plant andanimal remains. The most useful tracers in this respect are silt-sized particles and pollen.

(Table 4) Estimated surface water phosphate concentrations for the past 200 kyr at the location of sediment core GeoB1016-3

We compared ocean atlas values of surface water [PO4]3- and [CO2(aq)] against the carbon isotopic fractionation (ep) of alkenones obtained from surface sediments of the South Atlantic and the central Pacific (Pacific data are from Pagani et al. 2002, doi:10.1029/2002PA000756). We observed a positive correlation between ep and 1/[CO2(aq)], which is opposite of what would be expected if the concentration of CO2(aq) were the major factor controlling the carbon isotopic fractionation of C37:2 alkenones. Instead, we found inverse relationships between ep and [PO4]3- for the two ocean basins (for the Atlantic, ep = -4.6*[PO4]3- + 15.1, R = 0.76; for the Pacific, ep = -4.1*[PO4]3- + 13.7, R = 0.64), suggesting that ep is predominantly controlled by growth rate, which in turn is related to nutrient concentration. The similarity of the slopes implies that a general relationship between both parameters may exist. Using the relationship obtained from the South Atlantic, we estimated surface water nutrient concentrations for the past 200,000 years from a deep-sea sediment core recovered off Angola. Low ep values, indicating high nutrient concentrations, coincide with high contents of total organic carbon and C37 alkenones, low surface water temperatures, and decreased bulk d15N values, suggesting an increased upwelling of nutrient-rich cool subsurface waters as the main cause for the observed ep decrease.

Terrestrial organic matter in marine sediments of the Central Equatorial Atlantic

Terrestrial organic matter (OM) in pelagic sediments is discussed with regard to depositional processes and land-sea interactions in the modern and past glacial/interglacial Equatorial Atlantic. Special emphasis is placed on a critical evaluation of different analytical approaches (C/N, Rock-Eval Pyrolysis, stable carbon isotopes, palynology, organic petrology, and selected biomarkers) which are currently used for the qualitative and quantitative assessment of terrigenous organic carbon. If binary mixing equations are used to calculate terrestrial and marine proportions of organic carbon, we consider the definition of endmember values to be most critical since these values may be biased by a great number of independent controls. A combination of geochemical methods including optical studies (organic petrology and palynology) is therefore suggested to evaluate each individual proxy. Organic geochemical analyses performed on sediments from the modern and Late Quaternary Equatorial Atlantic evidence fluctuations in eolian supply of terrigenous OM related to changes in intensity of the trade winds. Quantification of this organic fraction leads to differing proportions depending on the approach applied, i.e. the organic carbon isotopic composition or maceral analyses. Modern distribution of terrigenous OM reveals a decrease in supply towards the basin contributing less than a fifth of the total OM in pelagic areas. Organic geochemical data indicate that sedimentation in the modern northeastern Brasil Basin is affected by lateral advection of reworked OM probably from southern source areas. Glacial/interglacial deposits from the pelagic Equatorial Atlantic (ODP Site 663), covering isotopic stages 12 and 11, reveal that deposition of terrigenous OM was higher under past glacial conditions, in correspondence to generally enhanced dust fluxes. Proportions of terrigenous OM, however, never exceed 50% of the total OM according to maceral analyses. Other estimates, recently proposed by Verardo and Ruddiman (1996), are considered to be too high probably for analytical reasons. Palynological records in the Equatorial Atlantic parallel dust records. Increased portions of grass pollen suggest the admixture of C4-plant material under modern and past glacial conditions. It is therefore assumed, as one possible interpetation, that C4-plant debris has an effect on sedimentary d13Corg and might explain differences between isotopic and microscopic quantitative estimates. Using the difference between these two records, we calculate that maximum supply of C4-material remains below 20% of the total OM for the deep modern and past glacial/interglacial Equatorial Atlantic.

Compilation of 231Protactinium/230 Thorium excess ratios in various surface sediment samples from the South Atlantic

In large areas of the world's oceans, there is a relationship between the mass flux of particulate matter and the unsupported 231Pa/230Th (xs231Pa/xs230Th) activity ratio of recent sediments. This observation forms the basis for using the xs231Pa/xs230Th ratio as a proxy for past changes in export productivity. However, a simple relationship between xs231Pa/xs 230Th ratio and particle flux requires that the water residence time in an ocean basin is far in excess of the scavenging residence time of 231Pa, and that the composition of sinking particles maintains a strong preference for the adsorption of 230Th over 231Pa with a constant 230Th/231Pa fractionation factor (F). The best correlation between xs231Pa/xs230Th ratio and mass flux is found in the Pacific Ocean. In the Atlantic, the contrast in the xs231Pa/xs230Th ratios between open ocean (low flux regions) and ocean margins (high flux regions) is much less pronounced due to the shorter residence time of deep water, resulting in less effective boundary scavenging of 231Pa. In the Southern Ocean, south of the Polar Front, there is no more a simple relationship between xs231Pa/xs230Th and particle flux. This is a result of a southward decrease in F, probably reflecting the increased opal content of sinking particles. Opal does not fractionate 231Pa and 230Th significantly. This lack of fractionation results in high xs231Pa/xs230Th ratios in opal-dominated regions, even in areas of very low particle fluxes such as the Weddell Sea. The xs231Pa/xs230Th ratio can therefore only be used as a paleoproductivity proxy if, in the time interval of interest, changes in the basin ventilation rate and differential scavenging of both radionuclides due to changes in the chemical composition of particulate matter can be excluded.