Stable Carbon Isotopes in Benthic Foraminifera
237
the Antarctic contionental margin are also significantly enriched by more than 0.2 %0 relative to bottom water 813C~C02 values. The 813C~C02 values
were interpolated between stations measured on
the same section where the foraminifera samples
were taken (Mackensen et al. 1996). In any event,
a plot of the deviations from bottom water813C~C02
versus latitude clearly reveals the coincidence of
extreme positive deviations (i.e., 13C depleted
foraminifera relative to bottom water 813C~C02)
with the seasonally highly productive areas between
the Subtropical Front (STF) and the Subantarctic
Front (SAF), between the Polar Front (PF) and the
southern boundary of the Antarctic Circumpolar
Current (ACC), and close to the Antarctic continental margin (CM) in the areas of seasonal and
year round coastal polynyas (Fig. 5).
Ifthe deviation from the bottom water813C~C02
is plotted versus water depth, no correlation is obvious, instead, by far the most of the foraminiferal
data show a high negative deviation, (i.e., the difference between water 613C~C02 and foraminiferal
8 13 C is positive) or are within a ± 0.2 %0 range
which is widely accepted as a practical error range
in paleoceanographic interpretations (Fig. 6).
There are, however, some samples from about
70 0 S (Fig. 5) and water depths shallower than 1000
m (Fig. 6) that show a positive deviation, i.e. the
difference between water 813C~C02 and
foraminiferal8 13 C is negative. A closer look to the
raw data reveals that these values are obtained
from Lobatula lobatula (Walker & Jacob) from
the Antarctic continental margin. Since our data and
others (e.g. Hald and Vorren 1987) have shown
that this shelf and upper slope species, similar to
the closely related F. wuellerstorfi and other
cibicids and cibicidoids, usually faithfully records
bottom water 8I3C~C02' we regard its high values
as correctly reflecting saisonally occuring cold,
saline and I3C-enriched water masses, formated in
coastal polynyas by permanent sea ice freezing,
where thermodynamic fractionation between atmosphere and surface water 8 13 C is enhanced (see
above). This, however, does not explain why
8I3C~C02 values of bottom water at the time of
sampling, although high, were not as high as the
foraminiferal data indicate. It may be that the time
of foraminiferal test secretion is coupled to times
of increased thermodynamic fractionation via seasonal openings of polynyas and sea ice formation
within them.
Actually, new data from the Arctic Ocean
reveal a consistent positive deviation of
F. wuellerstorfi from ambient bottom water
8I3Cl:C02 (Mackensen 1997). Although these data
are preliminary, an interpretation is suggested that
links high benthic foraminiferal 8 13 C values with
brine release and sea ice formation in contact with
atmospheric CO 2 ,
Comparison ofthe original data of Duplessy et
al. (1984) used to calibrate the 8 13 C foraminiferal
proxy with our new data from the Arctic Ocean,
the South Atlantic Ocean and the Atlantic sector
of the Antarctic Ocean reveals that it is the low
Pacific 8 13 C data ofDuplessy et al. that makes the
correlation a one-to-one one (Fig. 7). There obviously is a strong negative deviation in the Southern
Ocean and a positive one in the Arctic Ocean data.
Omission ofthe Pacific data would result in a much
steeper relationship between foraminiferal and
water 8 13 C values. The slope ofthe regression line
finally depends on the geographical position of the
samples included, which means if applied to
paleoceanographic data, a use of different calibration factors for northern and southern paleonutrient
calculations. Since the positive deviation can be
avoided by not sampling high latitide continental
margin sediments, and overall this deviation is not
as significant as the negative one in the Southern
Ocean, we suggest for paleoceanographic work,
to carefully check whether during the time span
under study the paleoposition investigated was affected by seasonally high primary production with
possible periodic phytodetritus deposition on the sea
floor.
Paleoceanographic Applications
To illustrate the impact of the new data and findings from the Recent South Atlantic Ocean and the
adjacent Atlantic sector ofthe Antarctic Ocean on
the paleoceanographic interpretation of glacial deep
and bottom water circulation, in the following we
present two examples extracted and compiled from
237
the Antarctic contionental margin are also significantly enriched by more than 0.2 %0 relative to bottom water 813C~C02 values. The 813C~C02 values
were interpolated between stations measured on
the same section where the foraminifera samples
were taken (Mackensen et al. 1996). In any event,
a plot of the deviations from bottom water813C~C02
versus latitude clearly reveals the coincidence of
extreme positive deviations (i.e., 13C depleted
foraminifera relative to bottom water 813C~C02)
with the seasonally highly productive areas between
the Subtropical Front (STF) and the Subantarctic
Front (SAF), between the Polar Front (PF) and the
southern boundary of the Antarctic Circumpolar
Current (ACC), and close to the Antarctic continental margin (CM) in the areas of seasonal and
year round coastal polynyas (Fig. 5).
Ifthe deviation from the bottom water813C~C02
is plotted versus water depth, no correlation is obvious, instead, by far the most of the foraminiferal
data show a high negative deviation, (i.e., the difference between water 613C~C02 and foraminiferal
8 13 C is positive) or are within a ± 0.2 %0 range
which is widely accepted as a practical error range
in paleoceanographic interpretations (Fig. 6).
There are, however, some samples from about
70 0 S (Fig. 5) and water depths shallower than 1000
m (Fig. 6) that show a positive deviation, i.e. the
difference between water 813C~C02 and
foraminiferal8 13 C is negative. A closer look to the
raw data reveals that these values are obtained
from Lobatula lobatula (Walker & Jacob) from
the Antarctic continental margin. Since our data and
others (e.g. Hald and Vorren 1987) have shown
that this shelf and upper slope species, similar to
the closely related F. wuellerstorfi and other
cibicids and cibicidoids, usually faithfully records
bottom water 8I3C~C02' we regard its high values
as correctly reflecting saisonally occuring cold,
saline and I3C-enriched water masses, formated in
coastal polynyas by permanent sea ice freezing,
where thermodynamic fractionation between atmosphere and surface water 8 13 C is enhanced (see
above). This, however, does not explain why
8I3C~C02 values of bottom water at the time of
sampling, although high, were not as high as the
foraminiferal data indicate. It may be that the time
of foraminiferal test secretion is coupled to times
of increased thermodynamic fractionation via seasonal openings of polynyas and sea ice formation
within them.
Actually, new data from the Arctic Ocean
reveal a consistent positive deviation of
F. wuellerstorfi from ambient bottom water
8I3Cl:C02 (Mackensen 1997). Although these data
are preliminary, an interpretation is suggested that
links high benthic foraminiferal 8 13 C values with
brine release and sea ice formation in contact with
atmospheric CO 2 ,
Comparison ofthe original data of Duplessy et
al. (1984) used to calibrate the 8 13 C foraminiferal
proxy with our new data from the Arctic Ocean,
the South Atlantic Ocean and the Atlantic sector
of the Antarctic Ocean reveals that it is the low
Pacific 8 13 C data ofDuplessy et al. that makes the
correlation a one-to-one one (Fig. 7). There obviously is a strong negative deviation in the Southern
Ocean and a positive one in the Arctic Ocean data.
Omission ofthe Pacific data would result in a much
steeper relationship between foraminiferal and
water 8 13 C values. The slope ofthe regression line
finally depends on the geographical position of the
samples included, which means if applied to
paleoceanographic data, a use of different calibration factors for northern and southern paleonutrient
calculations. Since the positive deviation can be
avoided by not sampling high latitide continental
margin sediments, and overall this deviation is not
as significant as the negative one in the Southern
Ocean, we suggest for paleoceanographic work,
to carefully check whether during the time span
under study the paleoposition investigated was affected by seasonally high primary production with
possible periodic phytodetritus deposition on the sea
floor.
Paleoceanographic Applications
To illustrate the impact of the new data and findings from the Recent South Atlantic Ocean and the
adjacent Atlantic sector ofthe Antarctic Ocean on
the paleoceanographic interpretation of glacial deep
and bottom water circulation, in the following we
present two examples extracted and compiled from