The gas sample was injected into a Hewlett-Packard 5880 gas chromatograph equipped
with dual thermal conductivity detectors. The analytical column was a 2.75 m x 3 mm
stainless steel column packed with molecular sieve 5A. Argon carrier gas was used at a
flow rate of 30 ml min' 1 . Operating temperatures were (°C); injection port, 35; oven,
35 ; detector, 250.
The total hydrogen in a vial was calculated using the Bunsen solubility coefficients of
Wiesenburg and Guinasso (1979) and corrected for any gas removed during previous
analyses. The value for total hydrogen was converted to hydrogen per liter of original
unconcentrated seawater, and unless otherwise noted, results expressed as nl H
2
l -1 are on
the basis of a liter of unconcentrated seawater.
Isolation of hydrogen-producing bacteria.
Hydrogen-producing bacteria were isolated from the incubated gas-producing samples.
At sea, 0.25 ml of each hydrogen-producing sample was transferred to a serum vial
prepared as above with 25 ml sterile Instant Ocean (Aquarium Systems, Inc.) in place of
seawater. When transfers of these crude cultures were returned to the lab, they were plated
on Difco marine agar 2216 and grown aerobically. Thirty-eight representative colonies
were picked and restreaked three times to ensure purity. The pure isolates were then tested
for hydrogen production using the anaerobic serum vial system described above.
Hydrogen-producing isolates were keyed to genus using the taxonomic scheme of Oliver
(1982) for marine bacteria.
Bacterial enumeration. Portions of concentrated and unconcentrated water were preserved with glutaraldehyde (2% final concentration) for bacterial enumeration. Total cells
were counted using a modification of the acridine orange staining procedure (AODC)
described by Hobbie et al. (1977).
Fluorescent antibodies were prepared in rabbits against pure cultures of the three primary
hydrogen-producing strains present in the seawater samples. The procedure used for
antisera preparation and indirect immunofluorescent staining is a modification of that
described by Ward and Perry (1980) and Ward (1982). Portions of the preserved samples
were also reserved for observation by scanning electron microscopy (SEM) by Dr Paul
LaRock (Florida State University).
RESULTS
Physical and chemical data.
Nephelometry indicated the water column to be quite low in suspended particulate matter
(except for Station 4). Particle maxima were generally poorly developed and the microbiological samples were taken at depths of 22 to 57 m, where light-scattering data
indicated particle maxima. The microbiological samples confirmed the paucity of suspended matter in the Sargasso Sea; very few particles or planktonic organisms were
visible in the concentrate. Dissolved nutrient concentrations in the surface waters in the
Sargasso Sea were quite low as had been previously reported for this region (Carpenter
and McCarthy, 1975).
Bacterial counts (AODC).
Unconcentrated seawater contained from 7.3 x 10 5 to 2.4 x 10 8 cells ml -1 (Tab. 1). The
number of cells was lowest in the central portion of the Sargasso Sea, increasing towards
its periphery and highest off the Florida coast as Station 4. At station 18 bacteria were
counted in three different size fractions. The bacteria were distributed such that 12%, 7%,
and 80 % of the cells were in the > 30 µm, 3 - 30 µm and <3 pm fractions, respectively.
202
with dual thermal conductivity detectors. The analytical column was a 2.75 m x 3 mm
stainless steel column packed with molecular sieve 5A. Argon carrier gas was used at a
flow rate of 30 ml min' 1 . Operating temperatures were (°C); injection port, 35; oven,
35 ; detector, 250.
The total hydrogen in a vial was calculated using the Bunsen solubility coefficients of
Wiesenburg and Guinasso (1979) and corrected for any gas removed during previous
analyses. The value for total hydrogen was converted to hydrogen per liter of original
unconcentrated seawater, and unless otherwise noted, results expressed as nl H
2
l -1 are on
the basis of a liter of unconcentrated seawater.
Isolation of hydrogen-producing bacteria.
Hydrogen-producing bacteria were isolated from the incubated gas-producing samples.
At sea, 0.25 ml of each hydrogen-producing sample was transferred to a serum vial
prepared as above with 25 ml sterile Instant Ocean (Aquarium Systems, Inc.) in place of
seawater. When transfers of these crude cultures were returned to the lab, they were plated
on Difco marine agar 2216 and grown aerobically. Thirty-eight representative colonies
were picked and restreaked three times to ensure purity. The pure isolates were then tested
for hydrogen production using the anaerobic serum vial system described above.
Hydrogen-producing isolates were keyed to genus using the taxonomic scheme of Oliver
(1982) for marine bacteria.
Bacterial enumeration. Portions of concentrated and unconcentrated water were preserved with glutaraldehyde (2% final concentration) for bacterial enumeration. Total cells
were counted using a modification of the acridine orange staining procedure (AODC)
described by Hobbie et al. (1977).
Fluorescent antibodies were prepared in rabbits against pure cultures of the three primary
hydrogen-producing strains present in the seawater samples. The procedure used for
antisera preparation and indirect immunofluorescent staining is a modification of that
described by Ward and Perry (1980) and Ward (1982). Portions of the preserved samples
were also reserved for observation by scanning electron microscopy (SEM) by Dr Paul
LaRock (Florida State University).
RESULTS
Physical and chemical data.
Nephelometry indicated the water column to be quite low in suspended particulate matter
(except for Station 4). Particle maxima were generally poorly developed and the microbiological samples were taken at depths of 22 to 57 m, where light-scattering data
indicated particle maxima. The microbiological samples confirmed the paucity of suspended matter in the Sargasso Sea; very few particles or planktonic organisms were
visible in the concentrate. Dissolved nutrient concentrations in the surface waters in the
Sargasso Sea were quite low as had been previously reported for this region (Carpenter
and McCarthy, 1975).
Bacterial counts (AODC).
Unconcentrated seawater contained from 7.3 x 10 5 to 2.4 x 10 8 cells ml -1 (Tab. 1). The
number of cells was lowest in the central portion of the Sargasso Sea, increasing towards
its periphery and highest off the Florida coast as Station 4. At station 18 bacteria were
counted in three different size fractions. The bacteria were distributed such that 12%, 7%,
and 80 % of the cells were in the > 30 µm, 3 - 30 µm and <3 pm fractions, respectively.
202