Oxygen Isotope Studies Of Calcareous Marine Fauna

If calcium carbonate is crystallized slowly in water, lsO is slightly concentrated in the calcium carbonate relative to that in the water. The process is temperature-dependent, with the concentrating effect diminishing as temperature increases. In a nutshell, this is the basis for a very important branch of paleoclimatic research — the analysis of oxygen isotopes in the calcareous tests of marine microfauna (princi-

pally foraminifera, but also coccoliths). The approach was first enunciated by Urey (1947, 1948) who noted, "If an animal deposits calcium carbonate in equilibrium with the water in which it lies, and the shell sinks to the bottom of the sea . . . it is only necessary to determine the ratio of the isotopes of oxygen in the shell today in order to know the temperature at which the animal lived" (Urey, 1948).

He then went on to calculate, from thermodynamic principles, the magnitude of this temperature-dependent isotopic fractionation. Although the principle of Urey's argument is correct, the numerous complications that arise in the real world have made direct paleotemperature estimates rather problematic. In fact, the oxygen isotope record in marine sediments varies locally (with temperature, and to a lesser extent salinity) and globally with variations in continental ice volume. This global signal is the single most important record of past climatic variations for the entire Cenozoic.

6.3.1 Oxygen Isotopic Composition of the Oceans

The oxygen isotopic composition of a sample is generally expressed as a departure of the 180/160 ratio from an arbitrary standard15:

(180/160)standard X 10 {6A)

The resulting values are expressed in per mil (%o) units; negative values represent lower ratios in the sample (i.e., less 8lsO than 160 and, therefore, isotopically lighter) and positive values represent higher ratios in the sample (more 8lsO than 160 and, therefore, isotopically heavier).

Empirical studies relating the isotopic composition of calcium carbonate deposited by marine organisms to the temperature at the time of deposition have demonstrated a relationship that approximates the following16:

where T is water temperature in degrees Celsius, 8c is the per mil difference between the sample carbonate and the SMOW standard, and 8^ is the per mil difference between the 8lsO of water in which the sample was precipitated and the SMOW standard (Epstein et ai, 1953; Craig, 1965).

For modern samples, 8^ can be measured directly in oceanic water samples; in fossil samples, however, the isotopic composition of the water is unknown and cannot be assumed to have been as it is at the site today. In particular, during glacial periods the removal of isotopically light water from the oceans to form continental ice sheets (Section 5.2) led to an increase in the 180/160 ratio of the oceans as a whole by -1.1 ± 0.25%o. Thus, the expected increase in 8c of foraminiferal tests during glacial

15 See footnote 11 on page 131, Chapter 5; Isotopic studies based on carbonate fossils use as a standard a Cretaceous belemnite from the PeeDee Formation of North Carolina (PDB-1) or a cross-referenced U.S. NIST sample. Carbon dioxide released from PDB-1 = +0.2%o relative to C02 equilibrated with SMOW (Craig, 1961b).

16 The precise form of the relationship depends on the particular technique used in analysis and on the temperature at which fractionation occurs (for further discussion, see Shackleton, 1974; Mix, 1987).

periods due to lower water temperatures is complicated by the increase in 8^ of the ocean water at these times. How much of the increase in &c is the result of variations in 8^ can be assessed by analyzing 8180 in pore waters squeezed from sediments of the last glacial age (Schrag and DePaolo, 1993) or in the tests of benthic (bottom-dwelling) foraminifera. Bottom waters today (derived from cold, dense, polar water spreading through the deep ocean basins) are relatively close to the freezing point of seawater, so the bulk of the 8lsO increase in benthic forams in glacial times can not be due to significantly lower temperatures; rather, the evidence indicates that at least 70% of the increase results from the changing isotopic composition of the oceans (Duplessy, 1978; Mix, 1987). Hence, the isotopic changes recorded in benthic fora-miniferal tests are primarily a record of changing terrestrial ice volumes, or a "paleo-glaciation" record (Shackleton, 1967; Dansgaard and Tauber, 1969). The benthic isotope record thus demonstrates that there have been more than 20 periods of major continental glaciation during the Quaternary, with the largest changes in ice volume (from glacials to interglacials) within the last 900,000 yr (Shackleton et al., 1990). The record of changing 8lsO in relation to variations in continental ice volume (recorded in terms of eustatic sea level change) is discussed further in Section 6.3.4.

Changes in the isotopic composition of ocean water through time are not the only complications affecting a simple temperature interpretation of 8c (Mix, 1987). Urey's initial hypothesis developed from a consideration of calcium carbonate precipitated inorganically, where the carbonate forms in isotopic equilibrium with the water. However, in the formation of carbonate tests by living organisms, metaboli-cally produced carbon dioxide may be incorporated; in such cases, the carbonate would not be formed in isotopic equilibrium with the water and the resulting isotopic composition would differ from the thermodynamically predicted value, generally leading to 8180 (and 813C) lower than the expected equilibrium values (Duplessy et al., 1970a; Vinot-Bertouille and Duplessy, 1973; Shackleton et al., 1973). This was termed the vital effect by Urey (1947). The contribution of metabolic carbon dioxide to the test carbonate differs from one species of foraminifera to another (Grossman, 1987). Modern samples of Globigerinoides ruber, for example, give isotopic values 0.5%o lighter than expected from thermodynamic principles alone (based on analysis of water from their modern habitat). This is equivalent to a temperature error of -2.5 °C (Shackleton et al., 1973). On the other hand, not all forams exhibit this unfortunate characteristic. For example, samples of Pulleni-atina obliquiloculata and Uvigerina spp. (benthic forams) appear to be in isotopic equilibrium with surrounding water (Shackleton, 1974). In other species, where isotopic equilibrium is not achieved, there is evidence that the vital effect remains constant over time (Duplessy et al., 1970a). It is thus possible to circumvent this particular problem by careful selection of the species being studied, or by assessing its specific vital effect, and adjusting the measured isotopic values accordingly.

Another complication in calculating water temperatures from the isotopic composition of carbonate tests is the problem of variations in depth habitat of planktic foraminifera. Even if the ice effect and vital effects are known, there is still some uncertainty as to whether foraminifera lived at the same depth from glacial to interglacial times. Water temperatures in the upper few hundred meters of the ocean change rapidly with depth, particularly outside the Tropics (Table 6.1), so small

TABLE 6.1 Mean Vertical Temperature Distribution (°C) and Temperature Gradients in the Three Oceans Between 40°N and 40°S

Atlantic Ocean Indian Ocean Pacific Ocean

Temperature Gradient Temperature Gradient Temperature Gradient Depth (m) (°C) ("C/IOOm) (°C) (°C/I00m) (°C) (°C/I00m)

Mean

1000

1200

1600

2000

3000

4000

20.0

17.8

13.4

0.35

0.20

0.15

0.12

0.08

0.08

22.2

18.9

14.3

11.0

0.15

0.09

0.03

21.8

18.7

14.3

2.15

1.45

0.65

0.05

0.03

21.3

18.5

14.0

10.0

0.75

0.35

0.22

0.12

0.07

0.05

From Defant (1961). " Maximum gradient.

variations in depth habitat can be equivalent to a change in temperature of several degrees Celsius (i.e., a change perhaps as large as the glacial to interglacial change at the surface of the ocean). It is thus critical to know what factors control depth habitat of foraminifera, and in particular the depths at which tests are secreted (Emiliani, 1971). Several studies have concluded that water density (a function of temperature and salinity; Fig. 6.7) is of prime importance to individual species, as the same species may be found in different areas living at different depths, but in water of the same temperature and salinity (Emiliani, 1954, 1969; Hecht and Savin, 1972). During glacial periods, when the oceans were more saline (due to the removal of water to the continental ice sheets), foraminifera may have migrated upwards in the water column, to a zone of warmer water, in order to maintain a constant density environment. Conversely, they, may have migrated downwards (to cooler water) in interglacials (Fig. 6.7). Clearly, such vertical migrations would result in isotopic paleotemperature estimates of glacial to interglacial temperature differences, considerably less than the changes actually occurring in the water column (Savin and Stehli, 1974). Hence, if this model is correct, any residual paleotemperature signal obtained (after correcting for ice and vital effects) would have to be considered a minimum estimate only.17 This problem may, however, be a relatively

FIGURE 6.7 Temperature-salinity diagram: a = 103 (pw - I), where pw = water density. Density of seawa-ter is a function of both temperature and salinity. Lines of equal density are shown. During glacial periods, when the removal of water to the continental ice sheets would have made overall oceanic salinity higher, foraminifera may have migrated to warmer water (generally upwards in the water column) to maintain a constant-density environment (illustrated schematically as A - B - C). In an interglacial, the opposite situation may have prevailed and the response of foraminifera may have been to move downward in the water column to a cooler zone below (A - D - E).

17 In addition, the effects of sediment mixing (bioturbation), which tend to smooth out extremes in the record, also contribute to making actual glacial-interglacial 8lsO differences appear smaller (Shack-leton and Opdyke, 1976).

33 34 35 36 37 SALINITY, %o

minor one compared to the important effect of variations in depth habitat of foraminifera during their life cycle. There is now convincing evidence that although the tests of living forams contain CaC03 that has been secreted in isotopic equilibrium with the upper mixed water layer, in certain species foram tests from the sea floor are significantly enriched with 180 compared to their living counterparts (Du-plessy et al., 1981). This is apparently due to calcification of the tests at depths (>300 m) considerably below the upper mixed layer, during the process of gameto-genesis (reproduction). Gametogenic calcification may account for -20% of foram test weight in samples from the sea floor and, because calcium carbonate has been extracted from water that is much cooler than that nearer the surface, the overall 8lsO values indicate a mean temperature significantly lower than the near-surface temperature (Fig. 6.8). Obviously, the rate at which the organism descends through the water column and the relative extent of gametogenic calcification will greatly influence the final isotopic composition of the test calcite. Similarly, as certain species are distinctly seasonal, the water temperature at those times will be reflected in 8lsO. In order to use 8lgO in foram tests as a temperature indicator, it is necessary to establish in some way exactly which temperatures (by depth and season) are being recorded. To determine empirically the optimum relationship, one approach

CL UJ

0 0

Post a comment