Clues to Ocean History: a Brief Overview of Proxies
25
flux rates, given poorly constrained age models) the
equation relating productivity to C o'g content is of
the form:
(6),
where q is usually between 0.6 and 0.8 (Berger et
al. 1989). Taking C OCgb as the Holocene standard
value, the ratios down core, after exponentiation to
0.7, are a reasonable estimate of the factor of
change of productivity. The great precision suggested by more complicated formulations must be
largely doubted. The influx of organic material
redeposited from the shelves in regions close to
continental margins is another disturbing factor.
Within the sediment, C mg is constantly being destroyed by bacteria. This is especially true close to
the sea floor, but it is also evident several meters
below. This destruction first proceeds by using free
oxygen, but subsequently occurs by the reduction
of nitrate, manganese oxide, iron oxide, and sulfate.
Thus, an oxygen debt is built up within the sediment.
Instead of using C mg as a productivity indicator
(which will result in a general trend towards lower
estimates with increasing age of sediment) one can
reasonably substitute oxygen demand (reducing
power) as a proxy (Perks and Keeling 1998).
Uncertainties can arise through the admixture
of terrestrial organic matter which, if not subtracted, will be included in the computation for
paleoproductivity. The admixture of terrestrial carbon can be adjusted for, to a limited extent, by
measuring CIN ratios and 8 13 C in the organic matter using a simple mixing equation with two extreme
values (e.g. >- 21 % for marine material and -27 to
-28%0 for land plants; MUller et al. 1983, 1994).
Further information about the terrestrial portion is
provided by the CIN ratios. Values between 7 and
9 are considered to indicate marine and those> 15
are assumed to be ofterrestrial origin. Uncertainties in the endmember values, when using these
methods of identifying the proportion ofterrigenous
carbon, result in semiquantitive estimates only.
Fig. 15 shows the downcore variation of marine organic carbon (MOC) and terrigenous organic carbon (TOC) as calculated from the 8 13 C o'g
values and measured total organic carbon and applying a mixing equation (e.g. Fontugne and
Duplessy 1986). The figure shows a covariation of
the percentages ofMOC, biogenic barium and opal.
All three parameters may serve as productivity indicators at this particular location (Schneider et al.
1997). Such an approach using several proxies
("multi-proxy") for one parameter (i.e. productivity) greatly increases confidence concerning the
paleoceanographic reconstruction.
Regions where sediments have very low organic
carbon contents, that is oligotrophic ocean areas,
pose special problems because measurement error and varying preservation introduce large uncertainties. For such cases, especially, a number of
additional proxies are required to reconstruct productivity (see below).
Reconstruction of Productivity from
Carbonate, Opal, and Barite
A number of biologically derived substances other
than organic carbon also can be useful in assessing productivity fluctuations. Based on sediment trap
results it can be shown that there is a good correlation between organic carbon flux and carbonate
under certain conditions in the open ocean (Wefer
and Fischer 1993). Arrhenius (1952) first proposed
the use of carbonate accumulation rates in the
eastern equatorial Pacific to track changes in
upwelling. However, the intensity of carbonate dissolution varies greatly on glacial-interglacial
timescales, so that great caution is indicated in
making this type of argument.
Quite generally, the links between carbonate
production and (organic) carbon production are
more tenuous than the links internal to the organic
carbon cycle. As productivity increases, carbonate production goes through a maximum and decreases again for high values, when deposition of
biogenic opal (e.g. from diatoms) takes over. Also,
on the sea floor, an increased supply of organic
carbon results in increased dissolution of carbonate. Contrary to organic carbon, carbonate content
is high in the sediments in low productivity regions,
and relatively low in high productivity areas (through
dilution and dissolution). Thus, as a proxy for productivity, carbonate content is useless. Instead,
carbonate fluxes have to be considered.
25
flux rates, given poorly constrained age models) the
equation relating productivity to C o'g content is of
the form:
(6),
where q is usually between 0.6 and 0.8 (Berger et
al. 1989). Taking C OCgb as the Holocene standard
value, the ratios down core, after exponentiation to
0.7, are a reasonable estimate of the factor of
change of productivity. The great precision suggested by more complicated formulations must be
largely doubted. The influx of organic material
redeposited from the shelves in regions close to
continental margins is another disturbing factor.
Within the sediment, C mg is constantly being destroyed by bacteria. This is especially true close to
the sea floor, but it is also evident several meters
below. This destruction first proceeds by using free
oxygen, but subsequently occurs by the reduction
of nitrate, manganese oxide, iron oxide, and sulfate.
Thus, an oxygen debt is built up within the sediment.
Instead of using C mg as a productivity indicator
(which will result in a general trend towards lower
estimates with increasing age of sediment) one can
reasonably substitute oxygen demand (reducing
power) as a proxy (Perks and Keeling 1998).
Uncertainties can arise through the admixture
of terrestrial organic matter which, if not subtracted, will be included in the computation for
paleoproductivity. The admixture of terrestrial carbon can be adjusted for, to a limited extent, by
measuring CIN ratios and 8 13 C in the organic matter using a simple mixing equation with two extreme
values (e.g. >- 21 % for marine material and -27 to
-28%0 for land plants; MUller et al. 1983, 1994).
Further information about the terrestrial portion is
provided by the CIN ratios. Values between 7 and
9 are considered to indicate marine and those> 15
are assumed to be ofterrestrial origin. Uncertainties in the endmember values, when using these
methods of identifying the proportion ofterrigenous
carbon, result in semiquantitive estimates only.
Fig. 15 shows the downcore variation of marine organic carbon (MOC) and terrigenous organic carbon (TOC) as calculated from the 8 13 C o'g
values and measured total organic carbon and applying a mixing equation (e.g. Fontugne and
Duplessy 1986). The figure shows a covariation of
the percentages ofMOC, biogenic barium and opal.
All three parameters may serve as productivity indicators at this particular location (Schneider et al.
1997). Such an approach using several proxies
("multi-proxy") for one parameter (i.e. productivity) greatly increases confidence concerning the
paleoceanographic reconstruction.
Regions where sediments have very low organic
carbon contents, that is oligotrophic ocean areas,
pose special problems because measurement error and varying preservation introduce large uncertainties. For such cases, especially, a number of
additional proxies are required to reconstruct productivity (see below).
Reconstruction of Productivity from
Carbonate, Opal, and Barite
A number of biologically derived substances other
than organic carbon also can be useful in assessing productivity fluctuations. Based on sediment trap
results it can be shown that there is a good correlation between organic carbon flux and carbonate
under certain conditions in the open ocean (Wefer
and Fischer 1993). Arrhenius (1952) first proposed
the use of carbonate accumulation rates in the
eastern equatorial Pacific to track changes in
upwelling. However, the intensity of carbonate dissolution varies greatly on glacial-interglacial
timescales, so that great caution is indicated in
making this type of argument.
Quite generally, the links between carbonate
production and (organic) carbon production are
more tenuous than the links internal to the organic
carbon cycle. As productivity increases, carbonate production goes through a maximum and decreases again for high values, when deposition of
biogenic opal (e.g. from diatoms) takes over. Also,
on the sea floor, an increased supply of organic
carbon results in increased dissolution of carbonate. Contrary to organic carbon, carbonate content
is high in the sediments in low productivity regions,
and relatively low in high productivity areas (through
dilution and dissolution). Thus, as a proxy for productivity, carbonate content is useless. Instead,
carbonate fluxes have to be considered.