462
HolmesetaL
although their data did not discount the possibility
that there may have been no change in [NOl'l.
Contrarily, Bertrand et aL (1997) suggested thatthe
high 8 15 N values they found in the deglacial section of cores from the northwestern African margin were related not to increased denitrification, but
to changes in the ratio of nutrients exported to the
deep ocean to those recycled and upwelled in the
upper ocean which resulted from sea level rise.
Farrell et aL (1995) and Nakatsuka et aL (1995a)
concluded that changes in 15N/ 14 N ratios in their
cores (in the equatorial and western equatorial
Pacific, respectively) did not result from changes
in denitrification. Instead, they determined that
decreased relative nitrate utilization, caused by
stronger upwelling, led to low8 15 N. Similarly, sedimentary 8 15 N in the Bering Sea responded to
changes in inorganic nitrogen levels and consequent
relative nutrient utilization caused by changes in
vertical mixing of nutrients into surface water
(Nakatsuka et aL 1995b). Likewise, the 8 15 N signal in our cores is explainable by changes in relative nitrate utilization and does not suggest that a
change in global 8 15 N-NO J ' was an important influence.
Mean glacial 8 15 N was between 0.5 and 0.7%0
lighter, on average, than interglacial 8 1S N, suggesting elevated glacial nutrient levels. This corroborates other evidence for elevated glacial productivity in the Angola Basin, provided by organic carbon content and foraminiferal OIlC (Schneider
1991; Schneider et aL 1994), organic carbon accumulation rates (Bornhold 1973) and fossil diatom
assemblages (Pokras 1987).
The 8 15 N records are characterized by a 23 kyr
cyclicity (Fig. 12), indicating that precessional forcing controls nitrate availability in this region. When
the earth-sun distance in boreal summer is greatest, monsoonal intensity decreases, leading to an
increase in trade wind intensity. Trade winds, in
tum, drive the coastal upwelling which brings nutrient-rich water from depth to the surface, and
allow intensified northward advection of nitraterich waters into the Angola Basin, resulting in less
extensive depletion of the available nitrate. There
is also significant power in the 41 kyr obliquity period in the 8 15 N profiles. The presence of a 41 kyr
signal demonstrates that although precessional forcing appears to dominate nitrate availability and relative utilization, boreal climate changes also have an
affect on this low-latitude region.
Terrestrial organic matter in core GeoB 10163 is negligible, while core GeoB 1008-3, in the Zaire Fan, is influenced more strongly by the presence
of continental-derived material (MUller et aL 1994;
Holmes et aL 1996), as seen by the 8 11 C o'g offset
between these two cores (Fig. 11 a). Terrestrial
organic carbon is depleted in llC relative to marine organic carbon, suggesting that the offset of
about -1.5%0 offset ofGeoB 1008-3 sediments from
GeoB 1016-3 results from higher terrestrial loading of these sediments. The 8 13 C mg fluctuations in
core GeoB 1016-3 which are paralleled by core
GeoB 1008-3 were interpreted by MUller et al.
(1994) to be the result of changes in [C0 2 l aq in
surface water. A similar offset in the8 15 N profiles
from the two cores is conspicuously lacking (Fig.
11 b), suggesting that terrestrial matter entering the
Zaire Fan from the Zaire River has not noticeably
altered the 8 1S N of the sediments at site GeoB
1008-3 over the past 180 kyr. The absence of a
clear terrestrial 8 1S N signal can be explained by the
low concentrations of terrigenous organic matter
at this site (which is more than 200 km from the
coast), the low % nitrogen in terrestrial organic
matter and the apparent similarity ofthe presentday nitrogen isotopic signal of organic material
derived from the Zaire River to marine-derived
organic matter in this region (Fig. 9a).
Shifts in the position of the ABF can be surmised
by looking at the differences between cores GeoB
1008-3 and 1016-3. 8 15 N was lower in core GeoB
1016-3 than in 1008-3 throughout stage 4 and late
in stage 3 (Fig. II), indicating that the ABF was
then positioned between these two sites and that
the presence of nutrient-rich Benguela Current
water led to decreased 15N/l4N ratios at GeoB
1016-3 relative to 1008-3. Schneider et aL (1995)
and Jansen et aL (1996) reported similar findings
based on alkenone-SST and planktic foraminifera
data from sediments of the Angola Basin.
HolmesetaL
although their data did not discount the possibility
that there may have been no change in [NOl'l.
Contrarily, Bertrand et aL (1997) suggested thatthe
high 8 15 N values they found in the deglacial section of cores from the northwestern African margin were related not to increased denitrification, but
to changes in the ratio of nutrients exported to the
deep ocean to those recycled and upwelled in the
upper ocean which resulted from sea level rise.
Farrell et aL (1995) and Nakatsuka et aL (1995a)
concluded that changes in 15N/ 14 N ratios in their
cores (in the equatorial and western equatorial
Pacific, respectively) did not result from changes
in denitrification. Instead, they determined that
decreased relative nitrate utilization, caused by
stronger upwelling, led to low8 15 N. Similarly, sedimentary 8 15 N in the Bering Sea responded to
changes in inorganic nitrogen levels and consequent
relative nutrient utilization caused by changes in
vertical mixing of nutrients into surface water
(Nakatsuka et aL 1995b). Likewise, the 8 15 N signal in our cores is explainable by changes in relative nitrate utilization and does not suggest that a
change in global 8 15 N-NO J ' was an important influence.
Mean glacial 8 15 N was between 0.5 and 0.7%0
lighter, on average, than interglacial 8 1S N, suggesting elevated glacial nutrient levels. This corroborates other evidence for elevated glacial productivity in the Angola Basin, provided by organic carbon content and foraminiferal OIlC (Schneider
1991; Schneider et aL 1994), organic carbon accumulation rates (Bornhold 1973) and fossil diatom
assemblages (Pokras 1987).
The 8 15 N records are characterized by a 23 kyr
cyclicity (Fig. 12), indicating that precessional forcing controls nitrate availability in this region. When
the earth-sun distance in boreal summer is greatest, monsoonal intensity decreases, leading to an
increase in trade wind intensity. Trade winds, in
tum, drive the coastal upwelling which brings nutrient-rich water from depth to the surface, and
allow intensified northward advection of nitraterich waters into the Angola Basin, resulting in less
extensive depletion of the available nitrate. There
is also significant power in the 41 kyr obliquity period in the 8 15 N profiles. The presence of a 41 kyr
signal demonstrates that although precessional forcing appears to dominate nitrate availability and relative utilization, boreal climate changes also have an
affect on this low-latitude region.
Terrestrial organic matter in core GeoB 10163 is negligible, while core GeoB 1008-3, in the Zaire Fan, is influenced more strongly by the presence
of continental-derived material (MUller et aL 1994;
Holmes et aL 1996), as seen by the 8 11 C o'g offset
between these two cores (Fig. 11 a). Terrestrial
organic carbon is depleted in llC relative to marine organic carbon, suggesting that the offset of
about -1.5%0 offset ofGeoB 1008-3 sediments from
GeoB 1016-3 results from higher terrestrial loading of these sediments. The 8 13 C mg fluctuations in
core GeoB 1016-3 which are paralleled by core
GeoB 1008-3 were interpreted by MUller et al.
(1994) to be the result of changes in [C0 2 l aq in
surface water. A similar offset in the8 15 N profiles
from the two cores is conspicuously lacking (Fig.
11 b), suggesting that terrestrial matter entering the
Zaire Fan from the Zaire River has not noticeably
altered the 8 1S N of the sediments at site GeoB
1008-3 over the past 180 kyr. The absence of a
clear terrestrial 8 1S N signal can be explained by the
low concentrations of terrigenous organic matter
at this site (which is more than 200 km from the
coast), the low % nitrogen in terrestrial organic
matter and the apparent similarity ofthe presentday nitrogen isotopic signal of organic material
derived from the Zaire River to marine-derived
organic matter in this region (Fig. 9a).
Shifts in the position of the ABF can be surmised
by looking at the differences between cores GeoB
1008-3 and 1016-3. 8 15 N was lower in core GeoB
1016-3 than in 1008-3 throughout stage 4 and late
in stage 3 (Fig. II), indicating that the ABF was
then positioned between these two sites and that
the presence of nutrient-rich Benguela Current
water led to decreased 15N/l4N ratios at GeoB
1016-3 relative to 1008-3. Schneider et aL (1995)
and Jansen et aL (1996) reported similar findings
based on alkenone-SST and planktic foraminifera
data from sediments of the Angola Basin.