Reconstruction of Surface Ocean Nitrate Utilization
449
to surface ocean organic matter. Francois et al.
(1992) reported an enrichment in 'SN of5 to 7%0 in
organic matter at the sediment-water interface
compared to POM in the water column in the Southern Ocean. Similarly, Altabet (1996) found that
equatorial Pacific sediments were 4%0 heavier than
particles collected in sediment traps. Although the
isotopic composition of organic matter may be altered between its production in the euphotic zone
and burial at the sea floor, the change appears to
be consistent within a given area, and sedimentary
1)15N in the regions investigated to date reflects
surface ocean processes in at least a relative way
(Altabet and Francois 1994).
Several recent publications demonstrate the importance of sedimentary nitrogen isotopic ratios in
paleoceanography. Minoura et al. (1997) found that
surface productivity and sea level changes controlled sedimentary 1)15N changes in the Japan Sea.
1) 15 N in marine shales was used by Calvert et al.
(1996) to draw conclusions about variations in
nutrient drawdown and terrestrial input. Francois
et al. (1997) demonstrated that 1),sN and other
geochemical proxies in the sedimentary record from
the Southern Ocean provide evidence that surfacewater stratification south of the Polar Front contributed to a great extent to the glacial decrease in
atmospheric CO 2 . Here we will present results of
degradation experiments which give insight into the
mechanisms behind '5N-enrichment in decomposing particles. We will also show that1)'SN in sinking particles and surface sediments in the southeast
Atlantic reflect surface nutrient utilization in the
upper ocean, and then discuss our interpretation of
the 1)'SN records of two sediment cores from the
same region.
Materials and Methods
Sinking particles were collected during a sediment
trap experiment (WR-2) at the Walvis Ridge
(20 0 06'S, 09°11 'E) approximately 400 km offshore
(Fig. 1). The water depth was 2196 m and the trap
mooring was located 599 m below the water surface. The cone-shaped sediment trap consisted of
20 cup collectors and a collection area, with a grid
fitted to the top, of 0.5 m 2 (SMT 230; Salzgitter
Electronics, Kiel) and was deployed from 18 March
1989 to 13 March 1990. Samples (each representing an 18 day collection period) were preserved
with a 1 ml saturated solution of HgCl 2 in 100 ml
sea water (Fischer and Wefer 1991). NaCI was also
added to the cups to increase the salinity by 3%0
relative to the surrounding sea water. After recovery of the trap samples, 0.5 ml of concentrated
HgCl 2 solution was added per 100 ml trap solution.
For nitrogen isotopic analyses, fractions of sediment
trap samples (see Fischer and Wefer 1991 for details regarding splitting procedures) were filtered
onto precombusted GF IF filters. After filtration, the
samples were dried in a drying oven at 50°C overnight.
Surface sediment samples from the Angola
Basin were obtained during "METEOR" cruise
6/6 (Wefer et al. 1988). Cape Basin surface
sediments were retrieved during "METEOR"
cruises 6/6 and 20/2 (Wefer et al. 1988, 1992). All
surface sediments were collected using a box corer
(50 x 50 cm surface area) and subsamples were
taken immediately after core retrieval using plastic
syringes which were sealed with electrical tape and
stored at 4°C until further processing. Seven surface sediment transects were taken - three in the
Angola Basin and four in the Cape Basin (Fig. I).
Samples were freeze-dried and homogenized in an
agate mortar in a shore-based laboratory, then
stored in plastic containers until nitrogen isotope
analysis. Gravity cores (GeoB 1008-3 and GeoB
1016-3) were retrieved during "METEOR" cruise
6/6 (Wefer et al. 1988). GeoB 1008-3 was located
at6°35'S, 10 0 19'E in3124m water depth and GeoB
1016-3 at 11°46'S, 11 °41 'Eand3411 m water depth
(Fig. 1). Subsamples were taken every 5 cm and
treated as described for surface sediment samples.
Due to the relatively recent development of the
capability of measuring the nitrogen isotopic composition in samples of low nitrogen content, some
discussion of the analytical procedures involved is
warranted. One technique for measuring 1)'SN in
organic matter is the conversion of the organic nitrogen to ammonium sulfate by Kjeldahl digestion
and measuring the purified N2 gas obtained by oxidation (Bremner and Keeney 1966).1)'SN analyses can also be done on N2 gas obtained by combustion, which has been found to more completely
convert sample nitrogen to N2 (Minagawa et al.
449
to surface ocean organic matter. Francois et al.
(1992) reported an enrichment in 'SN of5 to 7%0 in
organic matter at the sediment-water interface
compared to POM in the water column in the Southern Ocean. Similarly, Altabet (1996) found that
equatorial Pacific sediments were 4%0 heavier than
particles collected in sediment traps. Although the
isotopic composition of organic matter may be altered between its production in the euphotic zone
and burial at the sea floor, the change appears to
be consistent within a given area, and sedimentary
1)15N in the regions investigated to date reflects
surface ocean processes in at least a relative way
(Altabet and Francois 1994).
Several recent publications demonstrate the importance of sedimentary nitrogen isotopic ratios in
paleoceanography. Minoura et al. (1997) found that
surface productivity and sea level changes controlled sedimentary 1)15N changes in the Japan Sea.
1) 15 N in marine shales was used by Calvert et al.
(1996) to draw conclusions about variations in
nutrient drawdown and terrestrial input. Francois
et al. (1997) demonstrated that 1),sN and other
geochemical proxies in the sedimentary record from
the Southern Ocean provide evidence that surfacewater stratification south of the Polar Front contributed to a great extent to the glacial decrease in
atmospheric CO 2 . Here we will present results of
degradation experiments which give insight into the
mechanisms behind '5N-enrichment in decomposing particles. We will also show that1)'SN in sinking particles and surface sediments in the southeast
Atlantic reflect surface nutrient utilization in the
upper ocean, and then discuss our interpretation of
the 1)'SN records of two sediment cores from the
same region.
Materials and Methods
Sinking particles were collected during a sediment
trap experiment (WR-2) at the Walvis Ridge
(20 0 06'S, 09°11 'E) approximately 400 km offshore
(Fig. 1). The water depth was 2196 m and the trap
mooring was located 599 m below the water surface. The cone-shaped sediment trap consisted of
20 cup collectors and a collection area, with a grid
fitted to the top, of 0.5 m 2 (SMT 230; Salzgitter
Electronics, Kiel) and was deployed from 18 March
1989 to 13 March 1990. Samples (each representing an 18 day collection period) were preserved
with a 1 ml saturated solution of HgCl 2 in 100 ml
sea water (Fischer and Wefer 1991). NaCI was also
added to the cups to increase the salinity by 3%0
relative to the surrounding sea water. After recovery of the trap samples, 0.5 ml of concentrated
HgCl 2 solution was added per 100 ml trap solution.
For nitrogen isotopic analyses, fractions of sediment
trap samples (see Fischer and Wefer 1991 for details regarding splitting procedures) were filtered
onto precombusted GF IF filters. After filtration, the
samples were dried in a drying oven at 50°C overnight.
Surface sediment samples from the Angola
Basin were obtained during "METEOR" cruise
6/6 (Wefer et al. 1988). Cape Basin surface
sediments were retrieved during "METEOR"
cruises 6/6 and 20/2 (Wefer et al. 1988, 1992). All
surface sediments were collected using a box corer
(50 x 50 cm surface area) and subsamples were
taken immediately after core retrieval using plastic
syringes which were sealed with electrical tape and
stored at 4°C until further processing. Seven surface sediment transects were taken - three in the
Angola Basin and four in the Cape Basin (Fig. I).
Samples were freeze-dried and homogenized in an
agate mortar in a shore-based laboratory, then
stored in plastic containers until nitrogen isotope
analysis. Gravity cores (GeoB 1008-3 and GeoB
1016-3) were retrieved during "METEOR" cruise
6/6 (Wefer et al. 1988). GeoB 1008-3 was located
at6°35'S, 10 0 19'E in3124m water depth and GeoB
1016-3 at 11°46'S, 11 °41 'Eand3411 m water depth
(Fig. 1). Subsamples were taken every 5 cm and
treated as described for surface sediment samples.
Due to the relatively recent development of the
capability of measuring the nitrogen isotopic composition in samples of low nitrogen content, some
discussion of the analytical procedures involved is
warranted. One technique for measuring 1)'SN in
organic matter is the conversion of the organic nitrogen to ammonium sulfate by Kjeldahl digestion
and measuring the purified N2 gas obtained by oxidation (Bremner and Keeney 1966).1)'SN analyses can also be done on N2 gas obtained by combustion, which has been found to more completely
convert sample nitrogen to N2 (Minagawa et al.