448
Holmesetal.
confirming the relation between nitrate concentrations and isotopic values with plankton. Similarly,
sediment trap studies in the North Atlantic Ocean
and Sargasso Sea have shown a link between the
surface ocean and ol5N in POM sinking to the deep
sea (Altabet and Deuser 1985; Altabet et al. 1991).
The isotopic signature ofthe nitrate pool available for phytoplankton uptake is the foundation for
the nitrogen isotopic composition of organic matter. Wada et al. (1975) and Liu and Kaplan (1989)
reported a mean ol5N value for nitrate from
oxygenated water in the deep sea of 5-6%0.
Denitrification is one process that can alter theo l5 N
ofN0 3 -. Under oxygen-depleted conditions, the
reduction of N0 3 - leads to enrichment of the residual N0 3 - in 15N relative to the mean deep-sea
value because 14N0 3 - is more readily reduced by
bacteria than 15N03- (Cline and Kaplan 1975). This
potentially allows the detection of a denitrification
signal based on sedimentary o15N.
The interpretation of ol5N in organic matter in
the water column or in sediments is usually complicated by one or more factors that must be taken
into account. For example, the transfer of nitrogen
through trophic levels is characterized by a systematic increase in o15N, with each trophic step resulting in an enrichmentof2 - 4%0 (DeNiro and Epstein
1981; Minagawa and Wada 1984; Fry 1988;
Hansson et al. 1997). Schell et al. (1998) reported
that zooplankton ol5N reflected varying nutrient
levels and could also be used to determine whether
the zooplankton was herbivorous or carnivorous.
Fecal pellets are also isotopically enriched relative
to the food source (Checkley and Entzeroth 1985;
Montoya et al. 1992). Both of these effects appear
to lead to higher ol5N values in sinking particles
relative to POM suspended in the euphotic zone,
at least in oligotrophic systems (Altabet 1988). In
contrast, near-surface suspended and sinking particles in the North Atlantic exhibited nearly identical isotopic values (Altabetetal. 1991). Ostrom et
al. (1997) reported that seston and sinking POM
were isotopically most similar during the spring
bloom in a coastal cold-ocean embayment. These
results may indicate that in eutrophic regions, especially during blooms (which would be expected
to contribute the most to the sedimentary record),
sinking organic matter closely reflectso l5 N of the
primary production. Altabet (1988) also showed
that in oligotrophic regions, where there is complete
utilization of available N0 3 -, the meano l5 N of particles sinking out ofthe upper ocean should be equal
to the mean ol5N of the N0 3 - utilized by
phytoplankton. However, in eutrophic systems such
as the Benguela region, the sedimentary isotopic
record probably consists of the flux weighted average ol5N of both phyto- and (isotopically heavier)
zooplankton, as suggested by Montoya (1994). On
the other hand, since food chains are relatively short
(i.e. fewer trophic steps) where surface water nitrate concentrations are high (Biggs et al. 1989; Wu
et al. 1997), the increase in o15N_ POM due to food
web effects can be expected to be smaller than in
other types of systems.
The proportion of terrestrially-derived organic
matter present in the sediments may significantly
affect its ol5N signal. 15N/14N ratios in sediments
have been used to trace the origin of sedimentary
nitrogen in areas where the isotopic composition of
the terrigenous material was substantially depleted
in ISN relative to marine-derived matter (Sweeney
and Kaplan 1980; Wada et al. 1987). Other studies
in which ol5N has been used to document the differences between riverine and marine systems include Cifuentes et al. (1996) and McClelland et al.
(1997). Clear evidence for the mixing of a marine
with a terrestrial ol5N end member is not always
the case, however, due either to the masking of such
a trend by processes occurring in the water column
or to the wide range of ol5N values measured in
terrestrial systems (Sweeney et at. 1978; Owens
1985; Thornton and McManus 1994; Holmes et a\.
1996)_
Conflicting reports on the direction of depthrelated isotopic changes occurring in POM point
out the need for more research on this subject. An
enrichment in 14N with depth in sinking particles was
demonstrated by Altabet et al. (1991), Voss et al.
(1996), and Nakatsuka et at. (1997). However,
Saino and Hattori (1980) and Altabet and McCarthy
(1985) suggested that remineralization of organic
matter in sinking particles results in an enrichment
in ISN of the residual organic matter. Regardless
of the uncertainty as to the nature of the changes
particles in the water column undergo, sediments
appear to be consistently enriched in 1 5 N relative
Holmesetal.
confirming the relation between nitrate concentrations and isotopic values with plankton. Similarly,
sediment trap studies in the North Atlantic Ocean
and Sargasso Sea have shown a link between the
surface ocean and ol5N in POM sinking to the deep
sea (Altabet and Deuser 1985; Altabet et al. 1991).
The isotopic signature ofthe nitrate pool available for phytoplankton uptake is the foundation for
the nitrogen isotopic composition of organic matter. Wada et al. (1975) and Liu and Kaplan (1989)
reported a mean ol5N value for nitrate from
oxygenated water in the deep sea of 5-6%0.
Denitrification is one process that can alter theo l5 N
ofN0 3 -. Under oxygen-depleted conditions, the
reduction of N0 3 - leads to enrichment of the residual N0 3 - in 15N relative to the mean deep-sea
value because 14N0 3 - is more readily reduced by
bacteria than 15N03- (Cline and Kaplan 1975). This
potentially allows the detection of a denitrification
signal based on sedimentary o15N.
The interpretation of ol5N in organic matter in
the water column or in sediments is usually complicated by one or more factors that must be taken
into account. For example, the transfer of nitrogen
through trophic levels is characterized by a systematic increase in o15N, with each trophic step resulting in an enrichmentof2 - 4%0 (DeNiro and Epstein
1981; Minagawa and Wada 1984; Fry 1988;
Hansson et al. 1997). Schell et al. (1998) reported
that zooplankton ol5N reflected varying nutrient
levels and could also be used to determine whether
the zooplankton was herbivorous or carnivorous.
Fecal pellets are also isotopically enriched relative
to the food source (Checkley and Entzeroth 1985;
Montoya et al. 1992). Both of these effects appear
to lead to higher ol5N values in sinking particles
relative to POM suspended in the euphotic zone,
at least in oligotrophic systems (Altabet 1988). In
contrast, near-surface suspended and sinking particles in the North Atlantic exhibited nearly identical isotopic values (Altabetetal. 1991). Ostrom et
al. (1997) reported that seston and sinking POM
were isotopically most similar during the spring
bloom in a coastal cold-ocean embayment. These
results may indicate that in eutrophic regions, especially during blooms (which would be expected
to contribute the most to the sedimentary record),
sinking organic matter closely reflectso l5 N of the
primary production. Altabet (1988) also showed
that in oligotrophic regions, where there is complete
utilization of available N0 3 -, the meano l5 N of particles sinking out ofthe upper ocean should be equal
to the mean ol5N of the N0 3 - utilized by
phytoplankton. However, in eutrophic systems such
as the Benguela region, the sedimentary isotopic
record probably consists of the flux weighted average ol5N of both phyto- and (isotopically heavier)
zooplankton, as suggested by Montoya (1994). On
the other hand, since food chains are relatively short
(i.e. fewer trophic steps) where surface water nitrate concentrations are high (Biggs et al. 1989; Wu
et al. 1997), the increase in o15N_ POM due to food
web effects can be expected to be smaller than in
other types of systems.
The proportion of terrestrially-derived organic
matter present in the sediments may significantly
affect its ol5N signal. 15N/14N ratios in sediments
have been used to trace the origin of sedimentary
nitrogen in areas where the isotopic composition of
the terrigenous material was substantially depleted
in ISN relative to marine-derived matter (Sweeney
and Kaplan 1980; Wada et al. 1987). Other studies
in which ol5N has been used to document the differences between riverine and marine systems include Cifuentes et al. (1996) and McClelland et al.
(1997). Clear evidence for the mixing of a marine
with a terrestrial ol5N end member is not always
the case, however, due either to the masking of such
a trend by processes occurring in the water column
or to the wide range of ol5N values measured in
terrestrial systems (Sweeney et at. 1978; Owens
1985; Thornton and McManus 1994; Holmes et a\.
1996)_
Conflicting reports on the direction of depthrelated isotopic changes occurring in POM point
out the need for more research on this subject. An
enrichment in 14N with depth in sinking particles was
demonstrated by Altabet et al. (1991), Voss et al.
(1996), and Nakatsuka et at. (1997). However,
Saino and Hattori (1980) and Altabet and McCarthy
(1985) suggested that remineralization of organic
matter in sinking particles results in an enrichment
in ISN of the residual organic matter. Regardless
of the uncertainty as to the nature of the changes
particles in the water column undergo, sediments
appear to be consistently enriched in 1 5 N relative