give a wide range (x 6,5) for each calculated value. This is due to the number of conversion
factors and assumptions required to convert 3 H-Thymidine incorporation into bacterial
biomass. Fuhrman and Azam (1980) estimated that 2.0x10 17 to 1.3x1018 cells are produced per mole of 3 H-Thymidine incorporated into DNA. Subsequently, Fuhrman and
Azam (1982) proposed that conversion factors of 1,7x10 18 and 2.4xl0 18 should be used
for nearshore and offshore waters respectively. This range in production values reflects, in
part, maximum and minimum estimates of chromosomal DNA corresponding to actively
growing and dormant cells. (Fuhrman and Azam, 1980, 1982). The accuracy of the
3 H-Thymidine incorporation method is also subject to variable isotopic dilution due to
internal synthesis of thymine and in consequence, this method is always likely to produce
an underestimate of production. Furthermore, we have recent evidence to show that some
bacteria (Flavobacteriacea) do not take up 3 H-Thymidine (Lucas, 1986). It may be that
Flavobacteria do not have an appropriate uptake mechanism (see also Pollard and
Moriarty, 1984).
Thirdly, it appears that the lower estimate of bacterial production by 3H-Thymidine
incorporation is in good agreement with estimates of bacterial production based on
predator-free incubations ; although it should be pointed out that the latter method may
give an underestimate of bacterial growth if small microflagellates (<3µm) are found to be
present in the incubation media (Fuhrman and McManus, 1984). Here, this agreement
can however be expressed as a linear regression of the lower estimate of 3H-Thymidine
production against production values based on predator-free incubations (Fig. 3). Two
regression slopes are given ; A and B. Regression slope B is fitted to all data points for the
three profiles and includes data from 5 m down to 75 m. Although there is a significant
relationship between the results of the two methods (r = 0.82 ; P<0.02), a better
correlation is obtained from slope A (r = 0.95 ; P<0.0005) in which the two highest
estimates of production associated with surface waters are omitted. Note that regression
A has a slope of 1.04 and a Y intercept close to zero (1.44) relative to the intercept of slope
B (15.4) indicating that regression A describes an almost one to one agreement between
production estimates based on predator-free incubations and the lower estimate of the
3 H- Thymidine incorporation method of Fuhrman and Azam (1980).
Bacterial and phytoplankton relationships
The more rigorous constraints that we have been able to impose on our estimates of
bacterial biomass and production now allow us to more confidently estimate the quantitative significance of the transfer of photosynthetically fixed carbon through the bacterioplankton community.
Bacterial consumption of carbon (Cc) can be estimated from the net growth yield of
bacteria (Williams, 1981) and expressed as a % of the carbon substrate utilised. Estimates
of bacterial net growth yield have however been shown to vary widely (between 10-90%)
depending on the complexity of the substrate being utilised and also on the extent of
nitrogen available to the bacteria. (Newell et al., 1981; Williams, 1981; Joint and
Morris, 1982 ; Linley and New'ell, 1984). Soluble molecules such as amino-acids and
glucose are efficiently utilised (60-90 %) relative to refractory detrital components
(10-15%).
To estimate carbon flow through bacterioplankton, based on primary production, it is
necessary to estimate bacterial carbon consumption on the basis of the differing net growth
yields recorded for the particulate and photosynthetic exudate (PDOC) fractions of
primary production. Photosynthate exudates as DOC (PDOC) have been shown to
account for as much as 30 % of the total fixed carbon (Williams, 1981) and have been
considered as the major source of carbon for bacteria in the euphotic zone (Joiris et al.,
380
factors and assumptions required to convert 3 H-Thymidine incorporation into bacterial
biomass. Fuhrman and Azam (1980) estimated that 2.0x10 17 to 1.3x1018 cells are produced per mole of 3 H-Thymidine incorporated into DNA. Subsequently, Fuhrman and
Azam (1982) proposed that conversion factors of 1,7x10 18 and 2.4xl0 18 should be used
for nearshore and offshore waters respectively. This range in production values reflects, in
part, maximum and minimum estimates of chromosomal DNA corresponding to actively
growing and dormant cells. (Fuhrman and Azam, 1980, 1982). The accuracy of the
3 H-Thymidine incorporation method is also subject to variable isotopic dilution due to
internal synthesis of thymine and in consequence, this method is always likely to produce
an underestimate of production. Furthermore, we have recent evidence to show that some
bacteria (Flavobacteriacea) do not take up 3 H-Thymidine (Lucas, 1986). It may be that
Flavobacteria do not have an appropriate uptake mechanism (see also Pollard and
Moriarty, 1984).
Thirdly, it appears that the lower estimate of bacterial production by 3H-Thymidine
incorporation is in good agreement with estimates of bacterial production based on
predator-free incubations ; although it should be pointed out that the latter method may
give an underestimate of bacterial growth if small microflagellates (<3µm) are found to be
present in the incubation media (Fuhrman and McManus, 1984). Here, this agreement
can however be expressed as a linear regression of the lower estimate of 3H-Thymidine
production against production values based on predator-free incubations (Fig. 3). Two
regression slopes are given ; A and B. Regression slope B is fitted to all data points for the
three profiles and includes data from 5 m down to 75 m. Although there is a significant
relationship between the results of the two methods (r = 0.82 ; P<0.02), a better
correlation is obtained from slope A (r = 0.95 ; P<0.0005) in which the two highest
estimates of production associated with surface waters are omitted. Note that regression
A has a slope of 1.04 and a Y intercept close to zero (1.44) relative to the intercept of slope
B (15.4) indicating that regression A describes an almost one to one agreement between
production estimates based on predator-free incubations and the lower estimate of the
3 H- Thymidine incorporation method of Fuhrman and Azam (1980).
Bacterial and phytoplankton relationships
The more rigorous constraints that we have been able to impose on our estimates of
bacterial biomass and production now allow us to more confidently estimate the quantitative significance of the transfer of photosynthetically fixed carbon through the bacterioplankton community.
Bacterial consumption of carbon (Cc) can be estimated from the net growth yield of
bacteria (Williams, 1981) and expressed as a % of the carbon substrate utilised. Estimates
of bacterial net growth yield have however been shown to vary widely (between 10-90%)
depending on the complexity of the substrate being utilised and also on the extent of
nitrogen available to the bacteria. (Newell et al., 1981; Williams, 1981; Joint and
Morris, 1982 ; Linley and New'ell, 1984). Soluble molecules such as amino-acids and
glucose are efficiently utilised (60-90 %) relative to refractory detrital components
(10-15%).
To estimate carbon flow through bacterioplankton, based on primary production, it is
necessary to estimate bacterial carbon consumption on the basis of the differing net growth
yields recorded for the particulate and photosynthetic exudate (PDOC) fractions of
primary production. Photosynthate exudates as DOC (PDOC) have been shown to
account for as much as 30 % of the total fixed carbon (Williams, 1981) and have been
considered as the major source of carbon for bacteria in the euphotic zone (Joiris et al.,
380