88
based on diagenetic models. Such models are generally not an alternative to direct rate measurements, mainly due to limitations in the mathematical description of transport through bioturbation.
In combination with direct measurements, however, models are an extremely useful tool for, for
example, verifying experimental results, extracting
information on processes that cannot be measured
directly, and testing hypothetical scenarios (Berner
1980; Wang and Van Cappellen 1996; Boudreau
1997).
Total Benthic Mineralization,
Flux Measurements
The earliest and still most widely used measure of
total benthic mineralization is the oxygen (02) uptake of the sediment (Pamatmat 1977; Smith and
Hinga 1983). Its use, even in sediments where anaerobic respiration predominates, is based on the
general observation that in sediments underlying
well-oxygenated waters, reduced inorganic metabolites are largely reoxidized within the sediment,
and that O 2 is therefore the ultimate acceptor of
nearly all electron equivalents from organic carbon
(C) (see Fig. 6.1). The nitrogen gas (N 2 ) formed
during denitrification is an obvious exception,
which, however, only accounts for a major fraction
of mineralization in sediments that contact bottom
water with a low oxygen/nitrate (021N0;) ratio
(see later section on nitrate reduction). In sediments
with methanogenesis at shallow depth, electron
equivalents may also escape the sediment by methane (C~) ebullition (Martens and Klump 1980).
Benthic fluxes of reduced manganese (Mn 2 +), iron
(Fe 2 +), and hydrogen sulfide (H 2 S) are insignificant as long as O2 is not depleted (Balzer 1982;
Hall et al. 1989; Thamdrup et al. 1994b). Sulfide
burial, mainly in the form of pyrite, is the largest
sink for electron equivalents from C oxidation next
to O 2 consumption in most sediments, but typically
represents 10% or less of C mineralization (Jf/lrgensen 1982). The minor sinks listed here can all be
individually quantified and added to the O 2 consumption for a more accurate determination of benthic C mineralization. The O2 uptake should further
be corrected for the contribution from nitrification,
which can be estimated as the sum of denitrification
Bo Thamdrup and Donald E. Canfield
and NO; efflux from the sediment (e.g., Canfield
et al. 1993a). During the aerobic mineralization of
typical, "Redfieldian" organic matter, the oxygen
demand from nitrification makes up about 20% of
the total (Anderson and Sarmiento 1994).
An important uncertainty in using O2 uptake as
a measure of mineralization lies in the assumption
that anaerobic respiration and reoxidation of inorganic species are at steady state. Most reduced Mn,
Fe, and sulfur (S) is in particulate form, and sediment reworking, for example, by bioturbation, is
crucial for its reoxidation. The nature of particle
transport in sediments is poorly understood, and it
is not clear whether reoxidation occurs in pulses or
continuously relative to the time scale of O2 uptake
measurements. On monthly or seasonal time scales
the inventory of reduced inorganic compounds in a
sediment may vary in response to changes in supply
of organic substrate or oxidants (e.g., Luther et al.
1991; Thamdrup et al. 1994a), which implies a temporal uncoupling of anaerobic mineralization and
reoxidation. In special cases, reoxidation may be
dominated by discrete events such as wind or
current-induced resuspension, so that the benthic
O 2 uptake during calm periods may underestimate
total mineralization (Aller et al. 1996; Jf/lrgensen
1996).
The uncertainties related to reoxidation are
avoided if the benthic flux of dissolved inorganic
C, ~C02' rather than O2 is used as a measure of C
mineralization (Teal and Kanwisher 1961; Hargrave and Phillips 1981). Dissolved inorganic C is
the immediate product of all mineralization pathways, and ~C02 distributions respond rapidly to
changes in production rates. However, roughly a
10-fold higher precision is required for ~C02 compared with O2 analysis due to a higher background
concentration. Also, ~C02 fluxes may be affected
by carbonate dissolution or precipitation (Hammond et al. 1985; Anderson et al. 1986), though
this contribution has rarely been quantified in shelf
sediments. In active coastal sediments, calcium carbonate (CaC0 3 ) dissolution has been estimated to
account for approximately 15% of the ~C02 flux
as an annual average (Hammond et al. 1985; Jf/lrgensen 1996). The dissolution of foraminiferal tests
was estimated to be >30% of ~C02 fluxes from
Long Island Sound sediments during some seasons,
while at other times net precipitation occurred
(Green et al. 1993). On the continental slope,
based on diagenetic models. Such models are generally not an alternative to direct rate measurements, mainly due to limitations in the mathematical description of transport through bioturbation.
In combination with direct measurements, however, models are an extremely useful tool for, for
example, verifying experimental results, extracting
information on processes that cannot be measured
directly, and testing hypothetical scenarios (Berner
1980; Wang and Van Cappellen 1996; Boudreau
1997).
Total Benthic Mineralization,
Flux Measurements
The earliest and still most widely used measure of
total benthic mineralization is the oxygen (02) uptake of the sediment (Pamatmat 1977; Smith and
Hinga 1983). Its use, even in sediments where anaerobic respiration predominates, is based on the
general observation that in sediments underlying
well-oxygenated waters, reduced inorganic metabolites are largely reoxidized within the sediment,
and that O 2 is therefore the ultimate acceptor of
nearly all electron equivalents from organic carbon
(C) (see Fig. 6.1). The nitrogen gas (N 2 ) formed
during denitrification is an obvious exception,
which, however, only accounts for a major fraction
of mineralization in sediments that contact bottom
water with a low oxygen/nitrate (021N0;) ratio
(see later section on nitrate reduction). In sediments
with methanogenesis at shallow depth, electron
equivalents may also escape the sediment by methane (C~) ebullition (Martens and Klump 1980).
Benthic fluxes of reduced manganese (Mn 2 +), iron
(Fe 2 +), and hydrogen sulfide (H 2 S) are insignificant as long as O2 is not depleted (Balzer 1982;
Hall et al. 1989; Thamdrup et al. 1994b). Sulfide
burial, mainly in the form of pyrite, is the largest
sink for electron equivalents from C oxidation next
to O 2 consumption in most sediments, but typically
represents 10% or less of C mineralization (Jf/lrgensen 1982). The minor sinks listed here can all be
individually quantified and added to the O 2 consumption for a more accurate determination of benthic C mineralization. The O2 uptake should further
be corrected for the contribution from nitrification,
which can be estimated as the sum of denitrification
Bo Thamdrup and Donald E. Canfield
and NO; efflux from the sediment (e.g., Canfield
et al. 1993a). During the aerobic mineralization of
typical, "Redfieldian" organic matter, the oxygen
demand from nitrification makes up about 20% of
the total (Anderson and Sarmiento 1994).
An important uncertainty in using O2 uptake as
a measure of mineralization lies in the assumption
that anaerobic respiration and reoxidation of inorganic species are at steady state. Most reduced Mn,
Fe, and sulfur (S) is in particulate form, and sediment reworking, for example, by bioturbation, is
crucial for its reoxidation. The nature of particle
transport in sediments is poorly understood, and it
is not clear whether reoxidation occurs in pulses or
continuously relative to the time scale of O2 uptake
measurements. On monthly or seasonal time scales
the inventory of reduced inorganic compounds in a
sediment may vary in response to changes in supply
of organic substrate or oxidants (e.g., Luther et al.
1991; Thamdrup et al. 1994a), which implies a temporal uncoupling of anaerobic mineralization and
reoxidation. In special cases, reoxidation may be
dominated by discrete events such as wind or
current-induced resuspension, so that the benthic
O 2 uptake during calm periods may underestimate
total mineralization (Aller et al. 1996; Jf/lrgensen
1996).
The uncertainties related to reoxidation are
avoided if the benthic flux of dissolved inorganic
C, ~C02' rather than O2 is used as a measure of C
mineralization (Teal and Kanwisher 1961; Hargrave and Phillips 1981). Dissolved inorganic C is
the immediate product of all mineralization pathways, and ~C02 distributions respond rapidly to
changes in production rates. However, roughly a
10-fold higher precision is required for ~C02 compared with O2 analysis due to a higher background
concentration. Also, ~C02 fluxes may be affected
by carbonate dissolution or precipitation (Hammond et al. 1985; Anderson et al. 1986), though
this contribution has rarely been quantified in shelf
sediments. In active coastal sediments, calcium carbonate (CaC0 3 ) dissolution has been estimated to
account for approximately 15% of the ~C02 flux
as an annual average (Hammond et al. 1985; Jf/lrgensen 1996). The dissolution of foraminiferal tests
was estimated to be >30% of ~C02 fluxes from
Long Island Sound sediments during some seasons,
while at other times net precipitation occurred
(Green et al. 1993). On the continental slope,