Kaolinite and Chlorite as Tracers of Modern and Late Quaternary Deep Water Circulation
295
Hofmann in press) and Antarctica (Petit et al.
1990; Grousset et al. 1992) during cold climatic
stages.
However, a significant influx ofterrigenous matter by winds of the southern westerlies seems to
be confined to the ocean basins offthe Patagonian
coast (Balsam et al. 1995). In the Indian sector of
the Southern Ocean only a maximum of 5% oftotal
detrital influx can be attributed to direct dust input
(Bareille et al. 1994).
Our sediment core studies in the ACC region,
which is characterized by its dense nepheloid suspension layers (Biscaye and Eittreim 1977), support the prominent role of deep water circulation
in the distribution of terrigenous particles, also during glacial periods. It is true that, in the same region, a higher input of iron-bearing terrigenous dust
during glacials may fertilize surface-water masses
and so enhances biological production, export production and terrigenous particle fluxes to the sea
floor (Kumar et al. 1995). However, the settlement
of wind-borne particles from the surface through
the water column contributes only one part of the
total particle supply to the sea-floor and cannot
solely explain strong fluxes ofterrigenous matter
at the sea-floor. Several lines of evidence argue for
significant lateral near-bottom current transport of
sedimentary particles rather than primary long-distance eolian influx to the pelagic ACC region during glacial times.
A first approach is the comparison of AR tw -
values in the ACC region with ARt",-values at
sites in the equatorial Atlantic, where eolian influx
of terrigenous matter from the Sahara and Sahel
zone predominates (Tiedemann et al. 1989;
Ruddiman 1997). The sediment core locations on
submarine topographic highs in the ACC region, as
for instance those atthe Mid-Atlantic Ridge, should
provide the best record of eolian sedimentation.
Actually, ARtw-values of the Last Glacial Maximum documented in PS2495-3, PS2498-1, and
PS2499-1 from the Mid-Atlantic Ridge (Al\,,,values: 2-11 g cm· 2 ka·t)are higher and more variable than at sites in the equatorial Atlantic (AR t ,,,values: <2.0 g cm· 2 ka-') with similar distances to
the source areas of continental dust. Even if considering an enlargement of the Patagonian deflation area as owing to low stands of sea-level and
shelf exposure, this modified area size cannot be
compared to the size of the wide arid regions of
Northern Africa and thus the great potential sediment yield from there.
AR terr -values in the ACC region differ markedly
between the investigated sites. Especially the three
sediment cores from the Mid-Atlantic Ridge reveal
dramatic shifts of ARtw-valueswithin relatively
short distance (3 0 latitude) that rule out an uniform
eolian-controlled pelagic sedimentation (Fig. 4).
Glacial AR -values in the northernmost sediment
lerr
core (PS2495-3) is half a magnitude lower
than in PS2498-1 and PS2499-5 south of that
position. Moreover, grain sizes ofterrigenous silt
in PS2495-3 are smaller than in the other sediment
cores and may be explained by lesser current velocities at the position of PS2495-3 that lies in the
marginal jets of the ACC (Diekmann and Kuhn
1997). Variations of ARtw-values in a similar manner as the grain size parameters are consistent with
current-controlled sedimentation.
The conclusion that a lot of bulk particle fluxes
in the ACC region may be attributed to lateral sediment transport and reworking through winnowing
and focusing rather than vertical settling is also supported by the interpretation of excess Thorium -230
(230Th,) data measured on several sediment cores.
Some of these cores we also used for this
clay mineralogical study (PS 1768-8, PS2082-1,
PS2498-1, PS2499-5) (Frank et al. this volume). Although the 230Th ex method cannot distinguish between modes of biogenic and terrigenous particle
fluxes, a great part of bulk sediment redeposition
inferred from 230Thex measurements probably concerns the terrigenous fraction of the bulk sediment.
Looking atthese arguments, the substantial role
of deep water advection as a carrier and distributor ofterrigenous particles in the ACC region becomes clear. Nevertheless, we do not question the
importance of enhanced eolian influx to the ocean
during glacial times. But the sedimentological signals (clay mineralogy, grain-size distributions) recorded in most sediment cores of the ACC region
are a current signal. This means a signal that reflects bottom current transport and redeposition of
terrigenous particles delivered by wind, ice, and
turbidity currents in the vicinity of terrestrial source
areas.
295
Hofmann in press) and Antarctica (Petit et al.
1990; Grousset et al. 1992) during cold climatic
stages.
However, a significant influx ofterrigenous matter by winds of the southern westerlies seems to
be confined to the ocean basins offthe Patagonian
coast (Balsam et al. 1995). In the Indian sector of
the Southern Ocean only a maximum of 5% oftotal
detrital influx can be attributed to direct dust input
(Bareille et al. 1994).
Our sediment core studies in the ACC region,
which is characterized by its dense nepheloid suspension layers (Biscaye and Eittreim 1977), support the prominent role of deep water circulation
in the distribution of terrigenous particles, also during glacial periods. It is true that, in the same region, a higher input of iron-bearing terrigenous dust
during glacials may fertilize surface-water masses
and so enhances biological production, export production and terrigenous particle fluxes to the sea
floor (Kumar et al. 1995). However, the settlement
of wind-borne particles from the surface through
the water column contributes only one part of the
total particle supply to the sea-floor and cannot
solely explain strong fluxes ofterrigenous matter
at the sea-floor. Several lines of evidence argue for
significant lateral near-bottom current transport of
sedimentary particles rather than primary long-distance eolian influx to the pelagic ACC region during glacial times.
A first approach is the comparison of AR tw -
values in the ACC region with ARt",-values at
sites in the equatorial Atlantic, where eolian influx
of terrigenous matter from the Sahara and Sahel
zone predominates (Tiedemann et al. 1989;
Ruddiman 1997). The sediment core locations on
submarine topographic highs in the ACC region, as
for instance those atthe Mid-Atlantic Ridge, should
provide the best record of eolian sedimentation.
Actually, ARtw-values of the Last Glacial Maximum documented in PS2495-3, PS2498-1, and
PS2499-1 from the Mid-Atlantic Ridge (Al\,,,values: 2-11 g cm· 2 ka·t)are higher and more variable than at sites in the equatorial Atlantic (AR t ,,,values: <2.0 g cm· 2 ka-') with similar distances to
the source areas of continental dust. Even if considering an enlargement of the Patagonian deflation area as owing to low stands of sea-level and
shelf exposure, this modified area size cannot be
compared to the size of the wide arid regions of
Northern Africa and thus the great potential sediment yield from there.
AR terr -values in the ACC region differ markedly
between the investigated sites. Especially the three
sediment cores from the Mid-Atlantic Ridge reveal
dramatic shifts of ARtw-valueswithin relatively
short distance (3 0 latitude) that rule out an uniform
eolian-controlled pelagic sedimentation (Fig. 4).
Glacial AR -values in the northernmost sediment
lerr
core (PS2495-3) is half a magnitude lower
than in PS2498-1 and PS2499-5 south of that
position. Moreover, grain sizes ofterrigenous silt
in PS2495-3 are smaller than in the other sediment
cores and may be explained by lesser current velocities at the position of PS2495-3 that lies in the
marginal jets of the ACC (Diekmann and Kuhn
1997). Variations of ARtw-values in a similar manner as the grain size parameters are consistent with
current-controlled sedimentation.
The conclusion that a lot of bulk particle fluxes
in the ACC region may be attributed to lateral sediment transport and reworking through winnowing
and focusing rather than vertical settling is also supported by the interpretation of excess Thorium -230
(230Th,) data measured on several sediment cores.
Some of these cores we also used for this
clay mineralogical study (PS 1768-8, PS2082-1,
PS2498-1, PS2499-5) (Frank et al. this volume). Although the 230Th ex method cannot distinguish between modes of biogenic and terrigenous particle
fluxes, a great part of bulk sediment redeposition
inferred from 230Thex measurements probably concerns the terrigenous fraction of the bulk sediment.
Looking atthese arguments, the substantial role
of deep water advection as a carrier and distributor ofterrigenous particles in the ACC region becomes clear. Nevertheless, we do not question the
importance of enhanced eolian influx to the ocean
during glacial times. But the sedimentological signals (clay mineralogy, grain-size distributions) recorded in most sediment cores of the ACC region
are a current signal. This means a signal that reflects bottom current transport and redeposition of
terrigenous particles delivered by wind, ice, and
turbidity currents in the vicinity of terrestrial source
areas.