308
Diekmann et al.
Our results are sound with common concepts
of paleo-deep water advection suggesting a
shallowing and northward retreat of northernsource deep water and wider expansion of
southern-source deep water masses during glacial
times relative to interglacial times. The kaolinite/
chlorite proxy applies well in the Agulhas Basin of
the southeastern Atlantic sector of the Southern
Ocean where variations of the kaolinite/chloriteratio in the marine sediments show greatest latitudinal and temporal gradients. This region is situated
beyond significant dust supply. Furthermore, meridional shifts of deep water mass extensions within
the jets of the ACC are not hampered by topographic highs as observed in the western South
Atlantic.
From the kaolinite/chlorite proxy we cannot
decipher, which water mass is responsible for controlling water mass configurations. Time-series
analyses and phase relationships between clay
mineral parameters and other climate and NADW
proxies do not permit an unequivocal interpretation
of the timing and the response of the northern- and
southern-source deep water masses to circulation
and climate changes. Shifts of water mass boundaries are caused by variations of deep water mass
production rates in both northern and southern polar regions. It is generally accepted that the production of northern-source deep water was reduced during glacial stages. In turn, not much is
known about the behaviour of southern-source
deep and bottom water mass production. In the
western South Atlantic and the adjoining Southern
Ocean, grain-size properties of silt of the terrigenous sediment fraction are used as a proxy of relative paleo-bottom water current strengths that may
reflect the volume transport and production rates
of southern-source bottom water masses (e.g.
Ledbetter 1986; Pudsey 1992; Diekmann and Kuhn
1997). These grain-size studies, however, yielded
no consistent timing of maximal relative velocities
of the bottom currents at the various sites
(Diekmann and Kuhn 1997).
Refinements of our clay mineralogical approach
are desired. The validity of the kaolinite/chlorite
proxy has to be confirmed and supported by computer-aided modelling. An ocean circulation model
(Schlitzer 1996) could simulate particle fluxes and
transport directions within distinct water masses,
taking sources and settling velocities of terrigenous
particles into account.
A great effort has to be focused on the qualitative and quantitative determination of primary eolian
components in marine sediments of the Southern
Ocean that potentially mask the deep water circulation signal. In sediment cores of the Scotia Sea
(Hofmann in press), and less pronounced in sediment cores farther to the east used in this study
(Kuhn, unpublished), signatures of magnetic susceptibility coincide with the dust record of the
Vostok ice core (Petit et al. 1990). The susceptibility signal most likely originates from
titanomagnetite-rich dust particles derived from
Patagonia (Hofmann in press). Although our sediment core studies presented in this paper do not
support a significant primary dust supply to Southern ocean sediments of the Atlantic sector, the
observed patterns of magnetic susceptibility
suggest that at least a small proportion of the
terrigenous sediment fraction must originate from
eolian particles. Therefore, an important requirement is to resolve coherencies and differences in
records of magnetic susceptibility with mineralogical and geochemical data. Such investigations
would help to clarify whether the clay mineral signal could be identical with a wind signal.
In future, the kaolinite/chlorite proxy will be
applied on sediment cores with high temporal resolution, in order to obtain new insights into the relationship between rapid climate changes and variations of deep water mass advection on the southern hemisphere that seemed to lag rapid climate
changes during the last 80 ka (Charles et al. 1996).
Furthermore, the kaolinite/chlorite proxy will be
used to reconstruct long-term changes of Pliocene
to Pleistocene deep water circulation. Sediment
cores representing temporal resolutions on both
short-term and long-term time scales were obtained
during Leg 177 of the 'Ocean Drilling Program'
from the southeastern Atlantic sector ofthe Southern Ocean (Gersonde and Hodell 1997), the most
suitable region for the application of the kaolinite/
chlorite proxy.
Diekmann et al.
Our results are sound with common concepts
of paleo-deep water advection suggesting a
shallowing and northward retreat of northernsource deep water and wider expansion of
southern-source deep water masses during glacial
times relative to interglacial times. The kaolinite/
chlorite proxy applies well in the Agulhas Basin of
the southeastern Atlantic sector of the Southern
Ocean where variations of the kaolinite/chloriteratio in the marine sediments show greatest latitudinal and temporal gradients. This region is situated
beyond significant dust supply. Furthermore, meridional shifts of deep water mass extensions within
the jets of the ACC are not hampered by topographic highs as observed in the western South
Atlantic.
From the kaolinite/chlorite proxy we cannot
decipher, which water mass is responsible for controlling water mass configurations. Time-series
analyses and phase relationships between clay
mineral parameters and other climate and NADW
proxies do not permit an unequivocal interpretation
of the timing and the response of the northern- and
southern-source deep water masses to circulation
and climate changes. Shifts of water mass boundaries are caused by variations of deep water mass
production rates in both northern and southern polar regions. It is generally accepted that the production of northern-source deep water was reduced during glacial stages. In turn, not much is
known about the behaviour of southern-source
deep and bottom water mass production. In the
western South Atlantic and the adjoining Southern
Ocean, grain-size properties of silt of the terrigenous sediment fraction are used as a proxy of relative paleo-bottom water current strengths that may
reflect the volume transport and production rates
of southern-source bottom water masses (e.g.
Ledbetter 1986; Pudsey 1992; Diekmann and Kuhn
1997). These grain-size studies, however, yielded
no consistent timing of maximal relative velocities
of the bottom currents at the various sites
(Diekmann and Kuhn 1997).
Refinements of our clay mineralogical approach
are desired. The validity of the kaolinite/chlorite
proxy has to be confirmed and supported by computer-aided modelling. An ocean circulation model
(Schlitzer 1996) could simulate particle fluxes and
transport directions within distinct water masses,
taking sources and settling velocities of terrigenous
particles into account.
A great effort has to be focused on the qualitative and quantitative determination of primary eolian
components in marine sediments of the Southern
Ocean that potentially mask the deep water circulation signal. In sediment cores of the Scotia Sea
(Hofmann in press), and less pronounced in sediment cores farther to the east used in this study
(Kuhn, unpublished), signatures of magnetic susceptibility coincide with the dust record of the
Vostok ice core (Petit et al. 1990). The susceptibility signal most likely originates from
titanomagnetite-rich dust particles derived from
Patagonia (Hofmann in press). Although our sediment core studies presented in this paper do not
support a significant primary dust supply to Southern ocean sediments of the Atlantic sector, the
observed patterns of magnetic susceptibility
suggest that at least a small proportion of the
terrigenous sediment fraction must originate from
eolian particles. Therefore, an important requirement is to resolve coherencies and differences in
records of magnetic susceptibility with mineralogical and geochemical data. Such investigations
would help to clarify whether the clay mineral signal could be identical with a wind signal.
In future, the kaolinite/chlorite proxy will be
applied on sediment cores with high temporal resolution, in order to obtain new insights into the relationship between rapid climate changes and variations of deep water mass advection on the southern hemisphere that seemed to lag rapid climate
changes during the last 80 ka (Charles et al. 1996).
Furthermore, the kaolinite/chlorite proxy will be
used to reconstruct long-term changes of Pliocene
to Pleistocene deep water circulation. Sediment
cores representing temporal resolutions on both
short-term and long-term time scales were obtained
during Leg 177 of the 'Ocean Drilling Program'
from the southeastern Atlantic sector ofthe Southern Ocean (Gersonde and Hodell 1997), the most
suitable region for the application of the kaolinite/
chlorite proxy.