54
UORDON A . RILEY
was obtained, but the peak was reached more slowly. In this experiment the dissolved organic carbon was initially 2.7 mg C/litre. At the
time of the peak it had declined to 1.95 mgllitre. Thus about half of
the observed decrease represented conversion to particulate matter,
living or dead, and the remainder can be ascribed to metabolic utilization. During the remainder of the experiment the dissolved organic
carbon remained constant within the limits of error of the analytical
method. Thus the decrease in particulate carbon after its peak is
presumed to be due to bacterial utilization, and this is followed by a
steady state situation in which the metabolism of the bacteria proceeds
at a rate that is too low to be detected by the methods employed.
Studies of in situ accumulation of particulate matter have been
carried out by filtering sea water through Millipore or silver filters with
a pore size of 0-45 p or larger, which allow part of the bacterial flora
t o pass through and repopulate the water. Filtration has been carried
out by gravity, or sometimes using pressure generated by a peristaltic
pump. In an attempt to reduce air-water interfaces as much as possible,
most of the experiments have utilized polyethylene bags as experimental
containers. These bags were of about 4-litre capacity and were sealed
at the top, with a polyethylene tube sealed into each corner. One tube
extended to the bottom and served as an inlet. A short tube at the other
corner was used to evacuate the bag slightly. Thus the inflowing water
gradually filled and expanded the bags and had no contact with air
except for occasional accidental bubbles.
Glass bottles were used in some experiments for comparison with
results obtained in bags and also for experiments involving long period
storage. The bottles were filled by gravity filtration with as little
disturbance as possible and were allowed to overflow before sealing,
in the same way that bottles are filled for oxygen analysis.
Fig. 6 shows two examples of the kind of results obtained in bag
experiments, using water that was filtered but with no bacteria added.
Each point represents the mean carbon content of a bag that was refiltered at the time indicated on the graph. The particulate carbon
during the first few days was indistinguishable from the amount
obtained when water is refiltered immediately. Subsequently the
carbon content increased to roughly the same value that was found
when samples were inoculated artificially, the only difference being in
the length of time required to reach the peak. Other experiments that
lasted longer have shown that in this type of experiment, as in others
described above, the particulate carbon decreased later and tended to
level off. However, in both types there have been a few experiments
which did not follow the usual pattern in that the carbon increased and
UORDON A . RILEY
was obtained, but the peak was reached more slowly. In this experiment the dissolved organic carbon was initially 2.7 mg C/litre. At the
time of the peak it had declined to 1.95 mgllitre. Thus about half of
the observed decrease represented conversion to particulate matter,
living or dead, and the remainder can be ascribed to metabolic utilization. During the remainder of the experiment the dissolved organic
carbon remained constant within the limits of error of the analytical
method. Thus the decrease in particulate carbon after its peak is
presumed to be due to bacterial utilization, and this is followed by a
steady state situation in which the metabolism of the bacteria proceeds
at a rate that is too low to be detected by the methods employed.
Studies of in situ accumulation of particulate matter have been
carried out by filtering sea water through Millipore or silver filters with
a pore size of 0-45 p or larger, which allow part of the bacterial flora
t o pass through and repopulate the water. Filtration has been carried
out by gravity, or sometimes using pressure generated by a peristaltic
pump. In an attempt to reduce air-water interfaces as much as possible,
most of the experiments have utilized polyethylene bags as experimental
containers. These bags were of about 4-litre capacity and were sealed
at the top, with a polyethylene tube sealed into each corner. One tube
extended to the bottom and served as an inlet. A short tube at the other
corner was used to evacuate the bag slightly. Thus the inflowing water
gradually filled and expanded the bags and had no contact with air
except for occasional accidental bubbles.
Glass bottles were used in some experiments for comparison with
results obtained in bags and also for experiments involving long period
storage. The bottles were filled by gravity filtration with as little
disturbance as possible and were allowed to overflow before sealing,
in the same way that bottles are filled for oxygen analysis.
Fig. 6 shows two examples of the kind of results obtained in bag
experiments, using water that was filtered but with no bacteria added.
Each point represents the mean carbon content of a bag that was refiltered at the time indicated on the graph. The particulate carbon
during the first few days was indistinguishable from the amount
obtained when water is refiltered immediately. Subsequently the
carbon content increased to roughly the same value that was found
when samples were inoculated artificially, the only difference being in
the length of time required to reach the peak. Other experiments that
lasted longer have shown that in this type of experiment, as in others
described above, the particulate carbon decreased later and tended to
level off. However, in both types there have been a few experiments
which did not follow the usual pattern in that the carbon increased and