Implications of a Carbonate Ion Effect on Shell Carbon and Oxygen Isotopes
515
of culturing experiments, this increase in sea surface carbonate ion is expected to influence the isotopic composition of planktic foraminifera.
We have calculated surface water carbonate
ion anomalies at temperatures of lOOC and 25°C
for two hypotheses put forth to explain the glacial
decrease in pC0 2 : a stronger biological pump, and
a rise in whole ocean pH driven by changes in
CaC0 3 cycling (Berger 1982; Broecker 1982b;
Archer and Maier-Reimer 1994; Sanyal et al. 1995).
The change in surface ocean carbonate ion is very
different in these two scenarios because the ratio
of carbonate ion increase to pC0 2 decrease depends on the surface ocean alkalinity: CO 2 change
(Fig. 1). A smaller change is associated with the
drawdown of dissolved carbon dioxide under conditions of unchanging alkalinity, as would be the case
if the biological pump were strengthened solely
through an increase in the Redfield ratio of
phytoplankton, without any calcite compensation
(Broecker 1982b; Broecker and Henderson 1998).
In this case, a40 (at lOOC) to SO (at2S°C) Ilmollkg
surface water carbonate ion increase equates to a
pC0 2 of 200 ppmV. However, if the same pC0 2
level is achieved by the addition of calcium carbonate to seawater, as in the coral reef hypothesis,
carbonate ion increases by 80 (at lO°C) to 110 (at
25°C) Ilmollkg (Fig. 1 a, b). The change in carbonate ion required to achieve a specific pC0 2 levei is
smaller in cold waters because they contain less
carbonate ion and therefore require smaller fractional changes to match the glacial pcq reduction.
The calculated changes in surface water carbonate ion for the two scenarios are converted to
8 J3 C anomalies using the experimental response for
G. bulloides, a sub-polar species, for the 10°C
calculation and the response for 0. universa, a
sub-tropical species, for the 25°C calculation (Fig.
lc, d). Predicted anomalies in she1l8 13 C due to increased glacial carbonate ion range between -0.3
and -0.9%0, with the largest anomalies predicted for
G. bulloides under the conditions of CaC0 3 addition. In general, the predicted anomalies in shell
8 J3 C for the CaC0 3 addition scenario are twice
that ofthe anomalies predicted for the CO 2 extraction scenario. It should be noted that the contrast
in seawater pH increase for the two scenarios is
smaller: 0.11 pH units for the CO 2 drawdown scenario versus 0.15 for the CaC0 3 addition scenario.
Therefore, if pH controls the isotopic deviations in
the foraminifer shells, we would predict a less dramatic 8 J3 C anomaly difference for the two scenarios.
We have also generated general circulation
model simulations of surface water carbonate ion
concentrations for the interglacial ocean (Fig. 2a)
and two glacial simulations with lowered atmospheric pC0 2 (Fig. 2b, c) (Archer et al. in review).
The first glacial simulation increases oceanic NO;
by 50% (Fig. 2b) and assumes that plankton productivity increases in proportion to the nitrate increase (Broecker and Henderson 1998). There is
no CaC0 3 compensation in this simulation, which
is analogous to the CO 2 extraction scenario discussed above. The second glacial simulation invokes
a doubling of ocean H 4 Si0 4 (Fig. 2c), which perturbs calcite burial in such a way that oceanic alkalinity increases through CaC0 3 compensation
(Archer and Maier-Reimer 1994). This simulation
is analogous to the CaC0 3 addition scenario discussed above. As predicted from the calculations
above, the change in surface water carbonate ion
generated by the doubling of oceanic H 4 Si0 4
(CaC0 3 addition) is much larger than for a 50%
nitrate increase (C0 2 extraction). In line with these
simulations, the predicted negative anomalies in
foraminifer she1l8 13 C due to the carbonate ion effect would be consequently greater for pC0 2 reduction achieved via ocean alkalinity increase as
opposed to a stronger biological pump.
Predicted 1) 13 C Anomalies Compared to
Downcore Records
High resolution isotopic data from core RC 11-83 in
the sub-polar South Atlantic provides a useful basis of comparison with the 8 J3 C anomalies predicted for G. bulloides (Charles et al. 1996). Carbon isotope anomalies due to the influence of changing surface water carbonate ion concentration are
computed for the last 200 kyrs based on the Vostok
pC0 2 record and the assumption that surface waters were in equilibrium with the atmosphere (Fig.
3). Our calculations demonstrate that8 13 C anomalies, when computed for the condition of CaC0 3
addition (i.e. maximum carbonate anomaly), give
515
of culturing experiments, this increase in sea surface carbonate ion is expected to influence the isotopic composition of planktic foraminifera.
We have calculated surface water carbonate
ion anomalies at temperatures of lOOC and 25°C
for two hypotheses put forth to explain the glacial
decrease in pC0 2 : a stronger biological pump, and
a rise in whole ocean pH driven by changes in
CaC0 3 cycling (Berger 1982; Broecker 1982b;
Archer and Maier-Reimer 1994; Sanyal et al. 1995).
The change in surface ocean carbonate ion is very
different in these two scenarios because the ratio
of carbonate ion increase to pC0 2 decrease depends on the surface ocean alkalinity: CO 2 change
(Fig. 1). A smaller change is associated with the
drawdown of dissolved carbon dioxide under conditions of unchanging alkalinity, as would be the case
if the biological pump were strengthened solely
through an increase in the Redfield ratio of
phytoplankton, without any calcite compensation
(Broecker 1982b; Broecker and Henderson 1998).
In this case, a40 (at lOOC) to SO (at2S°C) Ilmollkg
surface water carbonate ion increase equates to a
pC0 2 of 200 ppmV. However, if the same pC0 2
level is achieved by the addition of calcium carbonate to seawater, as in the coral reef hypothesis,
carbonate ion increases by 80 (at lO°C) to 110 (at
25°C) Ilmollkg (Fig. 1 a, b). The change in carbonate ion required to achieve a specific pC0 2 levei is
smaller in cold waters because they contain less
carbonate ion and therefore require smaller fractional changes to match the glacial pcq reduction.
The calculated changes in surface water carbonate ion for the two scenarios are converted to
8 J3 C anomalies using the experimental response for
G. bulloides, a sub-polar species, for the 10°C
calculation and the response for 0. universa, a
sub-tropical species, for the 25°C calculation (Fig.
lc, d). Predicted anomalies in she1l8 13 C due to increased glacial carbonate ion range between -0.3
and -0.9%0, with the largest anomalies predicted for
G. bulloides under the conditions of CaC0 3 addition. In general, the predicted anomalies in shell
8 J3 C for the CaC0 3 addition scenario are twice
that ofthe anomalies predicted for the CO 2 extraction scenario. It should be noted that the contrast
in seawater pH increase for the two scenarios is
smaller: 0.11 pH units for the CO 2 drawdown scenario versus 0.15 for the CaC0 3 addition scenario.
Therefore, if pH controls the isotopic deviations in
the foraminifer shells, we would predict a less dramatic 8 J3 C anomaly difference for the two scenarios.
We have also generated general circulation
model simulations of surface water carbonate ion
concentrations for the interglacial ocean (Fig. 2a)
and two glacial simulations with lowered atmospheric pC0 2 (Fig. 2b, c) (Archer et al. in review).
The first glacial simulation increases oceanic NO;
by 50% (Fig. 2b) and assumes that plankton productivity increases in proportion to the nitrate increase (Broecker and Henderson 1998). There is
no CaC0 3 compensation in this simulation, which
is analogous to the CO 2 extraction scenario discussed above. The second glacial simulation invokes
a doubling of ocean H 4 Si0 4 (Fig. 2c), which perturbs calcite burial in such a way that oceanic alkalinity increases through CaC0 3 compensation
(Archer and Maier-Reimer 1994). This simulation
is analogous to the CaC0 3 addition scenario discussed above. As predicted from the calculations
above, the change in surface water carbonate ion
generated by the doubling of oceanic H 4 Si0 4
(CaC0 3 addition) is much larger than for a 50%
nitrate increase (C0 2 extraction). In line with these
simulations, the predicted negative anomalies in
foraminifer she1l8 13 C due to the carbonate ion effect would be consequently greater for pC0 2 reduction achieved via ocean alkalinity increase as
opposed to a stronger biological pump.
Predicted 1) 13 C Anomalies Compared to
Downcore Records
High resolution isotopic data from core RC 11-83 in
the sub-polar South Atlantic provides a useful basis of comparison with the 8 J3 C anomalies predicted for G. bulloides (Charles et al. 1996). Carbon isotope anomalies due to the influence of changing surface water carbonate ion concentration are
computed for the last 200 kyrs based on the Vostok
pC0 2 record and the assumption that surface waters were in equilibrium with the atmosphere (Fig.
3). Our calculations demonstrate that8 13 C anomalies, when computed for the condition of CaC0 3
addition (i.e. maximum carbonate anomaly), give