PP41F-01
Experimental Perspective on Cause of Planktonic Foraminifera Intrashell Mg/Ca Variability and Impact on Paleoceanographic Applications
Analysis of Me/Ca ratios in fossil foraminifera has become one of the primary methods used to reconstruct paleoceanographic conditions across the Cenozoic. Published laser ablation, electron probe, and nanoSIMS data suggest Mg varies considerably within a shell, and that the interpretation of Mg/Ca and other metals thus may be more complicated than previously thought. During the summers of 2007 and 2008, we conducted laboratory experiments on living Orbulina universa to quantify intrashell Mg/Ca variability and establish a mechanistic understanding to explain these geochemical observations. Specimens were cultured on a 12h:12h light:dark cycle in three water temperatures and individually analyzed with an Eximer laser/ICP- MS ablation system. Because the Ba/Ca of foraminifera calcite is homogeneous across a shell, we labeled day- and night-precipitated calcite by varying the Ba/Ca of the culture water during each 12h cycle and explored the impact of symbiont physiology and seawater carbonate chemistry on shell Mg/Ca. Our results demonstrate that in symbiotic O. universa, high Mg/Ca bands are precipitated during the night whereas low Mg/Ca bands are precipitated during the day. Night calcite has 2-3 times more Mg than day calcite. Foraminifera cultured in ambient seawater during the day and seawater with elevated [CO3] at night displayed a significant attenuation of the banding cycle whereas foraminifera grown in high pH seawater during the day record an increase in banding amplitude. Because Mg/Ca of both day and night calcite increases with temperature, the average Mg/Ca of an O. universa shell agrees well with predicted values based on published empirical calibrations for this species. Importantly, these results show that the high Mg/Ca ratios observed in this species relative to other foraminifera is due to the fact that Orbulina precipitates ~30% of its shell at night, which is significantly more than other species.
PP41F-02
Towards a global calibration and validation of the G. ruber (white) Mg/Ca paleothermometer
Paired planktonic foraminiferal Mg/Ca and d18O analyses are commonly applied to jointly estimate sea surface temperature and d18Oseawater, a proxy for ocean salinity. Our recent findings from an Atlantic meridional coretop sample transect indicated that shell Mg/Ca values were significantly elevated in higher salinity gyre waters. We found that "excess Mg/Ca" (the residual between the observed and expected Mg/Ca composition at the d18O calcification temperature) was very highly correlated with surface salinity. Here, we test this relationship for the global oceans. Using a new database of ~150 coretop samples selected from the Indian, Pacific, and Atlantic basins, we present a global coretop calibration of surface ocean temperature and salinity using shell Mg/Ca, d18O, and bottom water ΔCO3= as predictors. All shells were cleaned for Mg/Ca analyses using the full oxidative-reductive cleaning procedure and shell weights were recorded. We will discuss the validation of these new equations using published G. ruber Mg/Ca and d18O data from all basins.
PP41F-03 INVITED
Li/Mg in biogenic aragonite as a potential paleotemperature proxy
Although planktic foraminiferal Mg/Ca has been widely applied as a paleotemperature proxy, benthic
foraminiferal Mg/Ca has seen much more limited application. Recent benthic calibrations have shown that
seawater saturation with respect to CaCO3 (ΔCO32-) is likely an important factor
influencing the incorporation of Mg, thereby limiting the utility of Mg/Ca as a simple paleotemperature tool.
We have shown (Bryan and Marchitto, Paleoceanography, 2008) that in the Florida Straits, benthic
foraminiferal Mg/Li is better correlated to temperature than Mg/Ca is. We suggested that Mg and Li are
influenced similarly by ΔCO32-, such that dividing Mg by Li largely corrects for this effect
and isolates the temperature influence. The promise of this approach is most dramatic for the aragonitic
foraminifer Hoeglundina elegans, in which the variance explained by temperature is only 49% for Mg/Ca
but 90% for Mg/Li .
Recent work on coral Mg and Li at fine-scale resolution (Montagna et al., Goldschmidt Conference,
2008) appears to tell a similar story. Li/Ca is inversely correlated to temperature in both the high-latitude
zooxanthellate coral Cladocora caespitosa and the deep-water coral Lophelia pertusa, but each
taxon exhibits a unique relationship. However the Li/Mg (or Mg/Li) relationship to temperature is
indistinguishable between the two taxa, with a combined r2 of 0.96. The slope of this relationship is
identical within one standard error of the H. elegans Li/Mg relationship, and the intercept is similar.
We further explore the promise of this method by (1) measuring H. elegans Li/Mg in additional core tops
from around the world ocean; (2) measuring H. elegans Li/Mg in Last Glacial Maximum sediments from
the Florida Straits, where Mg-rich overgrowths appear to be a problem for calcitic benthic foraminifera; and
(3) measuring Li/Mg in various aragonitic gastropods from Florida Straits core tops, to see if the temperature
relationship can be generalized to other biogenic aragonites.
PP41F-04
Mean Ocean Temperature Deduced from Kr and Xe Covaries with Atmospheric CO2
Ocean temperature plays a vital role in the climate system, and may play a role in regulating atmospheric CO2, but its past variations are poorly constrained. Krypton (Kr) and xenon (Xe) are highly soluble, and more soluble in colder water. The total amount of Kr and Xe in the atmosphere and ocean together are constant through time, so variations in mean ocean temperature should modulate atmospheric Kr and Xe abundances. Kr and Xe ratios to nitrogen (N2) are measured in air bubbles in ice cores to reconstruct atmospheric Kr/N2 and Xe/N2 histories, which can be interpreted in terms of past mean ocean temperature using known solubilities. Initial GISP2 Kr/N2 data from the LGM indicates that mean ocean temperatures were ~2.7°C colder at that time, which is consistent with independent estimates of local deep ocean temperatures. Time series of δKr/N2 and δXe/N2 during the last glacial termination (from GISP2 and Dome F) and inception (from Vostok) indicate that mean ocean temperature appears to vary in step with atmospheric CO2, but lagged Antarctic mean-annual temperature by several thousand years during the inception. A possible explanation is that summer temperatures cooled first, and atmospheric CO2 is controlled by winter-linked processes in the Antarctic ocean such as sea ice (Stephens and Keeling, 2000) or salty-bottom-water-formation (Toggweiler, 1999), rather than summer-linked processes such as biology. Likewise, mean ocean temperature is expected to be controlled by winter temperatures, because most deep- water formation occurs in winter. Of course, the temperature-dependent solubility of CO2 also makes a small contribution to its observed correlation with mean ocean temperature. A large, early warming of the ocean (~2°C) is implied during the last deglaciation, between about 18-15 ka (the so-called 'Mystery Interval'), synchronous with the CO2 rise at that time. This much warming appears to conflict with the hypothesis that deep waters in the Pacific and Indian oceans are composed of a 50-50 mix of Antarctic- and North Atlantic-sourced waters, due to the fact that the north was cold during this time. Rather, it is consistent with the notion that ~2/3 of all deep-water volume forms around Antarctica (Johnson, 2008), because Antarctica warmed during the Mystery Interval. A 'revised conveyor belt', in which N. Atlantic Deep Water upwells around Antarctica briefly and cools before descending into the abyss to contribute to Indo-Pacific deep water, appears to satisfy all constraints (Lumpkin and Speer, 2007).
PP41F-05
Development of a New Temperature Proxy Based on Diatom-Derived Long-Chain Diols
Diatoms of the genus Proboscia have been shown to biosynthesize C28 and C30 1,14-diols and, therefore, constitute a likely source for these very common marine natural products which are well preserved in sediments. The relative ratio of long-chain 1,14-diols versus long-chain 1,15- or 1,13-diols, derived from other algae, has been used as a productivity proxy of Proboscia diatoms, and appeared to be very useful to track past variations of upwelling intensities in the Arabian Sea and of the Westerly Winds intensity in the Antarctic Peninsula. We cultured Proboscia diatoms at different temperatures revealing that the Proboscia diols composition changes with temperature, i.e. the C30-diols increases relative to the C28-diols with increasing temperature. Indeed, the relative ratio of C30 over C28 1,14-diols correlated well with temperature (r2=0.96, n=8, four Proboscia species) . This raises the potential of these compounds to be used as a paleothermometer to track past temperature changes. To further investigate this we are currently analyzing the distribution of Proboscia diols in more than 200 surface sediment samples worldwide distributed to potentially provide a core top calibration for this new temperature proxy. Our preliminary results show a good correlation (r2= 0.78, n=31) between Proboscia diols and sea surface temperature.
PP41F-06
Sedimentary Silver as a Paleoproductivity Proxy
Surface sediments from continental margin settings along the eastern margin of the Pacific Ocean exhibit a strong, positive correlation between Ag and Ba. Based on this observation it has been hypothesized that Ag, like Ba, is taken up by and delivered to the sediment with sinking organic particles. Consequently, Ag might be useful as a paleoproductivity proxy. To test this idea we studied six cores (MD02-2496, JT96-09, W8709A-13, ODP 1017E, ODP 893A and ME0005A-11) collected from the eastern North Pacific. The downcore Ag records for these cores are compared with various paleoproductivity proxies (organic carbon, opal, carbonate and biogenic Ba), as well as redox-sensitive trace metals (Re, Cd, Mo). Obvious differences exist between the Ag records and those of Re and Mo that record changes in reducing conditions in near-surface sediments. This observation is consistent with surface sediment data that suggest sedimentary redox conditions are not the primary control on Ag accumulation. There are some similarities between the Ag records and those of Cd, a fraction of which is delivered to the sediment with the organic flux. However, the most striking similarities are between Ag and the organic carbon and opal records, supporting the idea that sedimentary Ag is useful as a paleoproductivity proxy. Interestingly, in contrast to surface sediments, the downcore Ag and Ba records are very different, most probably due to the remobilization of Ba when strongly reducing conditions develop after deposition. In this situation, however, the Ag paleoproductivity signal is preserved making Ag an ideal paleo-proxy in continental margin sediments.
PP41F-07
The Zn Isotopic Composition of Diatom Frustules: an Archive of Past Trace Metal Depletion in HNLC Zones?
Zinc is among the essential trace-metal micronutrients for phytoplankton. In common with some other bio- active trace metals, Zn concentrations are highly depleted in those parts of the surface ocean that are replete in the major nutrients (so-called High Nutrient-Low chlorophyll, or HNLC, zones), such as the Southern Ocean. The proposed release of these HNLC zones from trace metal limitation is one of the key hypotheses to explain lower atmospheric CO2 during glacial periods. The preferential incorporation of light Zn isotopes into phytoplankton organic material [1] is expected to leave residual seawater Zn isotopically heavy. The isotopic heaviness of the residual seawater Zn could track the degree of trace metal depletion in the past if a suitable archive of surface seawater Zn isotopes could be identified. Here we investigate diatom opal as a record of the Zn isotopic composition of surface seawater. We have measured Zn isotopic compositions in cleaned diatom frustules from a sequence of core-top samples across the Southern Ocean [2]. All diatom opal exhibits Zn isotopic compositions (δ66ZnLyons- JMC) heavier than the continental input of +0.3 ‰ (at +0.7-1.4 ‰), as expected from surface waters in highly trace metal-depleted HNLC zones. Furthermore the Zn isotope composition tracks decreasing diatom opal burial rates with progressively heavier Zn isotope compositions. These results suggest that Zn isotopes in diatom frustules have potential as a record of past trace metal depletion in HNLC zones, and down-core results from the Southern Ocean will be presented. [1] John et al. (2007) Lim. Oceanogr. 52(6), 2710-2714 [2] Chase et al. (2003) Deep Sea Research II, 799-832
PP41F-08
El Niño-like conditions, decreased ocean productivity and ice decay during the Pliocene warmth
The Pliocene climate optimum (about 4.5 to 3 Ma), with mean global surface temperatures ca. 3° C warmer than today and ca. 7 m higher sea level, has been identified as a possible analogue for future climate change. We used lipid biomarkers in sediment cores from the Caribbean Sea (Ocean Drilling Program sites 999 and 1000) and from the East Equatorial Pacific (EEP; Site 1241) to examine Pliocene changes in global climate and regional tectonics in the Central American Seaway (CAS) and their influence on paleoceanography. Alkenone sea surface temperatures (SST) in both regions document a long-term 3° C cooling from the Late Miocene/Early Pliocene to the late Pleistocene, suggesting that the two regions responded similarly and dramatically to Plio-Pleistocene global climate change. In contrast, TEX86-derived temperatures are much lower and exhibit dramatic 10° C variability in the EEP, which apparently reflects changes in thermocline temperature. Thus, the difference between alkenone-derived SSTs and TEX-derived thermocline temperatures represents a potentially powerful tool for reconstructing the thermal structure of the ocean and upwelling intensity. In the EEP, smallest differences – and thus, warmest inferred thermocline temperatures – occur prior to 7 Ma and during the Pliocene warm period from 4.5 to 3.2 Ma. These intervals coincide with minima in the diol index, a proxy for the abundance for Proboscia diatoms and, thus, upwelling intensity. Both records confirm previous work suggesting that the Pliocene warm period represents a time of protracted El-Niño-like conditions. Moreover, the increase in upwelling intensity at Site 1241 occurred prior to a significant drop in pCO2, inferred to have occurred between 3 and 2 Ma. Future work will explore whether increased EEP upwelling since 3Ma was a consequence of declining pCO2 or perhaps drove the pCO2 decrease via increased organic carbon burial.