The South Atlantic Carbon Isotope Record of Planktic Foraminifera
439
increase in nutrient levels might result from any
combination of (1) decreased biological activity, (2)
increased upwelling rates or (3) increased nutrient
concentrations of source waters. Charles and
Fairbanks (1990) proposed that the increased nutrient concentrations in the Southern Ocean during the LGM were due to a reduced contribution
by nutrient poor NADW. This hypothesis is supported by by the isotopic composition of benthic
foraminifera, which record a comparable high
glacial-intergiacial8 1l C shift of about 1.3 %0 in the
Southern Ocean (see compilation and references
in Michel et al. 1995). Cd/Ca and BafCa ratios of
benthic foraminifera, however, suggest that the
nutrient concentrations within CPDW remained
close to the modern level (Lea 1995; Boyle and
Rosenthal 1996). Additionally, the nutrient utilization was increased south of the APF and remained
unchanged north of the APF (Francois et al. 1997).
This should result, at least south of the APF, in
higher 8 ll C-values of surface waters.
The discrepancy between the data might be
resolved by considering the habitat of N.
pachyderma and the hydrologic changes in the
Southern Ocean during the LGM. Results from
plankton tows indicate that N. pachyderma secretes % of its shell in water depth between 50 and
200 m (Kohfeld et al. 1997; Carstens et al. 1997).
Duplessy et al. (1996) have shown that the salinity
of surface waters in the Antarctic Zone were considerably reduced during the LGM which probably
resulted in an increased vertical stratification. Since
the modem upper water column is relatively well
mixed in the Southern Ocean, N. pachyderma
records surface water conditions despite its deeper
habitat. Under more stratified conditions, however,
their habitat would be enriched in nutrients and
hence their shells would record a lower8 ll C. Support for this hypothesis comes from shallow dwelling
G. bulloides which records a much lower glaciall
interglacial amplitude in the subantarctic zone
(Ninnemann and Charles 1997).
Due to the higher stratification or sea ice cover,
the gas exchange between atmosphere and ocean
might have been considerably reduced during the
glacial. This would cause an additional decrease of
the 8 1l C of surface waters and probably also in the
water masses sinking down from Southern Ocean
surface waters. Fig. lc indicates that nowadays the
exchange with the atmosphere increases the 8 1l C
of surface waters in the vicinity ofthe Polar Front
by about 0.4 %0. Hence, a reduction in air-sea exchange could only explain part of the deglacial and
glacial minimum and does not resolve the cadmium/
8 1l C discrepancy (Michel et al. 1995).
The deglacial minimum can also be explained
with changes in surface water hydrology. Zielinski
et al. (1998) have shown that deglacial temperatures recorded in core PS1768-8 were up to 3°C
warmer than during the late Holocene. This warming and an additional surface freshening (Niebler
1995) may have produced the huge negative 8 18 0
anomaly in N. pachyderma during the last deglaciation as measured in core PS1768-8 (arrows
in Fig. 9a). The warmer temperatures may indicate
that surface waters originated further north, where
equilibration with atmospheric CO 2 took place at
warmer temperatures.
An alternative explanation for the 8 1l C variability in planktic foraminifera was provided by Lea et
al. (this volume). They attributed the low glacial
8 1l C values of G. bulloides and N. pachyderma
to higher carbon ion contents of Southern Ocean
surface waters. Although this effect might cause
the glacial reduction in foraminiferal8 ll C, it cannot explain the huge deglacial 8 1l C - minimum,
because increasing atmospheric CO 2 during
deglaciation constrains decreased carbon ion content of deglacial subantarctic surface waters (Lea
et al. this volume).
Benguela Upwelling. To what extent is the glacial-interglacial variability of high latitudes transferred to the World Ocean? The answer to this
question provides hints about changes in the general circulation pattern. In this regard, upwelling
regions are important, because nutrients are normally not fully utilized and changes in the chemistry of upwelled waters should directly be transferred to surface waters. Most carbon isotope
studies, intended to monitor the upwelling regions
in the past, used G. bulloides because this species
is well adapted to upwelling conditions. Isotope
studies of plankton tows and core tops have shown
that the 8 1l C value of G. bulloides increases in
upwelling regions relative to open ocean conditions
(Ganssen 1983; Kroon and Ganssen 1988;
439
increase in nutrient levels might result from any
combination of (1) decreased biological activity, (2)
increased upwelling rates or (3) increased nutrient
concentrations of source waters. Charles and
Fairbanks (1990) proposed that the increased nutrient concentrations in the Southern Ocean during the LGM were due to a reduced contribution
by nutrient poor NADW. This hypothesis is supported by by the isotopic composition of benthic
foraminifera, which record a comparable high
glacial-intergiacial8 1l C shift of about 1.3 %0 in the
Southern Ocean (see compilation and references
in Michel et al. 1995). Cd/Ca and BafCa ratios of
benthic foraminifera, however, suggest that the
nutrient concentrations within CPDW remained
close to the modern level (Lea 1995; Boyle and
Rosenthal 1996). Additionally, the nutrient utilization was increased south of the APF and remained
unchanged north of the APF (Francois et al. 1997).
This should result, at least south of the APF, in
higher 8 ll C-values of surface waters.
The discrepancy between the data might be
resolved by considering the habitat of N.
pachyderma and the hydrologic changes in the
Southern Ocean during the LGM. Results from
plankton tows indicate that N. pachyderma secretes % of its shell in water depth between 50 and
200 m (Kohfeld et al. 1997; Carstens et al. 1997).
Duplessy et al. (1996) have shown that the salinity
of surface waters in the Antarctic Zone were considerably reduced during the LGM which probably
resulted in an increased vertical stratification. Since
the modem upper water column is relatively well
mixed in the Southern Ocean, N. pachyderma
records surface water conditions despite its deeper
habitat. Under more stratified conditions, however,
their habitat would be enriched in nutrients and
hence their shells would record a lower8 ll C. Support for this hypothesis comes from shallow dwelling
G. bulloides which records a much lower glaciall
interglacial amplitude in the subantarctic zone
(Ninnemann and Charles 1997).
Due to the higher stratification or sea ice cover,
the gas exchange between atmosphere and ocean
might have been considerably reduced during the
glacial. This would cause an additional decrease of
the 8 1l C of surface waters and probably also in the
water masses sinking down from Southern Ocean
surface waters. Fig. lc indicates that nowadays the
exchange with the atmosphere increases the 8 1l C
of surface waters in the vicinity ofthe Polar Front
by about 0.4 %0. Hence, a reduction in air-sea exchange could only explain part of the deglacial and
glacial minimum and does not resolve the cadmium/
8 1l C discrepancy (Michel et al. 1995).
The deglacial minimum can also be explained
with changes in surface water hydrology. Zielinski
et al. (1998) have shown that deglacial temperatures recorded in core PS1768-8 were up to 3°C
warmer than during the late Holocene. This warming and an additional surface freshening (Niebler
1995) may have produced the huge negative 8 18 0
anomaly in N. pachyderma during the last deglaciation as measured in core PS1768-8 (arrows
in Fig. 9a). The warmer temperatures may indicate
that surface waters originated further north, where
equilibration with atmospheric CO 2 took place at
warmer temperatures.
An alternative explanation for the 8 1l C variability in planktic foraminifera was provided by Lea et
al. (this volume). They attributed the low glacial
8 1l C values of G. bulloides and N. pachyderma
to higher carbon ion contents of Southern Ocean
surface waters. Although this effect might cause
the glacial reduction in foraminiferal8 ll C, it cannot explain the huge deglacial 8 1l C - minimum,
because increasing atmospheric CO 2 during
deglaciation constrains decreased carbon ion content of deglacial subantarctic surface waters (Lea
et al. this volume).
Benguela Upwelling. To what extent is the glacial-interglacial variability of high latitudes transferred to the World Ocean? The answer to this
question provides hints about changes in the general circulation pattern. In this regard, upwelling
regions are important, because nutrients are normally not fully utilized and changes in the chemistry of upwelled waters should directly be transferred to surface waters. Most carbon isotope
studies, intended to monitor the upwelling regions
in the past, used G. bulloides because this species
is well adapted to upwelling conditions. Isotope
studies of plankton tows and core tops have shown
that the 8 1l C value of G. bulloides increases in
upwelling regions relative to open ocean conditions
(Ganssen 1983; Kroon and Ganssen 1988;