Dinitrogen
fixation was estimated using the acetylene-reduction/ethylene-production
method (Stewart et al.
1967;
Hardy et al. 1986;
Larkum et al. 1988).
Short (4 h) and long-term (12-24 h) measurements were
performed, using single injections of acetylene. Dark N2
fixation was measured both during the night and during the day
by wrapping incubation bottles in aluminium foil and black
plastic.
|
|
Sediment samples 2 cm thick were placed
in an incubation vessel. The incubation vessel was a 500 ml glass jar
with a spring-loading Perspex lid and with a rubber injection
port (Larkum et al.
1988).
|
The
vessels contained a liquid volume of 100 ml and an air
volume of 400 ml. Experiments were started by the removal
of 50 ml air and the injection of 50 ml of C2H2,
followed by swirling for several minutes. Most of the
experiments were carried out in bottles incubated in floating
racks just below the sea surface (0.30 cm), to assure a
temperature similar to that in situ. Wind and wave action caused
gentle stirring. However, to estimate the influence of light,
some incubations were made at different depths. Gas samples were
taken from the incubation jars using 4 ml evacuated glass
sampling tubes, after swirling for several minutes.
Ethylene
production was assayed using a portable gas chromatograph with a
stainless steel column (Poropak T) and a sensitive thermal
conductivity detector for C2H4 detection. Gas (50-500 µl)
was injected, and the C4H4 peak was followed on a strip-chart
recorder. The system was calibrated with known dilutions of C2H4
after every ten analyses.
It
is important to calibrate the acetylene reduction method, using 15N measurements as carried out for each of the studied
communities. This was done by incubating benthic cyanobacterial
samples for 24 h under a culture chamber with continuous
light of 50 µE m-2 h-1 in the presence of
C2H2. After the C2H4
measurements, gaseous N2 was removed from sample flasks by
flushing the slurry with 100% He (Seitzinger and Garber
1987).
Fifty ml of He was then removed and replaced by 50 ml of 15N2. Each of the two replicate samples was then re-incubated
for 24h in the same conditions as before but with 20.2 Pa
of 98 at% 15N2. At the end of the incubation, a gas sample
was taken using 4 ml evacuated tubes, and the incubated
community was frozen immediately. The incubation water and
organisms were digested by Kjeldahl procedures, and the ammonia
was steam-distilled into boric acid and titrated with dilute HCl
as described by Bergersen (1980).
Micro-diffusion techniques were used for recovery of the
dissolved inorganic and organic nitrogen (Slawyk and Raimbault
1995).
15N measurements were made on a tracer mass spectrometer (Europa
Cientific).
Atoll_site_