140
1000-;----------------:l~-__j
10
a.. 10
CD
0.1
0 . 0 1 + - - - - r - - - - - - , - - - - . - - - - - - - t
10
100
1000
10000
NPP
FIGURE 9.2. Bacterial secondary production versus phytoplankton primary production across a series of freshwater and marine systems. Each point represents the average of seasonal NPP and BP measurements in a
different lake, estuary, or marine site. The equation for
the regression is log (BP) = -0.483 + 0.814 X log
(NPP); r = 0.6; n = 54 systems; p < 0.0001. The
Model II slope is 1.06. (Redrawn from the data of Cole
et al. 1988.)
be the major reservoirs of some limiting elements,
such as phosphorus (P) (Cho and Azam 1990; Cole
and Caraco 1993) (Figure 9.3). Because bacterial
abundance varies by several orders of magnitude
across systems, it is often necessary to quantify it
in a system of interest.
Epifluorescent Direct Count
Since the late 1970s, bacterial abundance has been
measured by direct count using epifluorescent microscopy. Such procedures are designed to give the
total direct count, that is the total number of intact
bacterial cells in a sample. It was long ago realized
that various culturing techniques greatly underestimated the total number of bacterial cells in natural
waters (Hobbie et al. 1977). Thus, the total direct
count, usually by epifluorescent microscopy, has
been the standard way to enumerate bacteria for
more than 20 years.
In these procedures, cells are stained with one of
several types of fluorescent stains, filtered onto a
0.2-J-lm pore size filter and enumerated in known
field sizes, typically at magnifications of 1000 to
Jonathan J. Cole
4) 100
C)
til
'E 80
Q)
e
r---_
Q)
60
-,s
- -
phosphorus
-
a;
- -
C)
40
- -
a;
-.... 20
a;
carbon
-
.t:
Q)
0
ts
til
0.1
1
10
100
II)
Chlorophyll-a (J.l9 liter"l)
FIGURE 9.3. Modeled ratio of bacterial to algal carbon
(solid line) and phosphorus (dashed line) along a gradient
from ultraoligotrophic to hypereutrophic. The curves are
derived from the relationships between bacterial abundance and chlorophyll-a (Bird and Kalff 1984); an assumed C/chlorophyll ratio (for phytoplankton) of 50:1;
an assumed CIP for phytoplankton of 106: 1. We further
assumed a bacterial cell size of 0.12 ).lm; bacterial C content of 1.2 X 1O- 7 ).lg ).lm- 3 ; and a bacterial CIP ratio
of 50: 1. (Modified from Cole and Caraco 1993 and data
therein.)
1200 diameters using epifluorescent microscopy
(Daley and Hobbie 1975). The type of filter used
can greatly affect the outcome because some kinds
of filters (those with a cellulose-acetate tortuous
weave, such as Millipore-HA) have enough thickness that many cells are hidden from view (Hobbie
et al. 1977). The use of polycarbonate filters (which
have discrete, punctured holes and very little depth,
e.g., NUclepore or Poretics) yields significantly
higher counts (Hobbie et al. 1977). Since the cells
are viewed from above using the fluorescence of
the stained bacteria, best results are achieved when
the filters are stained black. A suitable black stain
can be achieved using Irgalan Black (Hobbie et al.
1977) or by purchasing black filters from Poretics.
An excellent, detailed protocol for sample preparation and staining is provided by Turley (1993).
The two stains in greatest use at present are acridine orange (AO; Hobbie et al. 1977) and 4'6diamidino-2-phenylindole (DAPI) (Porter and Feig
1980). Both are nucleic acid fluorochromes but neither binds only to nucleic acids. In most samples,
this nonspecific binding is a minor annoyance. In
some samples, however, especially those from eutrophic waters, highly turbid waters or waters rich
1000-;----------------:l~-__j
10
a.. 10
CD
0.1
0 . 0 1 + - - - - r - - - - - - , - - - - . - - - - - - - t
10
100
1000
10000
NPP
FIGURE 9.2. Bacterial secondary production versus phytoplankton primary production across a series of freshwater and marine systems. Each point represents the average of seasonal NPP and BP measurements in a
different lake, estuary, or marine site. The equation for
the regression is log (BP) = -0.483 + 0.814 X log
(NPP); r = 0.6; n = 54 systems; p < 0.0001. The
Model II slope is 1.06. (Redrawn from the data of Cole
et al. 1988.)
be the major reservoirs of some limiting elements,
such as phosphorus (P) (Cho and Azam 1990; Cole
and Caraco 1993) (Figure 9.3). Because bacterial
abundance varies by several orders of magnitude
across systems, it is often necessary to quantify it
in a system of interest.
Epifluorescent Direct Count
Since the late 1970s, bacterial abundance has been
measured by direct count using epifluorescent microscopy. Such procedures are designed to give the
total direct count, that is the total number of intact
bacterial cells in a sample. It was long ago realized
that various culturing techniques greatly underestimated the total number of bacterial cells in natural
waters (Hobbie et al. 1977). Thus, the total direct
count, usually by epifluorescent microscopy, has
been the standard way to enumerate bacteria for
more than 20 years.
In these procedures, cells are stained with one of
several types of fluorescent stains, filtered onto a
0.2-J-lm pore size filter and enumerated in known
field sizes, typically at magnifications of 1000 to
Jonathan J. Cole
4) 100
C)
til
'E 80
Q)
e
r---_
Q)
60
-,s
- -
phosphorus
-
a;
- -
C)
40
- -
a;
-.... 20
a;
carbon
-
.t:
Q)
0
ts
til
0.1
1
10
100
II)
Chlorophyll-a (J.l9 liter"l)
FIGURE 9.3. Modeled ratio of bacterial to algal carbon
(solid line) and phosphorus (dashed line) along a gradient
from ultraoligotrophic to hypereutrophic. The curves are
derived from the relationships between bacterial abundance and chlorophyll-a (Bird and Kalff 1984); an assumed C/chlorophyll ratio (for phytoplankton) of 50:1;
an assumed CIP for phytoplankton of 106: 1. We further
assumed a bacterial cell size of 0.12 ).lm; bacterial C content of 1.2 X 1O- 7 ).lg ).lm- 3 ; and a bacterial CIP ratio
of 50: 1. (Modified from Cole and Caraco 1993 and data
therein.)
1200 diameters using epifluorescent microscopy
(Daley and Hobbie 1975). The type of filter used
can greatly affect the outcome because some kinds
of filters (those with a cellulose-acetate tortuous
weave, such as Millipore-HA) have enough thickness that many cells are hidden from view (Hobbie
et al. 1977). The use of polycarbonate filters (which
have discrete, punctured holes and very little depth,
e.g., NUclepore or Poretics) yields significantly
higher counts (Hobbie et al. 1977). Since the cells
are viewed from above using the fluorescence of
the stained bacteria, best results are achieved when
the filters are stained black. A suitable black stain
can be achieved using Irgalan Black (Hobbie et al.
1977) or by purchasing black filters from Poretics.
An excellent, detailed protocol for sample preparation and staining is provided by Turley (1993).
The two stains in greatest use at present are acridine orange (AO; Hobbie et al. 1977) and 4'6diamidino-2-phenylindole (DAPI) (Porter and Feig
1980). Both are nucleic acid fluorochromes but neither binds only to nucleic acids. In most samples,
this nonspecific binding is a minor annoyance. In
some samples, however, especially those from eutrophic waters, highly turbid waters or waters rich