474
Andersen et al.
pre-industrial period; TakahwIll et al. 1992; Fischer
et al. 1998) from modern surface water PC0 2
values (Weiss et al. 1992). These pre-industrial
PCO, values were converted in [C0 2 (aq)]p".ind via
Henry's Law using the alkenone-derived SST and
assuming a constant salinity of35 %0 (see definitions and calculations).
Foraminiferal8 13 C values are not available for
all surface sediments from the South Atlantic to
estimate the pre-industrial8 13 C of the LCO, in the
surface waters. Therefore, a constant value of2.5
%0 (Kroopnick 1985) was used for all core-top
samples included in the calibration. To estimate the
isotopic composition of ancient LC0 2 , the 8 13 C
record ofG. ruber (pink) from GeoB 1016-3 was
used.
Extraction and Isotope Analysis of
Alkenones
Alkenones were extracted from 250 mg to 11 g
(depending on the alkenone content) of freeze-dried
sediment by ultrasonication using an UP 200H sonic
disruptor probe (200 W, amplitude 0.5, pulse 0.5)
and successively less polar mixtures of methanol
and methylene chloride (MeOH, MeOHlCH,C~
1: 1, CH 2 Cl), each for 3 min. The samples were
centrifuged and the three extracts com bined. They
were then desalted, dried over Na,S04' concentrated under N 2 and finally taken up in 25 - IS 0 ,.tI
of a 1: 1 MeOHlCH 2 C1 2 mixture.
The carbon isotopic composition of the C 37 • 2
alkenone, the dominant alkenone in our samples,
was determined by gas chromatography isotope
ratio mass spectrometry (GC-IRMS) (e.g., Hayes
et al. 1990; Merritt and Hayes 1994) using a
Finnigan MAT 252 mass spectrometer connected
to a combustion interface and a HP 5890A gas
chromatograph. One to eight ,.tI aliquots of each
extract were analyzed on 50 m x 0.32 mm high
temperature columns (SGE-BPX 5 and Optima I;
each 0.5 fim films) with a 2.5 m retention gap using on-column injection for extracts ofthe surface
sediments and splitlsplitless injection (1: 10 split
modus) for that of core GeoB 1016-3. The gas
chromatograph was programmed from 50°C (5 min
initial time for on-column) with 30° /min to 150°C,
with 8°/min to 230°C, with 6°/min to 320°C (final
time 48 min) and helium was used as the carrier
gas. Before GC-injection, the core-top extracts
were purified by passing them over a silica gel
column, followed by saponification (to remove esters) with 0.1 M methanolic KOH at 80°C for 2
hours. Based on a lab internal standard, there was
no significant difference between the isotopic composition obtained with the two different columns for
both purified and non-purified extracts.
The isotope ratio ofthe alkenone was obtained
by comparison with a lab-internal reference gas
(C0 2 ) and with coinjected n-alkanes (n-C w n-C36'
n-C 38 ) which were used as internal isotopic standards to correct for performance changes of the
oxidation oven. Usually each extract was measured
four times and a few samples were also extracted
in duplicate. Based on these analyses, the average
precision of the isotopic values is ± 0.2 %0.
Definitions and Calculations
According to Henry's law (eq. I), the pressure of
CO 2 in oceanic waters (PC0 2 ) is directly related
to the concentration of dissolved molecular CO 2 ,
[CO,(aq)], and inversely related to the CO 2 solubility coefficient a which is mainly a function of
temperature and, to a minor extent, also of salinity
(Skirrow 1975; Rau et al. 1991):
PCO, = [CO,(aq)]/ a
(I ).
The [CO,(aq)] was obtained from the carbon
isotopic composition ofC 3 7.z alkenones and coeval
carbonates. To describe the dependence of photosynthetic carbon metabolism on [CO,(aq)], the isotopic fractionation accompanying photosynthetic
fixation of carbon, Ep' was used (Freeman and
Hayes 1992). The Ep is defined in eq. 2, where 8 d
is the isotopic composition of carbon in CO,(aq)
and 8 p is the isotopic composition of carbon in the
photosynthate.
Ep = [(8 d + 1000)/(8 p + 1000) - 1] * 1000
(2)
For algae, where CO 2 is thought to reach the
photosynthetic site only by passive diffusion (Raven
and 10hnston 1991), Ep can be described by a simple equation (Bidigare et al. 1997). Under normal
Andersen et al.
pre-industrial period; TakahwIll et al. 1992; Fischer
et al. 1998) from modern surface water PC0 2
values (Weiss et al. 1992). These pre-industrial
PCO, values were converted in [C0 2 (aq)]p".ind via
Henry's Law using the alkenone-derived SST and
assuming a constant salinity of35 %0 (see definitions and calculations).
Foraminiferal8 13 C values are not available for
all surface sediments from the South Atlantic to
estimate the pre-industrial8 13 C of the LCO, in the
surface waters. Therefore, a constant value of2.5
%0 (Kroopnick 1985) was used for all core-top
samples included in the calibration. To estimate the
isotopic composition of ancient LC0 2 , the 8 13 C
record ofG. ruber (pink) from GeoB 1016-3 was
used.
Extraction and Isotope Analysis of
Alkenones
Alkenones were extracted from 250 mg to 11 g
(depending on the alkenone content) of freeze-dried
sediment by ultrasonication using an UP 200H sonic
disruptor probe (200 W, amplitude 0.5, pulse 0.5)
and successively less polar mixtures of methanol
and methylene chloride (MeOH, MeOHlCH,C~
1: 1, CH 2 Cl), each for 3 min. The samples were
centrifuged and the three extracts com bined. They
were then desalted, dried over Na,S04' concentrated under N 2 and finally taken up in 25 - IS 0 ,.tI
of a 1: 1 MeOHlCH 2 C1 2 mixture.
The carbon isotopic composition of the C 37 • 2
alkenone, the dominant alkenone in our samples,
was determined by gas chromatography isotope
ratio mass spectrometry (GC-IRMS) (e.g., Hayes
et al. 1990; Merritt and Hayes 1994) using a
Finnigan MAT 252 mass spectrometer connected
to a combustion interface and a HP 5890A gas
chromatograph. One to eight ,.tI aliquots of each
extract were analyzed on 50 m x 0.32 mm high
temperature columns (SGE-BPX 5 and Optima I;
each 0.5 fim films) with a 2.5 m retention gap using on-column injection for extracts ofthe surface
sediments and splitlsplitless injection (1: 10 split
modus) for that of core GeoB 1016-3. The gas
chromatograph was programmed from 50°C (5 min
initial time for on-column) with 30° /min to 150°C,
with 8°/min to 230°C, with 6°/min to 320°C (final
time 48 min) and helium was used as the carrier
gas. Before GC-injection, the core-top extracts
were purified by passing them over a silica gel
column, followed by saponification (to remove esters) with 0.1 M methanolic KOH at 80°C for 2
hours. Based on a lab internal standard, there was
no significant difference between the isotopic composition obtained with the two different columns for
both purified and non-purified extracts.
The isotope ratio ofthe alkenone was obtained
by comparison with a lab-internal reference gas
(C0 2 ) and with coinjected n-alkanes (n-C w n-C36'
n-C 38 ) which were used as internal isotopic standards to correct for performance changes of the
oxidation oven. Usually each extract was measured
four times and a few samples were also extracted
in duplicate. Based on these analyses, the average
precision of the isotopic values is ± 0.2 %0.
Definitions and Calculations
According to Henry's law (eq. I), the pressure of
CO 2 in oceanic waters (PC0 2 ) is directly related
to the concentration of dissolved molecular CO 2 ,
[CO,(aq)], and inversely related to the CO 2 solubility coefficient a which is mainly a function of
temperature and, to a minor extent, also of salinity
(Skirrow 1975; Rau et al. 1991):
PCO, = [CO,(aq)]/ a
(I ).
The [CO,(aq)] was obtained from the carbon
isotopic composition ofC 3 7.z alkenones and coeval
carbonates. To describe the dependence of photosynthetic carbon metabolism on [CO,(aq)], the isotopic fractionation accompanying photosynthetic
fixation of carbon, Ep' was used (Freeman and
Hayes 1992). The Ep is defined in eq. 2, where 8 d
is the isotopic composition of carbon in CO,(aq)
and 8 p is the isotopic composition of carbon in the
photosynthate.
Ep = [(8 d + 1000)/(8 p + 1000) - 1] * 1000
(2)
For algae, where CO 2 is thought to reach the
photosynthetic site only by passive diffusion (Raven
and 10hnston 1991), Ep can be described by a simple equation (Bidigare et al. 1997). Under normal