Clues to Ocean History: a Brief Overview of Proxies
27
parameters may be used as productivity proxies at
this site in the Zaire Fan. However, at other sites,
there is no indication that biogenic opal may serve
as a paleoproductivity proxy, e.g. in the western
equatorial Pacific Ocean (Lange and Berger 1993).
There, diatom abundance patterns are not parallel
to either flux or any nutrient proxy.
Biogenic opal production and accumulation in
the sediment is dependent on the silicate supply into
the photic layer and to postdepositional dissolution
which both vary regionally and temporally. In the
open ocean, Si regulation of diatom growth has
been proven in many studies (e.g. Dugdale and
Wilkerson 1998) but also iron availability may be
important (e.g. Martin and Fitzwater 1988). With
the model of the "silicate pump", Dugdale et al.
(1995) showed that open-ocean systems may become silicate-limited due to differential export of
Si relative to Nand P. This leads to vertical and
inter-ocean contrasts of the silicate versus nitrate/
phosphate and to the present-day productivity pattern of biogenic opal in the ocean. Experimental
data (e.g. from Hutchins and Bruland 1998) now
shows that also iron may affect the uptake of silicate relative to nitrate and phosphate. Under ironlimited conditions, diatoms grow thicker shells (uptake of more Si relative to N) and can be expected
to sink faster and may thus be better preserved in
the sediment. As a consequence, downcore variations in the opal accumulation would not reflect productivity but the changing supply of iron into the surface waters. In addition, iron supply would influence the vertical and inter-ocean nutrient distributions; in the last glacial, for instance, higher iron input
into the ocean could have resulted in less silicified
diatoms and a more rapid dissolution and recycling
of the skeletons in the water column compared to
the present-day conditions.
Regional variations in the biogenic silica content
of surface sediments may also be influnced by opal
recycling intensity on the seafloor. Areas with high
opal accumulation do not necessarily correspond
to high carbon and silica production. In the Southern Ocean, more than 50% of the global biogenic
opal accumulation occurs (DeMaster 1981) but
only about 4% of the total primary productivity is
observed (Nelson et al. 1995). The development
of biogenic opal as a paleoproductivity proxy will
therefore require a better understanding of the Si
cycle in the modern ocean, including changes in
pre-formed silicate within the thermocline waters
(Berger and Lange 1998). To obtain reliable results,
this proxy should only be used within a multi-proxy
approach to estimate marine paleoproductivity.
Barite has long been recognized as an indicator of productivity (see also Gingele et al. this volume). Goldberg and Arrhenius (1958), for example, documented a distinct peak of barite accumulation below the eastern Pacific equatorial
upwelling region in a N-S transect. Compared to
other paleoproductivity proxies (e.g. organic carbon), barium is relatively refractory. Around 30%
of the biogenic Ba flux is embedded in the sediment,
normally in the form of discrete barite crystals
(Dymond et al. 1992). It is not clear exactly how
Ba enters the export flux. In low latitudes, incorporation into celestite in acantharians (protists related to radiolarians) and subsequent precipitation
as microcrystals of barite within sinking aggregates
may be important. Elsewhere, Ba occurs in certain microenvironments or in diatoms. Some authors suggested that the release of sulfate from
decaying organic matter may supersaturate the
microenvironment with respect to barite (for a discussion of possible mechanisms see Bishop 1989).
In any case, changing barite abundances agree well
with fluctuations in productivity on glacial-interglacial time scales (Dehairs et al. 1980; Dymond et
al. 1992; Gingele and Dahmke 1994; Pay tan and
Kastner 1996). Quantitative computations of productivity are based on algorithms based on sediment
trap data (e.g. Dymond et al. 1992; Francois et al.
1995).
Barite as a proxy, like opal, is subject to variations in availibility and may not be a "pure" indicator of productivity, because availability of dissolved
Ba plays an important role in the formation of barite
(Ba attains very low concentrations in the ocean).
Problems arise when the relationship between Ba
content and paleoproductivity are applied to predominantly terrigenous sediments. A major source
of error are incorrect estimates of the terrigenous
Ba! Al ratio (Klump et al. subm). In addition, small
changes in oxygen concentration, at sensitive levels, can influence the abundance of sulfate ions in
interstitial waters, and hence the stability of micro-
27
parameters may be used as productivity proxies at
this site in the Zaire Fan. However, at other sites,
there is no indication that biogenic opal may serve
as a paleoproductivity proxy, e.g. in the western
equatorial Pacific Ocean (Lange and Berger 1993).
There, diatom abundance patterns are not parallel
to either flux or any nutrient proxy.
Biogenic opal production and accumulation in
the sediment is dependent on the silicate supply into
the photic layer and to postdepositional dissolution
which both vary regionally and temporally. In the
open ocean, Si regulation of diatom growth has
been proven in many studies (e.g. Dugdale and
Wilkerson 1998) but also iron availability may be
important (e.g. Martin and Fitzwater 1988). With
the model of the "silicate pump", Dugdale et al.
(1995) showed that open-ocean systems may become silicate-limited due to differential export of
Si relative to Nand P. This leads to vertical and
inter-ocean contrasts of the silicate versus nitrate/
phosphate and to the present-day productivity pattern of biogenic opal in the ocean. Experimental
data (e.g. from Hutchins and Bruland 1998) now
shows that also iron may affect the uptake of silicate relative to nitrate and phosphate. Under ironlimited conditions, diatoms grow thicker shells (uptake of more Si relative to N) and can be expected
to sink faster and may thus be better preserved in
the sediment. As a consequence, downcore variations in the opal accumulation would not reflect productivity but the changing supply of iron into the surface waters. In addition, iron supply would influence the vertical and inter-ocean nutrient distributions; in the last glacial, for instance, higher iron input
into the ocean could have resulted in less silicified
diatoms and a more rapid dissolution and recycling
of the skeletons in the water column compared to
the present-day conditions.
Regional variations in the biogenic silica content
of surface sediments may also be influnced by opal
recycling intensity on the seafloor. Areas with high
opal accumulation do not necessarily correspond
to high carbon and silica production. In the Southern Ocean, more than 50% of the global biogenic
opal accumulation occurs (DeMaster 1981) but
only about 4% of the total primary productivity is
observed (Nelson et al. 1995). The development
of biogenic opal as a paleoproductivity proxy will
therefore require a better understanding of the Si
cycle in the modern ocean, including changes in
pre-formed silicate within the thermocline waters
(Berger and Lange 1998). To obtain reliable results,
this proxy should only be used within a multi-proxy
approach to estimate marine paleoproductivity.
Barite has long been recognized as an indicator of productivity (see also Gingele et al. this volume). Goldberg and Arrhenius (1958), for example, documented a distinct peak of barite accumulation below the eastern Pacific equatorial
upwelling region in a N-S transect. Compared to
other paleoproductivity proxies (e.g. organic carbon), barium is relatively refractory. Around 30%
of the biogenic Ba flux is embedded in the sediment,
normally in the form of discrete barite crystals
(Dymond et al. 1992). It is not clear exactly how
Ba enters the export flux. In low latitudes, incorporation into celestite in acantharians (protists related to radiolarians) and subsequent precipitation
as microcrystals of barite within sinking aggregates
may be important. Elsewhere, Ba occurs in certain microenvironments or in diatoms. Some authors suggested that the release of sulfate from
decaying organic matter may supersaturate the
microenvironment with respect to barite (for a discussion of possible mechanisms see Bishop 1989).
In any case, changing barite abundances agree well
with fluctuations in productivity on glacial-interglacial time scales (Dehairs et al. 1980; Dymond et
al. 1992; Gingele and Dahmke 1994; Pay tan and
Kastner 1996). Quantitative computations of productivity are based on algorithms based on sediment
trap data (e.g. Dymond et al. 1992; Francois et al.
1995).
Barite as a proxy, like opal, is subject to variations in availibility and may not be a "pure" indicator of productivity, because availability of dissolved
Ba plays an important role in the formation of barite
(Ba attains very low concentrations in the ocean).
Problems arise when the relationship between Ba
content and paleoproductivity are applied to predominantly terrigenous sediments. A major source
of error are incorrect estimates of the terrigenous
Ba! Al ratio (Klump et al. subm). In addition, small
changes in oxygen concentration, at sensitive levels, can influence the abundance of sulfate ions in
interstitial waters, and hence the stability of micro-