62
GORDON A. RILEY
I n all cases examined, higher values were obtained subsequently.
The maximum count recorded was 87 x l o 3 cells/ml. Most experiments
had maximum readings of 30 x lo3 or higher, but a few were lower.
The sampling interval may not have been frequent enough t o record
true maxima, and this may have accounted for some of the variability.
However, there was an extreme case in which the maximum bacterial
count was only 8 x 103 cells/ml, and this was one of the experiments
in which organic carbon did not exhibit a n early peak either. The
maximum observed concentration during the 10-day experiment was
140 pg C/litre.
The earlier discussion indicated that addition of an inoculum of
bacteria is not necessary for formation of particulate carbon. Aggregation occurs more slowly when bacteria are not added, but there is
little difference in the peak concentration that is eventually achieved.
Plate counts showed that there was in fact an abundant increase of
bacteria in filtered sea water, but maximum concentrations were seldom
as large as those obtained when an inoculum had been introduced. This
is best demonstrated in six pairs of experiments which were replicates
with respect to the original source of sea water and mode of handling
and differed only in that one set was inoculated with Pseudomonas and
the other was not. Maximum observed concentrations averaged 39 x
lo3 cellslml in the first set and 16 x lo3 in the second.
I n experiments inoculated with Pseudomonas, plate counts declined
rapidly after the early maximum, and by the end of the first week most
counts were of the order of I2 x lo3 cells/ml or less. The total range
was of the order of 1-12 x 103, and a reasonable average for the series
is about 8 x 103. There was considerable variation from one plate
count to another in the same experiment, but comparisons of replicate
plates and successive dilutions suggest that the variation can be
ascribed largely to sampling error rather than to real variation in the
size of the population.
I n cases where the bacterial population was derived from natural
flora which escaped the filter, the experiments generally did not last
long enough to permit a good estimate of the population level subsequent to the maximum. This matter will be taken u p in connection
with long term experiments to be described later.
The relationship of plate counts to the real population is always
problematical. Direct counts in natural sea water commonly range
from 1 to 4 orders of magnitude larger than plate counts (Jannasch and
Jones, 1959). The culture medium and often the physical conditions
of the laboratory environment are not suitable for the growth of a
large majority of naturally occurring species. However, the Pseudo-
GORDON A. RILEY
I n all cases examined, higher values were obtained subsequently.
The maximum count recorded was 87 x l o 3 cells/ml. Most experiments
had maximum readings of 30 x lo3 or higher, but a few were lower.
The sampling interval may not have been frequent enough t o record
true maxima, and this may have accounted for some of the variability.
However, there was an extreme case in which the maximum bacterial
count was only 8 x 103 cells/ml, and this was one of the experiments
in which organic carbon did not exhibit a n early peak either. The
maximum observed concentration during the 10-day experiment was
140 pg C/litre.
The earlier discussion indicated that addition of an inoculum of
bacteria is not necessary for formation of particulate carbon. Aggregation occurs more slowly when bacteria are not added, but there is
little difference in the peak concentration that is eventually achieved.
Plate counts showed that there was in fact an abundant increase of
bacteria in filtered sea water, but maximum concentrations were seldom
as large as those obtained when an inoculum had been introduced. This
is best demonstrated in six pairs of experiments which were replicates
with respect to the original source of sea water and mode of handling
and differed only in that one set was inoculated with Pseudomonas and
the other was not. Maximum observed concentrations averaged 39 x
lo3 cellslml in the first set and 16 x lo3 in the second.
I n experiments inoculated with Pseudomonas, plate counts declined
rapidly after the early maximum, and by the end of the first week most
counts were of the order of I2 x lo3 cells/ml or less. The total range
was of the order of 1-12 x 103, and a reasonable average for the series
is about 8 x 103. There was considerable variation from one plate
count to another in the same experiment, but comparisons of replicate
plates and successive dilutions suggest that the variation can be
ascribed largely to sampling error rather than to real variation in the
size of the population.
I n cases where the bacterial population was derived from natural
flora which escaped the filter, the experiments generally did not last
long enough to permit a good estimate of the population level subsequent to the maximum. This matter will be taken u p in connection
with long term experiments to be described later.
The relationship of plate counts to the real population is always
problematical. Direct counts in natural sea water commonly range
from 1 to 4 orders of magnitude larger than plate counts (Jannasch and
Jones, 1959). The culture medium and often the physical conditions
of the laboratory environment are not suitable for the growth of a
large majority of naturally occurring species. However, the Pseudo-