420
Frank eta!.
glacial peaks south of the APF. At the northerly
location, focussing occurred in the whole core with
higher values during the interglacials yielding uniformly low rain rates around I g/cm 2 kyr throughout the core. This is interpreted in terms of a glacial northward extension of the opal productivity
belt to about the location of core PS2499-5 during
glacial stages 2 and 4 (see also Fig. 5). The high
opaline productivity area did not reach the location
of PS2498-1 during the last 130 kyr. This means
that the present day position of the SAF is about
corresponding to the glacial northern boundary of
the high biogenic particle flux area indicating a
maximum glacial northward shift of this area of
about 5° compared to today. This finding stands in
accordance with other published estimates (Charles
et at. 1991; Mortlock et at. 1991). The results are
also in accordance with the paleoproductivity reconstructions from the other cores obtained from
similar locations relative to the position ofthe SAF
described above. Only core PSI754-1, although
located within the PFZ, is still dominated by carbonate sedimentation during glacials which may be
explained by the core's shallow water depth of only
2400m.
Unfortunately, the use of carbonate, biogenic
opal, and biogenic Ba is restricted due to variable
dissolution effects, which are mainly a function of
saturation of the respective biogenic component in
the water column. For Ba diagentic remobilisation
in anoxic or suboxic sediments may also occur. It
is therefore suggested that the initial 231 Pa ,/,30Th"
ratio, although there may also be some uncertainties (Walter et at. this volume), is a more reliable
measure of biogenic particle flux (Asmus et
al.1999). The basic idea is that the occurrence of
high biogenic particle fluxes, especially biogenic
opal, induce a boundary scavenging of231Pa from
areas of lower particle fluxes. The relationship
between the 2l1Pael30Thex ratio and particle fluxes
was shown to be nearly linear from sediment trap
studies (Yu 1994; Kumar et at. 1995). These
radionuclide ratios have the advantage of being
independent from sediment redistribution and dissolution effects. The 231Pae/230Th ex ratio profile of
PS2499-5 (Fig. 7b) clearly coincides with the high
biogenic particle fluxes during the isotope stages 2
and 4 deduced from bulk rain rates (Fig. 7c) and
opal rain rates (G. Bohrmann, pers. comm.). However, recent results suggest that the fractionation
between Pa and Th breaks down in the high opaline
particle flux area south of the present day position
of the APF (Walter et at. 1997). This implies that
Pa is very efficiently scavenged by biogenic opal
and, in case of a change from calcareous to
opaline dominated particle flux, such as in Core
PS2499-5, a part ofthe signal may be caused by
the change ofthe material, so that the linear relation ofthe 231Pael30Thex ratio to particle flux is not
given any more. It follows that this ratio should only
be used for qualitative, rather than quantitative, estimations of export paleoproductivity in the area of
high biogenic opal flux in the Southern Ocean.
Conclusions and Outlook
As a consequence of the results shown it is crucial to measure 23 0 Th ex profiles in cores where the
accumulation rates of sedimentary proxies are intended to be used for paleoceanographic reconstructions or paleoproductivity estimates. The calculation of the 230Th ex fluxes from Southern Ocean
sediments has revealed highly variable amounts of
laterally redistributed material which has masked
the original pattern of mass accumulation rates and
any paleoproductivity reconstruction based on
those. The focussing factors mostly deviate by more
than 50 % from the expected 230Th ex flux (values
range from 12.5 to 0.2) which represents the estimated uncertainty of the method (based on sediment trap data). The cause of the uncertainty is
mainly advective transport of 230Th in the water
column which may influence the assumed constant
flux of 230Th ex ' Strong variations of redistribution
intensity occurred within single cores and have impressively demonstrated the importance of a correction for lateral sediment redistribution.
However, also results based on radionuclide
ratios (e.g. 231PaerOThe) do not deliver unambiguous quantitative results on their own. Therefore,
multitracer approaches of as many proxies as possible will always lead to more reliable and more
detailed results. Together with other proxies, the
radionuclides have a strong potential to contribute
to a better understanding of the complex processes
involved in marine paleoproductivity. This is par-
Frank eta!.
glacial peaks south of the APF. At the northerly
location, focussing occurred in the whole core with
higher values during the interglacials yielding uniformly low rain rates around I g/cm 2 kyr throughout the core. This is interpreted in terms of a glacial northward extension of the opal productivity
belt to about the location of core PS2499-5 during
glacial stages 2 and 4 (see also Fig. 5). The high
opaline productivity area did not reach the location
of PS2498-1 during the last 130 kyr. This means
that the present day position of the SAF is about
corresponding to the glacial northern boundary of
the high biogenic particle flux area indicating a
maximum glacial northward shift of this area of
about 5° compared to today. This finding stands in
accordance with other published estimates (Charles
et at. 1991; Mortlock et at. 1991). The results are
also in accordance with the paleoproductivity reconstructions from the other cores obtained from
similar locations relative to the position ofthe SAF
described above. Only core PSI754-1, although
located within the PFZ, is still dominated by carbonate sedimentation during glacials which may be
explained by the core's shallow water depth of only
2400m.
Unfortunately, the use of carbonate, biogenic
opal, and biogenic Ba is restricted due to variable
dissolution effects, which are mainly a function of
saturation of the respective biogenic component in
the water column. For Ba diagentic remobilisation
in anoxic or suboxic sediments may also occur. It
is therefore suggested that the initial 231 Pa ,/,30Th"
ratio, although there may also be some uncertainties (Walter et at. this volume), is a more reliable
measure of biogenic particle flux (Asmus et
al.1999). The basic idea is that the occurrence of
high biogenic particle fluxes, especially biogenic
opal, induce a boundary scavenging of231Pa from
areas of lower particle fluxes. The relationship
between the 2l1Pael30Thex ratio and particle fluxes
was shown to be nearly linear from sediment trap
studies (Yu 1994; Kumar et at. 1995). These
radionuclide ratios have the advantage of being
independent from sediment redistribution and dissolution effects. The 231Pae/230Th ex ratio profile of
PS2499-5 (Fig. 7b) clearly coincides with the high
biogenic particle fluxes during the isotope stages 2
and 4 deduced from bulk rain rates (Fig. 7c) and
opal rain rates (G. Bohrmann, pers. comm.). However, recent results suggest that the fractionation
between Pa and Th breaks down in the high opaline
particle flux area south of the present day position
of the APF (Walter et at. 1997). This implies that
Pa is very efficiently scavenged by biogenic opal
and, in case of a change from calcareous to
opaline dominated particle flux, such as in Core
PS2499-5, a part ofthe signal may be caused by
the change ofthe material, so that the linear relation ofthe 231Pael30Thex ratio to particle flux is not
given any more. It follows that this ratio should only
be used for qualitative, rather than quantitative, estimations of export paleoproductivity in the area of
high biogenic opal flux in the Southern Ocean.
Conclusions and Outlook
As a consequence of the results shown it is crucial to measure 23 0 Th ex profiles in cores where the
accumulation rates of sedimentary proxies are intended to be used for paleoceanographic reconstructions or paleoproductivity estimates. The calculation of the 230Th ex fluxes from Southern Ocean
sediments has revealed highly variable amounts of
laterally redistributed material which has masked
the original pattern of mass accumulation rates and
any paleoproductivity reconstruction based on
those. The focussing factors mostly deviate by more
than 50 % from the expected 230Th ex flux (values
range from 12.5 to 0.2) which represents the estimated uncertainty of the method (based on sediment trap data). The cause of the uncertainty is
mainly advective transport of 230Th in the water
column which may influence the assumed constant
flux of 230Th ex ' Strong variations of redistribution
intensity occurred within single cores and have impressively demonstrated the importance of a correction for lateral sediment redistribution.
However, also results based on radionuclide
ratios (e.g. 231PaerOThe) do not deliver unambiguous quantitative results on their own. Therefore,
multitracer approaches of as many proxies as possible will always lead to more reliable and more
detailed results. Together with other proxies, the
radionuclides have a strong potential to contribute
to a better understanding of the complex processes
involved in marine paleoproductivity. This is par-