422
Frank et al.
portant step towards quantitative reconstruction of
export paleoproductivity from the marine sedimentary record will be made.
One of the main disadvantages of the 23 0 Th"
method is its limited applicability to, at best, the last
300 kyr. A similar and reliable tool to evaluate sediment redistribution further back in time is required.
Recently, the measurement of stable 3He retained
in interplanetary dust particles (IDPs) within marine sediments was suggested to represent such a
tool (Marcantonio et al. 1995; 1996). For this idea
it has to be assumed that IDPs have constantly
rained down to the earth's surface. A comparison
of the 3He with 230Th ex yielded a variability of not
more than 75 %. The 3He-containing dust particles
are transferred to the marine sediments and, iftheir
accumulation rates vary with significantly more
than 75 %, this can only have been caused by sediment redistribution. Although more work needs to
be done to further corroborate these initial results
and to exclude possible uncertainties, the comparison with 2JOTh normalized accumulation rates
ex
shows a very good agreement and makes 3He a
very promising candidate to go further back in time
with the quantification oflateral sediment redistribution without the restrictions of radioactive decay.
Given the increasing amount of2 30 Th" data and
information on glacial interglacial changes of sediment redistribution intensity, a mapping of the \j1
data in the Atlantic Sector of the Southern Ocean
was carried out to see whether only random local
influences are responsible for the observed variations on glacial interglacial time scales, or whether
any systematic links to paleoclimate-driven changes
of the current regime are superimposed. Such systematics in the pattern of sediment redistribution
variations were not observed in the Norwegian Sea
(Scholten et al. 1994) but a link between sediment
redistribution intensity and climatically induced
changes of bottom current speeds apparently
exists in other areas, such as the equatorial Pacific
Ocean (systematically higher focussing intensity
during the glacial periods in two cores (Marcantonio
et al. 1996», or the Weddell Sea Continental Margin (systematically higher focussing intensity during interglacial stages 1, 5e and 7 in three cores;
Frank et al. 1995).
In Fig. 8a, a preliminary compilation 0[230Th -
based sediment redistribution data for the Atlantic
sector ofthe ACC is shown. The contours in these
three plots are certainly somewhat arbitrary due to
the low density ofthe data and because it is probable that there are cores in which the data only
represent information about very local and small
spatial scale redistribution. The Holocene (here
0-12 kyr B.P.) redistribution map (Fig. 8a) shows
that there are generally more locations with sediment focussing than winnowing. This may be partly
the consequence of a sediment sonar-system-based
choice of coring locations during the cruises. With
the exception of two locations just south of the APF
and one north of the SAF, the average amount of
sediment redistribution has been relatively small.
Looking at the map of the last glacial this pattern
is different (Fig. 8b). With the exception of the
southernmost core, the area south of 46°S remained almost unchanged compared to the
Holocene. However, further north, around the approximate glacial northern boundary of the high
particle flux area, strong focussing in six cores not
below a \j1 value of2.5 prevailed, which can not be
explained by the sonar-based choice of thick sediment cover for coring. This change of the pattern
can also be observed from the map showing the
ratios of the last glacial and the Holocene \j1 values (Fig. 8c). The largest changes occurred between 46°S and 43 oS. The high glacial to Holocene
ratio at the southernmost location (PS 1772-8), as
discussed previously, may not represent a real
change of focussing intensity, but might be due
to a stratigraphic artifact caused by the low age
resolution in this section of the core similar to
the effects described previously for core GeoB
1523-1 from the Ceara Rise.
Assuming that the change of the \j1 values has
exclusively been a function of sediment redistribution, a preliminary interpretation ofthese patterns
would be that there was a significant hydrographic
change from the last glacial to the Holocene around
the present day location of the SAF which might
be related to the speed ofthe ACC and the global
ocean's paleocirculation. Unfortunately, it is not
possible to unambiguously interpret a higher
focussing intensity in terms of higher or lower cur-
Frank et al.
portant step towards quantitative reconstruction of
export paleoproductivity from the marine sedimentary record will be made.
One of the main disadvantages of the 23 0 Th"
method is its limited applicability to, at best, the last
300 kyr. A similar and reliable tool to evaluate sediment redistribution further back in time is required.
Recently, the measurement of stable 3He retained
in interplanetary dust particles (IDPs) within marine sediments was suggested to represent such a
tool (Marcantonio et al. 1995; 1996). For this idea
it has to be assumed that IDPs have constantly
rained down to the earth's surface. A comparison
of the 3He with 230Th ex yielded a variability of not
more than 75 %. The 3He-containing dust particles
are transferred to the marine sediments and, iftheir
accumulation rates vary with significantly more
than 75 %, this can only have been caused by sediment redistribution. Although more work needs to
be done to further corroborate these initial results
and to exclude possible uncertainties, the comparison with 2JOTh normalized accumulation rates
ex
shows a very good agreement and makes 3He a
very promising candidate to go further back in time
with the quantification oflateral sediment redistribution without the restrictions of radioactive decay.
Given the increasing amount of2 30 Th" data and
information on glacial interglacial changes of sediment redistribution intensity, a mapping of the \j1
data in the Atlantic Sector of the Southern Ocean
was carried out to see whether only random local
influences are responsible for the observed variations on glacial interglacial time scales, or whether
any systematic links to paleoclimate-driven changes
of the current regime are superimposed. Such systematics in the pattern of sediment redistribution
variations were not observed in the Norwegian Sea
(Scholten et al. 1994) but a link between sediment
redistribution intensity and climatically induced
changes of bottom current speeds apparently
exists in other areas, such as the equatorial Pacific
Ocean (systematically higher focussing intensity
during the glacial periods in two cores (Marcantonio
et al. 1996», or the Weddell Sea Continental Margin (systematically higher focussing intensity during interglacial stages 1, 5e and 7 in three cores;
Frank et al. 1995).
In Fig. 8a, a preliminary compilation 0[230Th -
based sediment redistribution data for the Atlantic
sector ofthe ACC is shown. The contours in these
three plots are certainly somewhat arbitrary due to
the low density ofthe data and because it is probable that there are cores in which the data only
represent information about very local and small
spatial scale redistribution. The Holocene (here
0-12 kyr B.P.) redistribution map (Fig. 8a) shows
that there are generally more locations with sediment focussing than winnowing. This may be partly
the consequence of a sediment sonar-system-based
choice of coring locations during the cruises. With
the exception of two locations just south of the APF
and one north of the SAF, the average amount of
sediment redistribution has been relatively small.
Looking at the map of the last glacial this pattern
is different (Fig. 8b). With the exception of the
southernmost core, the area south of 46°S remained almost unchanged compared to the
Holocene. However, further north, around the approximate glacial northern boundary of the high
particle flux area, strong focussing in six cores not
below a \j1 value of2.5 prevailed, which can not be
explained by the sonar-based choice of thick sediment cover for coring. This change of the pattern
can also be observed from the map showing the
ratios of the last glacial and the Holocene \j1 values (Fig. 8c). The largest changes occurred between 46°S and 43 oS. The high glacial to Holocene
ratio at the southernmost location (PS 1772-8), as
discussed previously, may not represent a real
change of focussing intensity, but might be due
to a stratigraphic artifact caused by the low age
resolution in this section of the core similar to
the effects described previously for core GeoB
1523-1 from the Ceara Rise.
Assuming that the change of the \j1 values has
exclusively been a function of sediment redistribution, a preliminary interpretation ofthese patterns
would be that there was a significant hydrographic
change from the last glacial to the Holocene around
the present day location of the SAF which might
be related to the speed ofthe ACC and the global
ocean's paleocirculation. Unfortunately, it is not
possible to unambiguously interpret a higher
focussing intensity in terms of higher or lower cur-