5. The Measurement of Primary Production in Aquatic Ecosystems
75
of which comes directly from the products of the
light reaction. There is a similar enhancement of
the photosynthetic quotient when phytoplankton
produce large amounts of lipid or other highly reduced compounds.
Rather than measuring changes in O 2 in light and
dark bottles, changes in dissolved inorganic carbon
(DIC, or the sum of dissolved CO2 gas, HCO; and
CO~-) can be measured instead. However, in seawater and in freshwaters with significant carbonate
alkalinity, CO 2 is much more soluble than O2 and
DIC concentrations are very high. Thus, measuring
changes in DIC requires a large relative accuracy
and usually results in a much less sensitive method.
In lakes with low carbonate alkalinity, DIC concentrations can sometimes be less than O 2 concentrations, making changes in DIC a more sensitive
method.
Historically, the light and dark bottle method
could only be used in ecosystems where the rate of
GPP was reasonably high. The resolution of the
traditional Winkler titration for oxygen is of the
order 0.2 to 1 % in the surface ocean (0.4 to 2 /lmol
O 2 kg-I). In oligotrophic ecosystems, such as
many lakes and the majority of oceanic waters,
where the GPP rates over the range of the euphotic
zone are often only 1 to 10 /lg C liter - I day - I, the
oxygen change during the day is usually less than
1 /lmolliter - I. Traditional analytical techniques
for measuring oxygen concentrations simply lacked
the precision necessary to apply this technique in
oligotrophic, "blue" oceanic waters. However, during the past two decades, analytical techniques for
measuring oxygen have improved dramatically. Instruments for automated endpoint detection (e.g.,
Williams and Jenkinson 1982) have pushed the precision of the Winkler chemistry to 0.1 /lmolliter -I.
With the large amount of replication made relatively simple by rapid automated titrators, statistically significant comparisons of light and dark
treatments that vary by as little as 0.03 /lmol O2
liter- I are now possible. As a result, the light and
dark bottle technique could now be applied in systems with extremely low productivity, and this
measurement of GPP would be a powerful complement to the estimates from radiotracers and other
techniques.
There are a variety of problems with the light
and dark bottle technique that are associated with
scaling from volume to surface area, with scaling
in time, with sampling water, and with isolating a
parcel of water in a bottle during incubations. These
problems are all shared with the carbon-14 technique (presented next) and so are discussed in a
later section. A more fundamental problem with
applying this technique in open ocean ecosystems
is that the incubation bottles must be gasimpermeable. Also, the most common gas bottles
(BOD bottles) are extremely difficult to make
"trace-metal clean." Quartz bottles, which can be
appropriately cleaned, are extremely expensive.
This issue of the standard of trace-element cleanliness appropriate for oceanic ecosystems is discussed in more detail below.
Carbon-14 Technique
With the advent of readily available radioisotopes
following World War II, Steemen Nielsen (1952)
introduced the use of the carbon-14 e 4 C) method
for measuring primary production in planktonic
ecosystems. This approach increased the sensitivity
for measuring production by at least 1O-fold over
the light and dark bottle oxygen technique as it was
applied in the 1950s through the 1970s, and the 14C
method gained rapid favor. It remains today the
most common method for measuring primary production in both lakes and marine ecosystems, despite the fact that the light and dark bottle technique
has now been improved sufficiently so as to be sensitive enough in many oligotrophic ecosystems (see
above). The ease ofthe method, the high sensitivity,
and the ability to follow carbon through explicit
subcomponents of the ecosystem are all advantages
shared by no other method.
In the 14C method, water samples with ambient
plankton are placed in clear incubation bottles, and
a tracer quantity of 14C-Iabeled DIC is added (typically 1 to 20 /lCi liter-I). The samples are incubated, either in situ or in an incubator, as for the
light and dark bottle technique. Generally, at the
end of the incubation (typically 2 to 24 hours), the
water is filtered, the sample acidified to remove inorganic 14C and the 14C on the filter counted in a
scintillation counter. By measuring the ratio of 14C
to total inorganic carbon in the incubation, the rate
of accumulation of 14C in the particulate matter can
be converted to a rate of uptake of carbon.
75
of which comes directly from the products of the
light reaction. There is a similar enhancement of
the photosynthetic quotient when phytoplankton
produce large amounts of lipid or other highly reduced compounds.
Rather than measuring changes in O 2 in light and
dark bottles, changes in dissolved inorganic carbon
(DIC, or the sum of dissolved CO2 gas, HCO; and
CO~-) can be measured instead. However, in seawater and in freshwaters with significant carbonate
alkalinity, CO 2 is much more soluble than O2 and
DIC concentrations are very high. Thus, measuring
changes in DIC requires a large relative accuracy
and usually results in a much less sensitive method.
In lakes with low carbonate alkalinity, DIC concentrations can sometimes be less than O 2 concentrations, making changes in DIC a more sensitive
method.
Historically, the light and dark bottle method
could only be used in ecosystems where the rate of
GPP was reasonably high. The resolution of the
traditional Winkler titration for oxygen is of the
order 0.2 to 1 % in the surface ocean (0.4 to 2 /lmol
O 2 kg-I). In oligotrophic ecosystems, such as
many lakes and the majority of oceanic waters,
where the GPP rates over the range of the euphotic
zone are often only 1 to 10 /lg C liter - I day - I, the
oxygen change during the day is usually less than
1 /lmolliter - I. Traditional analytical techniques
for measuring oxygen concentrations simply lacked
the precision necessary to apply this technique in
oligotrophic, "blue" oceanic waters. However, during the past two decades, analytical techniques for
measuring oxygen have improved dramatically. Instruments for automated endpoint detection (e.g.,
Williams and Jenkinson 1982) have pushed the precision of the Winkler chemistry to 0.1 /lmolliter -I.
With the large amount of replication made relatively simple by rapid automated titrators, statistically significant comparisons of light and dark
treatments that vary by as little as 0.03 /lmol O2
liter- I are now possible. As a result, the light and
dark bottle technique could now be applied in systems with extremely low productivity, and this
measurement of GPP would be a powerful complement to the estimates from radiotracers and other
techniques.
There are a variety of problems with the light
and dark bottle technique that are associated with
scaling from volume to surface area, with scaling
in time, with sampling water, and with isolating a
parcel of water in a bottle during incubations. These
problems are all shared with the carbon-14 technique (presented next) and so are discussed in a
later section. A more fundamental problem with
applying this technique in open ocean ecosystems
is that the incubation bottles must be gasimpermeable. Also, the most common gas bottles
(BOD bottles) are extremely difficult to make
"trace-metal clean." Quartz bottles, which can be
appropriately cleaned, are extremely expensive.
This issue of the standard of trace-element cleanliness appropriate for oceanic ecosystems is discussed in more detail below.
Carbon-14 Technique
With the advent of readily available radioisotopes
following World War II, Steemen Nielsen (1952)
introduced the use of the carbon-14 e 4 C) method
for measuring primary production in planktonic
ecosystems. This approach increased the sensitivity
for measuring production by at least 1O-fold over
the light and dark bottle oxygen technique as it was
applied in the 1950s through the 1970s, and the 14C
method gained rapid favor. It remains today the
most common method for measuring primary production in both lakes and marine ecosystems, despite the fact that the light and dark bottle technique
has now been improved sufficiently so as to be sensitive enough in many oligotrophic ecosystems (see
above). The ease ofthe method, the high sensitivity,
and the ability to follow carbon through explicit
subcomponents of the ecosystem are all advantages
shared by no other method.
In the 14C method, water samples with ambient
plankton are placed in clear incubation bottles, and
a tracer quantity of 14C-Iabeled DIC is added (typically 1 to 20 /lCi liter-I). The samples are incubated, either in situ or in an incubator, as for the
light and dark bottle technique. Generally, at the
end of the incubation (typically 2 to 24 hours), the
water is filtered, the sample acidified to remove inorganic 14C and the 14C on the filter counted in a
scintillation counter. By measuring the ratio of 14C
to total inorganic carbon in the incubation, the rate
of accumulation of 14C in the particulate matter can
be converted to a rate of uptake of carbon.