Clues to Ocean History: a Brief Overview of Proxies
19
(1991) used such a method to reconstruct surface
water salinity of the North Atlantic based on a
comparison between transfer function estimates of
SST and the oxygen isotope ratio of planktic
foraminifersl species (Fig. 10). The reconstruction
for the LGM shows significant differences to the
modern pattern and the presence of a sharp gradient associated with the Polar Front.
Rostek et al. (1993) calculated temperatures for
the Indian Ocean using alkenones, and then used
the /i180 values of G. ruber (white) for determining the salinities (see also Wolff et al. this volume).
Even when the temperatures can be determined
with some confidence, there is still a large uncertainty in salinity calculations due to the different
/i180- salinity relationships in the modern ocean,
which suggests a lack of constancy for the past,
even at the same location. This may result in large
uncertainties in the salinity calculation. If we assume an error of at least 1°C in the temperature
estimate, then the error for the estimated salinity
is at least 1.2 0 / 00 which is in the range of presentday open ocean salinity fluctuations. Rostek et al.
(1993) estimated an average analytical error of
about ±0.3 0/ 00 for the reconstructed changes in
salinity. The/i180-salinityrelationship is particularly
sensitive to variations in climate, ocean currents,
and evaporation/precipitation conditions. At present,
reconstructed salinity patterns must therefore be
considered as mainly qualitative information.
Reconstruction of Bottom-Water Age and
Flow
The deep circulation of the present ocean is well
depicted in the progression of age of bottom waters from their sources, along their flow paths, toward the basins farthest removed from the sources
(Fig. 11). The highest oxygen content, over 5 mill,
is found in the Atlantic Ocean due to the sinking of
oxygen-saturated surface waters in the northern
Atlantic. It is possible for a part of the NADW to
come into contact with surface waters again in the
region of the Weddell Sea, and take up additional
oxygen. However, the time and amount are not
sufficient to reach saturation. This water is distributed to the other oceans by the Antarctic
Circumpolar Current (ACC), which mixes
deepwater masses of various origins. However,
because these water masses have not been in contact with the atmosphere for a long time (they have
"aged"), the oxygen content is lower and continues
to decrease with further transport into the North
Pacific, where the "oldest" deep water is found.
With the consumption of oxygen by the oxidation
of organic substance, nutrient salts and CO 2 are
released to the water. Oxygen and nutrient contents, along with the /iJ3C values in LC0 2 , are coupled through this process.
Based on radiocarbon and the rate of oxygen
uptake in deep waters, the range of ages is just short
of 2000 years, which means that the average residence time of water masses in the deep ocean is
about 1000 years. In reconstruction, it is of interest to obtain clues on changes in the flow patterns
and on changes in overall age. Radiocarbon methods only work for the last 40,000 years or so, because ofthe short half-life ofl4C (~6000 y), and
the many opportunities for contamination at low
activity levels. Thus, indirect clues, tied to changes
in oxygen and associated changes in nutrients and
/iJ3C, have to be used to trace the aging of bottom
waters. Direct effects of bottom transport can be
studied by standard methods of sedimentology, that
is, the displacement of particles (e.g. clay, silt,
microfossils; see Diekmann et al. this volume).
Measurements regarding the paleochemistry of
bottom waters are made on the organisms living
within that water: benthic foraminifers and
ostracods. It is important to know if these organisms live on or just below the sea floor, since interstitial waters look "older" than overlying bottom
water. There are two important geochemical tracers that are widely used: /iJ3C and Cd/Ca ratios.
Values of /iJ3C decrease with age within deep
waters, being negatively correlated with oxygen,
while Cd content increases, being positively correlated with phosphate.
How can the /iJ3C values of benthic
foraminifers be used for this kind of reconstruction?
This works because the isotope composition of the
dissolved inorganic carbon is linked to deep water
circulation by the processes of photosynthesis and
respiration (for a review see Berger and Vincent
19
(1991) used such a method to reconstruct surface
water salinity of the North Atlantic based on a
comparison between transfer function estimates of
SST and the oxygen isotope ratio of planktic
foraminifersl species (Fig. 10). The reconstruction
for the LGM shows significant differences to the
modern pattern and the presence of a sharp gradient associated with the Polar Front.
Rostek et al. (1993) calculated temperatures for
the Indian Ocean using alkenones, and then used
the /i180 values of G. ruber (white) for determining the salinities (see also Wolff et al. this volume).
Even when the temperatures can be determined
with some confidence, there is still a large uncertainty in salinity calculations due to the different
/i180- salinity relationships in the modern ocean,
which suggests a lack of constancy for the past,
even at the same location. This may result in large
uncertainties in the salinity calculation. If we assume an error of at least 1°C in the temperature
estimate, then the error for the estimated salinity
is at least 1.2 0 / 00 which is in the range of presentday open ocean salinity fluctuations. Rostek et al.
(1993) estimated an average analytical error of
about ±0.3 0/ 00 for the reconstructed changes in
salinity. The/i180-salinityrelationship is particularly
sensitive to variations in climate, ocean currents,
and evaporation/precipitation conditions. At present,
reconstructed salinity patterns must therefore be
considered as mainly qualitative information.
Reconstruction of Bottom-Water Age and
Flow
The deep circulation of the present ocean is well
depicted in the progression of age of bottom waters from their sources, along their flow paths, toward the basins farthest removed from the sources
(Fig. 11). The highest oxygen content, over 5 mill,
is found in the Atlantic Ocean due to the sinking of
oxygen-saturated surface waters in the northern
Atlantic. It is possible for a part of the NADW to
come into contact with surface waters again in the
region of the Weddell Sea, and take up additional
oxygen. However, the time and amount are not
sufficient to reach saturation. This water is distributed to the other oceans by the Antarctic
Circumpolar Current (ACC), which mixes
deepwater masses of various origins. However,
because these water masses have not been in contact with the atmosphere for a long time (they have
"aged"), the oxygen content is lower and continues
to decrease with further transport into the North
Pacific, where the "oldest" deep water is found.
With the consumption of oxygen by the oxidation
of organic substance, nutrient salts and CO 2 are
released to the water. Oxygen and nutrient contents, along with the /iJ3C values in LC0 2 , are coupled through this process.
Based on radiocarbon and the rate of oxygen
uptake in deep waters, the range of ages is just short
of 2000 years, which means that the average residence time of water masses in the deep ocean is
about 1000 years. In reconstruction, it is of interest to obtain clues on changes in the flow patterns
and on changes in overall age. Radiocarbon methods only work for the last 40,000 years or so, because ofthe short half-life ofl4C (~6000 y), and
the many opportunities for contamination at low
activity levels. Thus, indirect clues, tied to changes
in oxygen and associated changes in nutrients and
/iJ3C, have to be used to trace the aging of bottom
waters. Direct effects of bottom transport can be
studied by standard methods of sedimentology, that
is, the displacement of particles (e.g. clay, silt,
microfossils; see Diekmann et al. this volume).
Measurements regarding the paleochemistry of
bottom waters are made on the organisms living
within that water: benthic foraminifers and
ostracods. It is important to know if these organisms live on or just below the sea floor, since interstitial waters look "older" than overlying bottom
water. There are two important geochemical tracers that are widely used: /iJ3C and Cd/Ca ratios.
Values of /iJ3C decrease with age within deep
waters, being negatively correlated with oxygen,
while Cd content increases, being positively correlated with phosphate.
How can the /iJ3C values of benthic
foraminifers be used for this kind of reconstruction?
This works because the isotope composition of the
dissolved inorganic carbon is linked to deep water
circulation by the processes of photosynthesis and
respiration (for a review see Berger and Vincent