Deep-ocean Paleoceanography and Ocean Chemistry

1 PALEOCEANOGRAPHY Lecture 9 DEEP-OCEAN PALEOCEANOGRAPHY AND OCEAN CHEMISTRY: LINKS TO ATMOSPHERIC CARBON DIOXIDE and 14C/12C I. Deep water paleoceanography A. In the 1970’s, it was hoped that benthic foraminiferal populations could be used in "Imbrie-Kipp" mode to reconstruct deep ocean physical and chemical properties. This idea has been long on hope and slim on results, however. 1. There are many more species of benthic foraminifera, but these are scarce compared to planktonic foraminifera (B/P ~ 1%). Studies of benthic foraminifera require much larger samples (10g or more) and take much longer to process and identify. 2. Benthic foraminifera do show large-scale variations throughout the present ocean, and temporal changes are recorded in sediment cores. The first studies were of necessity limited to small regions, and the temptation to overdo the planktonic analogy was great; hence depth-correlated changes in the modern ocean were at first attributed to temperature. (e.g. see Streeter, Schnitker studies). 3. Subsequent work has shown that, unlike planktonic foraminifera, correlations with temperature do not hold globally. This lack of correlation probably is due to the extremely small range of variation of temperature (a few °C) in the deep ocean. There have been attempts to attribute the observed variations to other deep-water parameters (e.g. oxygen content, carbonate saturation, or even the vaguer term "Uvigerina water" which implies that something in the bottom water controls the Uvigerina population). So far, none of these attempts has a "residence time" of more than a few years. So far, benthic foraminifera have defied a systematic generalization of the factors controlling their species composition. A recent paper by P. Loubere characterizes benthic foram census data from the Pacific in terms of surface productivity and deep water oxygen content. 4. It is likely that factors other than deep water properties influence the benthic foraminiferal populations: e.g., food (from the surface) - a correlation between benthic foram flux and productivity has been observed; sedimentation of non-food constituents; other aspects of the sedimentary environment (deep currents, grain size, etc.). One question of some interest to this "other factors" issue is: where do benthic foraminifera live? If they live at some depth in the sediment, the chemical environment they respond to can be different from those at the bottom of the ocean: while T is the same, the chemical composition of the water, the food sources, oxygen, etc. may differ significantly from that of the bottom water. (To be expanded upon later.) 5. Few paleoceanographic interpretations based on benthic foraminifera populations have survived more than a few years. One partial exception is the study of Streeter and Shackleton (1979) who used glacial increases in the Uvigerina abundance to argue that North Atlantic Deep water ceased or drastically diminished. Although the extent of the decrease is now thought to be considerably less than the 2 foraminiferal populations would suggest, it is agreed that there was less North Atlantic Deep Water in the deepest North Atlantic during glacial times, and so at least the sign was right! Images removed due to copyright considerations. Source: Streeter and Shackleton, 1979, figure 1 and 2. 4 B. Carbon isotopes as a tracer of deep ocean paleoceanography 1. The distribution of δ13C in the modern ocean is linked to oxygen and nutrients. a. Organic matter is depleted in 13C. This depletion occurs mainly during the enyzymatic conversion of inorganic CO2 (itself somewhat depleted related to total dissolved CO2, which is mainly HCO3-) to organic matter by plants, mainly a kinetic effect. Much of the carbon uptake by marine photosynthetic organisms is achieved by the transport of free aqueous CO2 (which is depleted in 13C by ~10‰) across cell membranes (although there are also some ways to use anionic HCO3directly). Furthermore, because CO2(aq) concentrations are low (10-20 µmol/kg), the cell membrane environment can be depleted in CO2(aq) during rapid photosynthesis because the conversion of HCO3- to CO2. is relatively slow. Enzymatic (Rubisco) and other kinetic effects within plants add to this 12C depletion, resulting in the δ13C of marine organic ranging from -20‰ (tropical and subtropical waters) to -30‰ (Antarctic). b. When this organic matter sinks into the deep ocean, it quantitatively decomposes (~99% in the whole water column; ~80% in the upper 500 m), releasing the bound C, N, and P in the same ratio, and consuming O2 from the water in the process. The CO2 released during this respiration process is depleted in 13C, so that the δ13C of the deeper water lower than that of the original surface waters. A plot of δ13C vs. P for the modern ocean is linear. c. Hence it is possible to use the δ13C distribution in the ocean as a tracer of oceanic water masses. 2. Because δ13C is recorded in the shells of planktonic and benthic foraminifera, it is possible to determine the nutrient distribution in past oceans. a. Complication: δ13C in forams is not in equilibrium or even close to it; in fact, different species have significantly different offsets from the δ13C of the water. 5 Image removed due to copyright considerations. Source: Duplessy et al. (1984). b. Complication: at least some species (and perhaps all) (especially Uvigerina) appear to have variable offsets from bottom water composition related to the productivity of the overlying surface ocean. 6 Images removed due to copyright considerations. Source: Zahn et al. (1987). 7 c. Complication: where in the sediment do foraminifera live? If they live at some depth in the sediment, δ13C can be significantly different from that of the bottom water. d. Complication: The average δ13C of the ocean is not constant, due to a significant transfer of terrestrial reduced organic into the oceans during glacial periods (tropical aridity and destruction of high-latitude hardwood forests: Shackleton (1977). This means that a local downcore records of δ13C have a large (and often dominant) component due to changes in the ocean average δ13C, and it is necessary to somehow correct for this in determining the spatial δ13C distribution at a given time. 3. Because of these complications, early δ13C studies of benthic foraminifera have been characterized by "incomplete interpretations" which were modified later. For example, Shackleton (1977) attributed the C-13 changes from Uvigerina off Northwestern Africa (core M12392) (almost) solely to changes in the global oceanic δ13C due to a diminished LGM soil/biospheric carbon reservoir. Image removed due to copyright considerations. Source: Shackleton (1977), in Fate of Fossil Fuel CO2, figure 3. Note that this figure is slightly misleading because the “Uvigerina”record is actually a “Uvigerina + Cibicidoides converted to Uvigerina scale” composite record. It turned out later that a lot of the C13 “signal” in this core was due to the variable C13 offset for Uvigerina in high organic carbon glacial sediments (although the “continental biomass” argument is still accepted as part of the benthic C13 record. Later, Shackleton, Imbrie, and Hall (1983) attributed the difference 8 between Atlantic and Pacific records to NADW fluctuations (which would have therefore been a large portion of the total δ13C record in the Atlantic core. Zahn et al. (1986) showed that Uvigerina shows variable δ13C offsets from bottom water that are related to the organic carbon content of the sediment (which they suggested was due to an infernal habitat for Uvigerina). The Northwestern Africa (core M12392) Uvigerina record contains a large component due to organic-carbon source related non-constant δ13C offsets for Uvigerina. Hence the role of NADW fluctuations was smaller than it appeared in the 1983 paper. Images removed due to copyright considerations. Source: Shackleton, Imbrie, and Hall (1983), figure 2, 3, and 5. 10 4. However, so far it appears that the species. C. wuellerstorfi records the more reliable record of temporal and spatial variations in C-13 [though even this species shows negative δ13C offsets in some environments (e.g. Mackensen, 1993)]. C. Cd/Ca in foraminifera as a tracer of ocean chemistry and circulation. 1. Organisms also take up cadmium (Cd) in their organic tissues which is quantitatively released during decomposition just as C,N, and P. Hence Cd plots linearly with P; both are near-zero in subtropical surface waters. Image removed due to copyright considerations. Source: Boyle (1994). 11 2. Cd has nearly the same ionic radius as Ca; hence it is well-suited to solid-solution substitution in the calcium carbonate crystal. Chemically, at equilibrium, one would expect that XCd ___ XCa = Deq γCd fCd [Cd] ____________ γCa fCa [Ca] where: X = the mole fraction of the element in the solid Deq = the equilibrium constant (distribution coefficient) (determined by the relative free energies of the ion in solution and in the solid) γ = the activity coefficient of the free ion in solution f = the fraction of the element that exists in solution as the free ion (i.e., not complexed) [ ] = the total concentration of the element in solution For an ideal solid solution (i.e., where the free energy of CdCO3 is the same in solid solution in CaCO3 as it is in pure CdCO3), Keq is equal to the ratio of the solubility products of the pure carbonates Ksp(CdCO3)/KspCaCO3. (Most solid solutions are non-ideal, however). 3. Although γ, f, and K are functions of temperature, their temperature dependence is expected to be slight, so that we expect (Cd/Ca)carbonate = Dp (Cd/Ca)solution where Dp is the partition coefficient 4. Determination of the equilibrium partition coefficient is difficult (because of rampant disequilibrium and slow kinetics at low temperatures). And of course, organisms may choose to use their energy resources to create disequilibrium fractionations. Best course is to use empirical approach to see if the cadmium content of the foraminifera is proportional to the cadmium content of the water the organisms grown in. This has been approached by feeding radioactive cadmium to planktonic foraminifera (Delaney, Be) and by looking at the distribution of Cd in core-top benthic foraminifera relative to the bottom-water Cd content (Boyle, 1988). These studies show that the Cd content of a variety of deepwater (>2500m) calcitic foraminifera is proportional to the Cd content of the water, with Dp ≈ 2.9 ± 0.6 . In shallower cores, the situation for Cd is more complicated. Coretop calibration shows that D changes with depth from ~1.3 at z<1100 to D~2.9 at z>2900 m. Boyle (1992) proposed that cadmium in the water ("CdW") should be 12 estimated from this relationship when comparing data from depths crossing this transition. Fortunately, the aragonitic benthic foraminifera Hoeglundina elegans appears to little or no depth dependence for D (Boyle et al., 1995), which allows one to check whether the inferred depth dependence is stable. Comparison of Cd data from this species and the calcitic foraminifera indicates that the calcitic depthdependence has remained constant between the last glacial maximum and the present. Image removed due to copyright considerations. Source: Boyle, 1992. 5. The main difficulty with this technique is that the cadmium content of foraminifera is very low, ranging from less than 0.01 µmol/mol for surface planktonics to 0.23 µmol/mol for Pacific benthics. Boyle (1981) demonstrated that ultrasonicallycleaned foraminifera shells contain in total 100-1000 times more cadmium than is actually contained in the lattice. This cadmium is contained in ferromanganese coatings on the surface of the shells which can be removed by a reductive cleaning treatment. In some environments, precipitation of manganese carbonate (MnCO3) on the surfaces of foraminifera also leads to excessive Cd levels; it is more difficult 13 to selective remove this phase, although acid leaching can preferentially dissolve the outer surfaces relative to the interior calcite. 6. Studies by Boyle and Keigwin (1982; 1984 a,b; 1985) have shown that the δ13C results from C. wuellerstorfi and Cd/Ca in several species including C. wuellerstorfi, C. kullenbergi, Uvigerina, and N. umbonifera are consistent and indicate that the nutrient content of the deep North Atlantic was about twice as high during the glacials as it is during the interglacials, which is consistent with a significant reduction in NADW, but not with cessation. 7. Aragonitic Hoeglundina as a deep water Cd tracer. 8. McCorkle, Lea et al. (1995)showed that C. wuellerstorfi from very deep Pacific cores had lower Cd than expected from bottom water chemistry 9. Marchitto (2004) confirmed that there was no significant temperature effect on benthic foraminiferal Cd in the range 4°C-18°C. D. Paired δ13C and Cd/Ca data as tracers of ocean circulation and chemistry. 1. Even if both tracers contained exactly the same information, the combination of both (each with different own artifacts and problems) can make for a more convincing case when they both agree. 2. Although both δ13C and Cd plot linearly against P (and hence contain similar information on oceanic chemical fractionation due to deep ocean circulation patters), paired data on both tracers provides more information because the slope of paired isochronous data provides information on the Cd:P ratio of the ocean and the intercept provides information on the P content of the ocean. (Boyle, 1985). 3. "δ13Cas" concept (Lynch-Stieglitz et al., 1995): the role of gas exchange in influencing oceanic δ13C. At equilibrium, the d13C of seawater in equilibrium with the atmosphere is a function of the temperature: Temperature-dependent equilibrium: fixing the carbon isotope composition of the atmosphere and allowing seawater to equilibrate: 14 25°C 0°C -6.4 -6.4 gaseous CO2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -7.6 -7.7 aqueous CO2 HCO3- +1.5 +4.4 CO3= -0.5 +0.8 (CaCO3) ______ TCO2 +3.0 +0.0 +1.3 +4.1 Because of this temperature-dependent isotopic fractionation, gas exchange tends to make higher latitude waters heavier in δ13C (and lower latitude waters lighter). This gas exchange process does not affect phosphorus or cadmium, and hence is a potential mechanism for making the oceanic δ13C distribution different from that of P or Cd. Lynch-Stieglitz et al., 1995 proposed that the "preformed" δ13C distribution could be estimated by correcting benthic δ13C for deep respiration using benthic Cd/Ca data and assuming a Cd:C "Redfield Ratio". For this method to work, both tracers must be free of artifacts, and at the present time this cannot be guaranteed. E. Foraminiferal Ba/Ca as a paleoceanographic tracer 1. Ba also has "nutrient-like" behavior in the ocean (lower in surface waters, higher in deep waters), but its oceanic distribution resembles that of Si or Alk more than that of P or N; this distribution is explained as due to the redissolution of the Ba-Si-Alk tracers near or on the ocean floor rather than shallower in the water column as for P, N, Cd. 2. Foraminiferal Ba also has a cleaning problem; main culprit is barite (BaSO4); special cleaning methods need to be employed to remove this selectively from CaCO3. 3. Benthic foraminiferal Ba correlates with bottom water Ba, and the three species studied give similar Ba values; it looks like Ba will be a useful paleo tracer. 4. Early results of Ba studies show some similarities and some differences compared to Cd and δ13C. F. Foraminiferal Zn/Ca (Marchitto et al. 2000, 2002). Zn is incorporated into benthic foraminifera but the amount depends on both the Zn concentration and the carbonate ion concentration of the bottom water. G. Results from deep-water paleochemistry studies 15 1. North Atlantic Deep Water flow was diminished, but did not cease (at least on time scales approachable for 3 cm/kyr sediment cores). Evidence: Cd in several benthic species and δ13C in C. wuellerstorfi show that NADW remains nutrient-depleted relative to Pacific Deep Water (Boyle and Keigwin, 1982; 1985/6). 2. Intermediate-depth water in the North Atlantic ocean was more depleted in nutrients than at present. Evidence: Cd and δ13C in Caribbean Sea benthic forams; continental margin cores (East and Western margins); little change in nutrients at 2100m on Mid-Atlantic Ridge (in contrast to deeper increase) (Boyle and Keigwin, 1987). 3. Curry and Lohmann (1983) study of bathymetric gradients in δ13C in the eastern equatorial Atlantic. 4. Duplessy et al. (1988) and Curry et al. (1988) studies of Atlantic and global deepwater δ13C distributions during the last 18,000 years. 5. Kallel et al. (1988) study of northern Indian ocean shallow cores. Boyle et al. (1992) amplification using Cd in aragonitic and calcitic benthics. 6. Other basins: the jury is still out. Some special interest in the North Pacific, since some people have wondered if deepwater might have formed there during glacial times. II. 14C as a deep-water and ventilation rate tracer A. The basic idea is simple: benthic and planktonic foraminifera incorporate 14C into their shells and hence represent deep and surface water 14C. Since 14C decreases as a water mass "ages", the contrast between surface and deep waters can be used as a circulation tracer. Coexisting pairs of benthic and planktonic foraminifera from sediment cores allows us to do this in the past, in principle. B. In practice, there are some major obstacles to this idea. The most serious are mixing, statistics, and dissolution. Because a range of ages is represented by specimens, one must collect enough to compensate for the sampling statistics; this problem is more serious than for chemical tracers (because 14C always changes by a factor of two every 5730 years, while Cd or δ18O may not). Furthermore, changes in P/B abundance patterns and differential dissolution can create artifacts. We also need to understand something more about the behavior of 14C on the surface of the earth. But working with deep sea corals (no bioturbation) or planktonic-benthic pairs from abundance maxima (hence minimizing the effects of bioturbation, this method leads to some very interesting results (Keigwin and Schlegel, 2002; Robinson et al., 2005). 16 Image removed due to copyright considerations. Source: Duplessy et al. (1988). 17 Figure removed due to copyright considerations. Please see: Figure 8 in Keigwin and Schlegel. “Ventilation in the Western North Atlantic.” Paleoc 19 (2004): 4012. Figure removed due to copyright considerations. Please see: Figure 3 in Robinson L. F., J. F. Adkins, L. D. Keigwin, J. Southon, D. P. Fernandez, S. L. Wang, and D. S. Scheirer. “Variability in the Western North Atlantic During the Last Deglaciation.” Science 310 (2005): 1469-1473. 18 III. 231Pa/230Th as a tracer of deep ocean ventilation rates (Yu et al.,1996; Marchal et al.,2000; McManus et al. 2004). These isotopes are generated from the decay of 235U and 234U at a precisely known rate that is uniform throughout the ocean. But 230Th is more "particle reactive than 231Pa (deep sea Th residence time is ~30 years whereas the Pa residence time is ~150 years). Hence ocean currents such as NADW can transport 231Pa away from their origin to regions where the particle flux (and scavenging) is higher. The sediments of the modern North Atlantic are depleted in 231Pa/230Th as a result of this deep water transport of 231Pa out of the basin. Yu et al. (1996) found that the LGM Atlantic was similarly depleted, hence they argued that there must have been a net flow of NADW out of the basin comparable to the modern rate. Subsequent model studies (Marchal et al., 2000) have shown that the constraint is probably a +50% estimate (i.e., NADW export from the Atlantic could have been as low as 50% of the present value, or 50% higher). Figure removed due to copyright considerations. Please see: Yu, E. -F., R. Francois, and M. P. Bacon. “Similar rates of modern and last-glacial ocean thermohaline circulation inferred from radiochemical data.” Nature 379 (1996): 689-694. 19 Figure removed due to copyright considerations. Please see: McManus J. F., R. Francois, J. –M. Gherardi, L. D. Keigwin, and Leger S. Brown. “Collapse and rapid resumption of Atlantic meridional circulation linked to deglacial climate changes.” Nature 428 (2004): 834-837. Readings and references Boyle, E.A. (1981) Cadmium, zinc, copper, and barium in foraminifera tests, E.A. Boyle, EPSL, 53, 11-35. Boyle, E.A., and L.D. Keigwin (1982) Deep circulation of the North Atlantic over the last 200,000 years: geochemical evidence, Science, 218:784-787. E.A. Boyle and L.D. Keigwin (1985/1986) Comparison of Atlantic and Pacific paleochemical records for the last 215,000 years: changes in deep ocean circulation and chemical inventories, EPSL 76, 135-150. Boyle, E.A. (1986) Paired carbon isotope and cadmium data from benthic foraminifera: implications for changes in deep ocean P and atmospheric carbon dioxide Geochim. Cosmochim. Acta 50: 265-276. Boyle, E.A. (1988) Cadmium: chemical tracer of deep-water paleoceanography. Paleoceanography, 3:471-489. * Boyle, E. and Keigwin, L.D., 1987. North Atlantic thermohaline circulation during the last 20,000 years linked to high latitude surface temperature. Nature, 330:35-40. Boyle, E.A., L. Labeyrie, and J.-C. Duplessy, Calcitic foraminiferal data confirmed by cadmium in aragonitic Hoeglundina: application to the last glacial maximum northern Indian Ocean, Paleoceanogr., 10, 881-900, 1995. Curry, W. and G.P. Lohmann, Reduced advection into Atlantic Ocean deep eastern basins during last glaciation maximum, Nature, 306, 577-580, 1983. * Curry, W.B., Duplessy, J.C., Labeyrie, L.D. and Shackleton, N.J., 1988. Changes in the distribution of C13 of deep water CO2 between the last glaciation and the Holocene. Paleoceanogr., 3:317-342. Duplessy, J.C., R.K. Matthews, W. Prell, W.F. Ruddiman, M. Caralp, and C.H. Hendy, C13 record of benthic foraminifera in the last interglacial ocean: implications for the carbon cycle and global deep water circulation, Quat. Res., 21, 225-243, 1984. Duplessy, J.C., Shackleton, N.J., Fairbanks, R.G., Labeyrie, L., Oppo, D. and Kallel (1988) Deepwater source variations during the last climatic cycle and their impact on the global deepwater circulation. Paleoceanography 3: 343-360 Emrich et al. (1970) Earth Planet. Sci. Lett. 8, 363-371 - but note they excluded their 20°C measurement as unreliable. 20 Keigwin, L.D. and E.A. Boyle (1985) Carbon isotopes in deep-sea benthic foraminifera: precession and changes in low-latitude biomass, , in The carbon cycle and atmospheric CO2: Natural Variations Archean to Present, E. Sundquist and W. Broecker, eds., 319328, AGU. Keigwin L. D. and Schlegel M. A. (2002) Ocean ventilation and sedimentation since the glacial maximum at 3 km in the western North Atlantic. Geochem. Geophys. Geosyst. 3, 1034, 10.1029/2001GC000283. Kallel, N., Labeyrie, L.D., Juillet-Laclerc, A. and Duplessy, J.C., 1988. A deep hydrological front between intermediate and deep water masses in the glacial Indian Ocean. Nature, 333:651-655. Loubere P. (1994) Quantitative estimation of surface ocean productivity and bottom water oxygen concentration using benthic foraminifera. Paleoceanogr. 9, 723-737. Lynch-Stieglitz, J., and R.G. Fairbanks, A conservative tracer for glacial ocean circulation from carbon isotope and paleo-nutrient measurements in benthic foraminifera, Nature, 369, 308-310, 1994. Mackensen A. (1993) δ13C in benthic foraminiferal tests of Fontbotia wuellerstorfi (Schwager) relative to the d13C of dissolved inorganic carbon in Southern Ocean deep water: implications for glacial ocean circulation models. Paleoceanogr. 6, 587-610. Marchal O., Francois R., Stocker T. F., and Joos F. (2000) Ocean thermohaline circulation and sedimentary 231Pa/230Th ratio. Paleoceanogr. 15, 625-641. Marchitto T. M. (2004) Lack of a significant temperature influence on the incorporation of Cd into benthic foraminiferal tests. Geochem. Geophys. Geosys. 5, Q10D11, doi:10.1029/2004GC000753. Marchitto T. M., Oppo D. W., and Curry W. B. (2002) Paired benthic foraminiferal Cd/Ca and Zn/Ca evidence for a greatly increased presence of Southern Ocean Water in the glacial North Atlantic Matsumoto, K. and J. Lynch-Stieglitz (1999) Similar glacial and Holocene deep water circulation inferred from southeast Pacific benthic foraminiferal carbon isotope composition, Paleoceanogr. 14:149-163 McManus J. F., Francois R., Gherardi J.-M., Keigwin L. D., and Brown-Leger S. (2004) Collapse and rapid resumption of Atlantic meridional circulation linked to deglacial climate changes. Nature 428, 834-837. Robinson L. F., Adkins J. F., Keigwin L. D., Southon J., Fernandez D. P., Wang S.-L., and Scheirer D. S. (2005) Variability in the Western North Atlantic During the Last Deglaciation, Science 310, 1469-1473. Shackleton, N.J. (1977) Carbon-13 in Uvigerina: Tropical rainforest history and the equatorial Pacific carbonate dissolution cycles, , in The Fate of Fossil Fuel CO2, edited by N.R. Andersen and A. Malahoff, 401-427, Plenum, N.Y. 21 Shackleton, N.J., J. Imbrie and M.A. Hall (1983) Oxygen and carbon isotope record of East Pacific core V19-30: implications for the formation of deep water in the late Pleistocene North Atlantic, EPSL 65, 233-244. Streeter, S.S. and N.J. Shackleton (1979) Paleocirculation of the deep North Atlantic: 150,000-year record of benthic foraminifera and oxygen-18, , Science, 203, 168-171. Yu, E.-F., R. Francois, and M.P. Bacon, Similar rates of modern and last-glacial ocean thermohaline circulation inferred from radiochemical data, Nature, 379, 689-694, 1996. Zahn, R., Winn, K. and Sarnthein, M., 1986. Benthic foraminiferal 13C and accumulation rates of organic carbon: Uvigerina peregrina group and Cibicidoides wuellerstorfi. paleoceanography, 1:27-42. Zhang, J., P.D. Quay, and D.O. Wilbur (1995) Carbon isotope fractionation during gaswater exchange and dissolution of CO2, Geochim. Cosmochim. Acta, 59:107-114.