142
(Bird and Kalff 1984; Cole et al. 1988; Simon et
al. 1992) or nutrients (Currie 1990). Despite these
relationships, total direct counts include populations of both active and inactive cells. Bacterial assemblages in natural aquatic systems are composed
of cells in a wide range of physiological conditions,
and the techniques used for total counts provide no
information on the physiological condition of individual bacterial cells. There is now widespread
consensus that traditional culture techniques greatly
underestimate the number of metabolically active
and viable cells (Roszak and Colwell 1997), so a
number of alternative approaches, some relatively
new, have been used to try to distinguish between
populations of highly active and dormant or inactive bacteria (reviewed by McFeters et al. 1995).
Briefly, these approaches include the use of:
1. microautoradiography, in which the uptake of
radiolabeled compounds by single cells can be
detected (Karner and Fuhrman 1997);
2. fluorogenic compounds that can diffuse through
bacterial cell walls and membranes only when
the integrity of the membranes has been compromised (Jepras et al. 1995);
3. universal 16s rRNA-targeted oligonucleotide
probes to identify bacterial cells that contain sufficient ribosomal ribonucleic acid (rRNA) to be
potentially metabolically active (Karner and
Fuhrman 1997);
4. a destaining method developed by Zweifel and
Hagstrom (1995) for specifically staining with
DAPI only the nucleoid region of in situ bacterial cells. This eliminates nonspecific DAPI
staining of other cellular components. Zweifel
and Hagstrom suggested that only those cells
with identifiably stained nucleoid regions should
be considered to be alive. The nonnucleoid cells
(called "ghosts") are presumably inactive since
they contain either no or little deoxyribonucleic
acid (DNA). Choi et al. (1996), working in
coastal waters off Oregon, confirmed the finding
that a large portion of the AO or DAPI countable
cells did not have visible nucleoids. However,
under conditions that stimulated cell growth,
cells with visible nucleoids became nearly 100%
of the population. The work of Choi et al. (1996)
suggests that the ghost cells are capable of becoming active cells; and
5. stains that change upon reduction by intercepting energy along the biological electron transJonathan J. Cole
port system (ETS). Among the various stains
used for this, tetrazolium salts have received the
most attention. The tetrazolium salt, 5-cyano2,3,-ditolyl tetrazolium chloride (CTC) , becomes fluorescent upon reduction by the ETS
system and can be visualized using epifluorescence microscopy (Rodriguez et al. 1992; Gasol
et al. 1996; Smith 1998) and by flow cytometry
(del Giorgio et al. 1997b). The newer work suggests that only a fraction of the total bacterial
cells are active and that this fraction varies from
<5% in ultraoligotrophic regions to >50% in
very productive waters (del Giorgio and Scarborough 1995).
Most of these techniques can be coupled to flow
cytometry (Davey and Kell 1996), and this effectively increases the sensitivity, speed, and precision
of most determinations. In addition, subpopulations
characterized by their level of activity or genetic
composition can be physically sorted and used for
further molecular or metabolic assays (Wallner et
al. 1997).
Cell Size and Biomass
For ecosystem scientists, biomass, rather than numbers per se, is usually of more significant interest.
The standard way to obtain the biomass of planktonic bacteria is to multiply the abundance by the
cell volume. Cell volume is most often determined
by direct measurement of size and shape either under the microscope itself or from photographic images of the microscopic preparation (Psenner
1993). One then has to adopt a model relating linear
dimensions to volume. The most common of these
is:
v = (nI4)W2(L - W/3)
(9.1)
where V is volume, L is length, and W is width.
This model assumes all cells to be straight rods, but
it works equally well for spherical cocci (Bratbak
1993). The various advantages of using epifluorescent microscopy or scanning or transmission electron microscopy have been reviewed by Bratbak
(1993).
The estimation of cell biomass from cell volume
is not straightforward. Cell volumes vary over three
orders of magnitude, and biomass and volume appear to be related (Norland 1993). Many research-
(Bird and Kalff 1984; Cole et al. 1988; Simon et
al. 1992) or nutrients (Currie 1990). Despite these
relationships, total direct counts include populations of both active and inactive cells. Bacterial assemblages in natural aquatic systems are composed
of cells in a wide range of physiological conditions,
and the techniques used for total counts provide no
information on the physiological condition of individual bacterial cells. There is now widespread
consensus that traditional culture techniques greatly
underestimate the number of metabolically active
and viable cells (Roszak and Colwell 1997), so a
number of alternative approaches, some relatively
new, have been used to try to distinguish between
populations of highly active and dormant or inactive bacteria (reviewed by McFeters et al. 1995).
Briefly, these approaches include the use of:
1. microautoradiography, in which the uptake of
radiolabeled compounds by single cells can be
detected (Karner and Fuhrman 1997);
2. fluorogenic compounds that can diffuse through
bacterial cell walls and membranes only when
the integrity of the membranes has been compromised (Jepras et al. 1995);
3. universal 16s rRNA-targeted oligonucleotide
probes to identify bacterial cells that contain sufficient ribosomal ribonucleic acid (rRNA) to be
potentially metabolically active (Karner and
Fuhrman 1997);
4. a destaining method developed by Zweifel and
Hagstrom (1995) for specifically staining with
DAPI only the nucleoid region of in situ bacterial cells. This eliminates nonspecific DAPI
staining of other cellular components. Zweifel
and Hagstrom suggested that only those cells
with identifiably stained nucleoid regions should
be considered to be alive. The nonnucleoid cells
(called "ghosts") are presumably inactive since
they contain either no or little deoxyribonucleic
acid (DNA). Choi et al. (1996), working in
coastal waters off Oregon, confirmed the finding
that a large portion of the AO or DAPI countable
cells did not have visible nucleoids. However,
under conditions that stimulated cell growth,
cells with visible nucleoids became nearly 100%
of the population. The work of Choi et al. (1996)
suggests that the ghost cells are capable of becoming active cells; and
5. stains that change upon reduction by intercepting energy along the biological electron transJonathan J. Cole
port system (ETS). Among the various stains
used for this, tetrazolium salts have received the
most attention. The tetrazolium salt, 5-cyano2,3,-ditolyl tetrazolium chloride (CTC) , becomes fluorescent upon reduction by the ETS
system and can be visualized using epifluorescence microscopy (Rodriguez et al. 1992; Gasol
et al. 1996; Smith 1998) and by flow cytometry
(del Giorgio et al. 1997b). The newer work suggests that only a fraction of the total bacterial
cells are active and that this fraction varies from
<5% in ultraoligotrophic regions to >50% in
very productive waters (del Giorgio and Scarborough 1995).
Most of these techniques can be coupled to flow
cytometry (Davey and Kell 1996), and this effectively increases the sensitivity, speed, and precision
of most determinations. In addition, subpopulations
characterized by their level of activity or genetic
composition can be physically sorted and used for
further molecular or metabolic assays (Wallner et
al. 1997).
Cell Size and Biomass
For ecosystem scientists, biomass, rather than numbers per se, is usually of more significant interest.
The standard way to obtain the biomass of planktonic bacteria is to multiply the abundance by the
cell volume. Cell volume is most often determined
by direct measurement of size and shape either under the microscope itself or from photographic images of the microscopic preparation (Psenner
1993). One then has to adopt a model relating linear
dimensions to volume. The most common of these
is:
v = (nI4)W2(L - W/3)
(9.1)
where V is volume, L is length, and W is width.
This model assumes all cells to be straight rods, but
it works equally well for spherical cocci (Bratbak
1993). The various advantages of using epifluorescent microscopy or scanning or transmission electron microscopy have been reviewed by Bratbak
(1993).
The estimation of cell biomass from cell volume
is not straightforward. Cell volumes vary over three
orders of magnitude, and biomass and volume appear to be related (Norland 1993). Many research-