Roberson EB, Firestone MK.. Relationship between desiccation and exopolysaccharide production in a soil Pseudomonas sp. Appl Environ Microbiol 58: 1284-1291

Abstract

The relationship between desiccation and the production of extracellular polysaccharides (EPS) by soil bacteria was investigated by using a Pseudomonas species isolated from soil. Cultures subjected to desiccation while growing in a sand matrix contained more EPS and less protein than those growing at high water potential, suggesting that resources were allocated to EPS production in response to desiccation. Desiccation did not have a significant effect on activity as measured by reduction of iodonitrotetrazolium. Purified EPS produced by the Pseudomonas culture contained several times its weight in water at low water potential. Sand amended with EPS held significantly more water and dried significantly more slowly than unamended sand, implying that an EPS matrix may buffer bacterial colonies from some effects of desiccation. We conclude that bacteria may use EPS production to alter their microenvironment to enhance survival of desiccation.

Full text

APPLIED
AND
ENVIRONMENTAL
MICROBIOLOGY,
Apr.
1992,
p.
1284-1291
0099-2240/92/041284-08$02.00/0
Copyright
©
1992,
American
Society
for
Microbiology
Relationship
between
Desiccation
and
Exopolysaccharide
Production
in
a
Soil
Pseudomonas
sp.
EMILY
B.
ROBERSON'
AND
MARY
K.
FIRESTONE2*
Abacus
Concepts,
Inc.,
1984
Bonita
Avenue,
Berkeley,
Califomnia
94704,1
and
Department
of
Soil
Science,
University
of
California,
Berkeley,
California
947202
Received
15
July
1991/Accepted
2
February
1992
The
relationship
between
desiccation
and
the
production
of
extracellular
polysaccharides
(EPS)
by
soil
bacteria
was
investigated
by
using
a
Pseudomonas
species
isolated
from
soil.
Cultures
subjected
to
desiccation
while
growing
in
a
sand
matrix
contained
more
EPS
and
less
protein
than
those
growing
at
high
water
potential,
suggesting
that
resources
were
allocated
to
EPS
production
in
response
to
desiccation.
Desiccation
did
not
have
a
significant
effect
on
activity
as
measured
by
reduction
of
iodonitrotetrazolium.
Purified
EPS
produced
by
the
Pseudomonas
culture
contained
several
times
its
weight
in
water
at
low
water
potential.
Sand
amended
with
EPS
held
significantly
more
water
and
dried
significantly
more
slowly
than
unamended
sand,
implying
that
an
EPS
matrix
may
buffer
bacterial
colonies
from
some
effects
of
desiccation.
We
conclude
that
bacteria
may
use
EPS
production
to
alter
their
microenvironment
to
enhance
survival
of
desiccation.
Bacteria
in
soil
generally
live
in
colonies
within
a
matrix
largely
composed
of
extracellular
polysaccharides
(EPS)
(10,
11,
16,
26).
Although
there
is
general
agreement
that
bacteria
in
many
environments,
including
soils
and
lungs,
live
within
an
EPS
matrix,
relatively
little
is
known
about
the
func-
tion(s)
of
this
EPS
matrix.
One
possibility
that
has
often
been
discussed
but
that
has
been
the
subject
of
few
studies
is
that
an
EPS
envelope
may
protect
bacteria
from
drying
and
from
fluctuations
in
water
potential
(15,
39).
Water
potential
is
the
potential
energy
of
water.
Because
water
moves
freely
across
microbial
cell
membranes,
ther-
modynamic
laws
require
that
the
internal
water
potential
of
unicellular
organisms
be
in
equilibrium
with
the
external
water
potential.
When
the
external
water
potential
decreases
in
a
drying
soil,
soil
microorganisms
may
retain
water
by
increasing
their
internal
solute
concentration,
or
they
may
lose
water
to
their
surroundings
(plasmolyze),
which
can
result
in
cell
death.
Soil
microorganisms
have
been
reported
to
synthesize
organic
compatible
solutes
(14,
24)
such
as
betaines
or
amino
acids
and
to
selectively
take
in
inorganic
solutes
such
as
K+
(21)
to
increase
their
internal
solute
concentrations
during
periods
of
low
external
water
poten-
tial.
Bacteria
may
also
change
the
structure
of
their
mem-
branes
(9)
or
make
other
physiological
changes
in
response
to
desiccation.
An
EPS
matrix
may
slow
the
rate
at
which
a
bacterial
colony
equilibrates
with
the
surrounding
soil.
Slow-
ing
the
rate
of
drying
within
the
colony
microenvironment
could
increase
bacterial
survival
by
increasing
the
time
available
for
metabolic
adjustment.
Clays,
which
slow
the
rate
of
drying
of
soil
(29),
have
been
shown
to
increase
the
ability
of
bacteria
to
survive
desiccation
in
soil
(6,
15).
An
EPS
matrix
may
provide
another
advantage
to
bacteria
living
within
it.
Decreasing
the
water
content
of
soil
restricts
diffusion
of
nutrients
to
microorganisms
(7,
14,
23,
34a).
Polysaccharides
are
hygroscopic
(25,
38)
and
therefore
may
maintain
a
higher
water
content
in
the
colony
microenviron-
ment
than
in
the
bulk
soil
as
water
potential declines.
This
increase
in
water
content
could
increase
nutrient
availability
within
the
bacterial
colony.
*
Corresponding
author.
There
is
some
evidence
in
plants
that
polysaccharides
increase
survival
and
activity
during
drying.
Morse
(25)
studied
two
varieties
of
the
composite
Hemizonia
luzulifolia,
whose
leaf
intercellular
spaces
contain
very
different
amounts
of
EPS.
Plants
with
high
levels
of
EPS
showed
higher
transpiration
rates
than
low-level-EPS
plants
during
midday
water
stress.
Morse
found
that
water
potential
within
the
leaves of
high-level-EPS
plants
at
midday
was
significantly
higher
than
in
the
leaves
of
low-level-EPS
plants,
allowing
the
increased
activity.
She
described
the
polysaccharides
as
"buffers"
which
held
water
and
pro-
tected
leaves
from
drying.
Several
studies
(5,
6,
15,
30)
have
investigated
the
rela-
tionship
between
bacterial
EPS
production
on
agar
plates
or
in
liquid
medium
and
ability
to
survive
desiccation
in
soil.
These
studies
have
generally
found
no
relationship
or
an
inverse
correlation
between
EPS
production
and
ability
to
survive
desiccation.
However,
to
our
knowledge,
no
one
has
examined
whether
bacteria
respond
to
desiccation
by
in-
creasing
production
of
EPS.
In
the
work
reported
here,
we
determined
(i)
whether
desiccation
stimulates
polysaccharide
production
by
a
soil
bacterial
isolate
and
(ii)
whether
the
water
retention
charac-
teristics
of
the
EPS
produced
by
the
isolate
might
help
bacteria
living
within
an
EPS
matrix
to
survive
desiccation.
MATERIALS
AND
METHODS
Bacteria.
The
strain
used
was
isolated
from
an
agricultural
soil
in
the
Sacramento
Valley
of
California.
The
Mediterra-
nean
climate
in
this
area
exposes
the
soil
microbial
commu-
nity
to
frequent
and
severe
desiccation.
The
strain
was
selected
for
mucoid
colony
appearance.
It
is
a
motile,
yellow-pigmented,
gram-negative,
oxidase-positive
rod
and
has
been
identified
as
a
Pseudomonas
species
by
its
fatty
acid
profile
(MIDI,
Inc.,
Newark,
Del.).
The
EPS
produced
by
this
strain
contains
both
neutral
monosaccharide
and
uronic
acid
subunits.
The
ratio
of
neutral
to
uronic
acid
subunits
in
the
EPS
varies
depending
on
the
medium
com-
position
(data
not
shown).
Two
series
of
experiments
were
performed.
The
first
group
monitored
bacterial
protein,
activity,
and
EPS
produc-
1284
Vol.
58,
No.
4

DESICCATION
AND
BACTERIAL
POLYSACCHARIDE
PRODUCTION
1285
tion
during
1
cycle of
wetting
and
drying.
The
second
group
examined
the
water
retention
characteristics
of
the
EPS
produced
by
the bacteria
and
the
effect
of
EPS
on
the
rate
of
drying.
Metabolic
response
of
bacteria
to
the
wet-dry
cycle.
A
sand
matrix
and
mineral
salts
growth
medium
was
chosen
for
this
experiment
so
that
nutrient
supply
could
be
tightly
con-
trolled
and
monitored
and
so
that
bacterial
protein
and
EPS
production
could
be
precisely
measured.
Bacteria
were
grown
to
mid-log
phase
(about
2
x
108
cells
per
ml)
in
a
mineral
salts
medium
consisting
of
4
g
of
NH4Cl,
6.98
g
of
Na2HPO4
7H20,
5.58
g
of
KH2PO4,
20
g
of
glucose,
and
40
ml
of
Huntner's
mineral
supplement
(13)
per
liter.
The
initial
pH
was
adjusted
to
6.8.
Sand
cultures
were
grown
in
99-mm-diameter
Pyrex
petri
dishes.
Fifty
grams
of
sterile
quartz
sand
which
had
been
acid
washed,
rinsed
with
deionized
water,
and
equilibrated
with
growth
medium
was
inoculated
with
sufficient
unwashed
culture
to
bring
the
total
water
potential
to
-0.025
MPa
(130
RI
of
culture
per
g
of
sand).
All
petri
plates
were
placed
in
sterile
desiccators
containing
solid
LiCl
to
produce
a
relatively
constant
low
relative
humidity
and
to
dry
the
plates
reproducibly.
The
cultures
were
allowed
to
dry
to
-1.5
MPa
and
then
read-
justed
to
-0.025
MPa
with
0.25x
growth
medium.
Three
plates
were
collected
for
measurements
on
day
0,
after
this
wetting.
The
bacteria
were
exposed
to
this
drying
and
wetting
cycle
before
the
experiment
in
order
to
acclimatize
them
to
growing
in
the
presence
of
surfaces.
On
day
0,
after
wetting,
half
of
the
remaining
plates
were
replaced
in
the
LiCl-containing
desiccators
and
half
were
placed
in
desiccators
containing
sterile
distilled
water.
The
distilled
water
maintained
a
high
relative
humidity
in
the
desiccators.
The
water
potential
in
the
cultures
in
the
water-containing
desiccators
remained
roughly
constant
dur-
ing
the
experiment,
and
this
was
used
as
a
control
treatment
group.
Three
petri
plates
from
the
LiCl
chambers
and
three
petri
plates
from
the
control
chambers
were
collected
each
day,
weighed
to
determine
water
content,
and
sampled
for
viable
counts,
protein,
electron
transport
activity,
total
carbohy-
drates,
glucose,
and
ammonium.
A
moisture
release
curve
was
used
to
convert
water
content
to
water
potential.
The
experiment
was
terminated
when
the
average
water
potential
in
the
desiccated
treatment
group
reached
-1.5
MPa,
a
water
potential
which
previously
has
been
reported
to
limit
bacterial
activity
(14,
32,
34a).
CFU
were
counted
by
spreading
dilutions
of
the
sand
cultures
on
agar
plates
containing
growth
medium.
One
drop
of
the
surfactant
Tween
80
was
added
to
the
initial
sand-
buffer
mixture,
and
the
mixture
was
gently
mixed
in
a
vortex
mixer
to
facilitate
detachment
of
bacteria
from
sand
surfaces
(33).
The
protein
content
of
the
cultures
was
determined
by
the
Coomassie
blue
method
(Bio-Rad
Laboratories,
Rich-
mond,
Calif.).
The
cultures
were
heated
to
100°C
for
10
min
in
1
N
NaOH
to
release
cell
contents
and
were
filtered,
and
the
amount
of
protein
was
measured
in
the
filtrate.
Lyso-
zyme
in
1
N
NaOH
was
used
as
the
standard.
The
results
are
reported
as
micrograms
of
lysozyme
equivalents
per
gram
of
sand.
Respiratory
activity
was
estimated
by
the
reduction
of
iodonitrotetrazolium
(INT)
to
INT-formazan
(36).
The
INT
solution
was
sterilized
by
passage
through
a
sterile
0.2-,um-
pore-size
Acrodisc
filter
(Gelman
Sciences,
Ann
Arbor,
Mich.).
Sand
culture
was
weighed
into
sterile
glass
vials,
and
a
0.4%
(wt/vol)
solution
of
INT
(grade
1;
Sigma
Chemical
Co.,
St.
Louis,
Mo.)
in
distilled
water
was
added.
The
suspension
was
well
mixed,
incubated
at
30°C
for
2
h
in
the
sealed
vial,
and
stored
at
-20°C
until
it
was
extracted
and
analyzed.
The
amount
of
INT-formazan
in
a
dimethylforma-
mide
extract
of
the
sand
cultures
was
measured.
Sand
was
extracted
twice
with
dimethylformamide
(12),
the
extracts
were
combined,
and
the
A485
was
read
with
a
spectropho-
tometer.
The
A485
was
converted
to
concentration
by
using
a
standard
curve
of
the
absorbance
of
INT-formazan
in
dim-
ethylformamide.
Polysaccharides
were
extracted
from
the
sand
by
a
method
similar
to
that
of
Oades
et
al.
(27).
Cultures
were
heated
to
120°C
with
5
N
H2SO4
for
30
min
and
filtered
through
a
glass
fiber
filter.
Samples
were
further
extracted
once
with
boiling
water.
The
extracts
were
pooled,
and
the
amounts
of
neutral
and
uronic
acid
carbohydrates
were
determined
with
anthrone
and
m-hydroxydiphenyl,
respec-
tively
(3,
4).
The
results
are
reported
as
micrograms
of
glucose
and
glucuronic
acid equivalents
per
gram
of
sand,
respectively.
Residual
amounts
of
glucose
and
NH4'
and
NO3
in
the
culture
solution
were
measured
in
distilled
water
extracts.
The
amount
of
glucose
was
determined
by
the
glucose
oxidase
assay
(19),
and
NH4'
and
NO3
amounts
were
measured
colorimetrically
(18)
with
a
Lachat
autoanalyzer.
Since
only
trace
concentrations
of
nitrate
were
found
in
the
cultures,
only
NH4'
is
reported.
The
amount
of
glucose
was
subtracted
from
that
of
total
neutral
carbohydrates
to
calcu-
late
the
amount
of
neutral
polysaccharides.
Microscopy.
Microscopic
observations
were
made
at
the
Station
de
Science
du
Sol,
Institute
National
de
la
Recherche
Agronomique,
Versailles,
France.
A
separate
but
identical
experiment
was
performed
to
produce
samples
for
examina-
tion
by
scanning
electron
microscopy
(SEM).
We
used
a
cryoscan
device
(Oxford
Cryotrans
Temperature
and
Prep-
aration
Controllers),
which
makes
it
unnecessary
to
dry
samples
before
they
are
placed
in
the
microscope.
Conven-
tional
SEM
preparation
techniques
such
as
alcohol
dehydra-
tion
and
critical
point
drying
have
been
found
to
severely
affect
the
structure
and
apparent
quantity
of
bacterial
EPS
(37).
Small
samples
were
collected
from
both
desiccated
and
control
treatment
groups
each
day,
frozen
to
-210°C
in
an
N2
slush,
and
transferred
into
a
liquid
N2-cooled
Phillips
525
M
SEM.
Ice
was
removed
from
the
surface
of
the
samples
by
20
min
of
sublimation
at
-80°C,
and
the
samples
were
coated
with
gold
and
observed
at
-
180°C.
By
using
cryoscan
SEM,
we
were
able
to
observe
qualitative
changes
in
both
the
amount
and
the
hydrated
structure
of
the
EPS
in
the
cultures
as
water
potential
changed
during
desiccation.
EPS
water
retention
characteristics.
(i)
Polysaccharide
moisture
release
curve.
Bacteria
were
grown
to
late
log
phase
in
mineral
salts
medium
and
separated
from
EPS
in
solution
by
two
14-min
centrifugations
at
10,000
x
g
at
4°C.
The
supernatant
from
the
first
centrifugation
was
decanted,
and
the
cells
were
resuspended
in
phosphate
buffer
and
centri-
fuged
again.
The
two
supernatants
were
combined
and
extensively
dialyzed
against
distilled
water
by
using
Spectra/
Por
1,000-Da-cutoff
dialysis
membranes
(Spectrum
Inc.,
Los
Angeles,
Calif.)
to
remove
residual
growth
medium.
Polysaccharides
were
then
concentrated
approximately
five-
fold
by
further
dialysis
against
solid
polyethylene
glycol
(average
molecular
weight,
8,000)
and
finally
evaporated
to
dryness
with
a
Speedvac
concentrator
(Savant
Medical
Industries,
Farmingdale,
N.Y.).
The
absence
of
protein
contamination
in
the
concentrated
solution
was
confirmed
by
the
Coomassie
blue
method
(Bio-Rad
Laboratories).
A
VOL.
58,
1992

1286
ROBERSON
AND
FIRESTONE
0
co
0
-0.4
-0.8
-1.2
-1.6
0
Day
2
3
FIG.
1.
Water
potential
changes
in
the
two
treatment
groups.
P
value
is
the
level
of
significance
for
difference
between
values
for
desiccated
and
control
treatment
groups
from
days
1
to
3.
I0
indicates
that
the
treatment
groups
were
not
separated
until
after
day
0
measurements.
range
of
volumes
(30
to
400
,u)
of
distilled
water
was
added
to
approximately
40
mg
of
dry
polysaccharide
in
gas-tight
1.5-ml
microcentrifuge
tubes.
The
mixtures
were
mixed
in
a
vortex
mixer
and
then
centrifuged
at
14,000
x
g
for
2
min
to
thoroughly
mix
the
water
with
the
polysaccharide,
and
they
were
allowed
to
equilibrate
overnight
at
4°C.
The
water
potential
at
each
water
content
was
determined.
(ii)
Effect
of
EPS
on
drying
rate.
EPS
were
purified
and
concentrated
by
dialysis
as
described
above
but
were
not
lyophilized.
Sufficient
concentrated
EPS
solution
(100
mg
of
EPS
per
ml)
was
added
to
three
replicate
petri
plates
containing
50
g
of
sterile
acid-washed
sand
to
bring
the
water
potential
to
-0.2
MPa.
Concentrated
EPS
solution
was
used
in
order
to
approximate
the
immediate
bacterial
microenvi-
ronment,
which
is
essentially
pure
polysaccharide.
Three
control
plates
were
brought
to
-0.2
MPa
with
sterile
deion-
ized
water.
Both
solutions
contained
0.2%
sodium
azide
to
maintain
sterility.
Pairs
of
sand-filled
petri
plates,
one
con-
taining
EPS
solution
and
one
containing
water,
were
placed
in
desiccators
containing
LiCl
and
allowed
to
evaporate.
Samples
were
taken
from
each
dish
at
each
hour,
and
the
water
content
and
water
potential
were
measured.
(iii)
Water
potential.
In
all
experiments,
water
potential
values
of
<-0.2
MPa
were
determined
with
an
HR33T
microvoltimeter
with
a
C32
thermocouple
psychrometer
sample
chamber
(Wescor
Co.,
Logan,
Utah)
by
using
an
equilibration
time
of
50
min
before
measurement.
Water
content
at
water
potential
values
of
>
-0.2
MPa
were
measured
with
a
pressure
plate
apparatus
(Soil
Moisture
Equipment
Corporation,
Santa
Barbara,
Calif.).
Sand
and
medium
were
sterilized
with
an
autoclave.
Statistical
analyses.
All
statistical
analyses
were
performed
and
graphs
were
created
with
StatView
and
SuperANOVA
software
(1,
2).
RESULTS
Metabolic
response
of
bacteria
to
the
wet-dry
cycle.
The
change
in
water
potential
in
the
two
treatment
groups
is
shown
in
Fig.
1.
All
points
in
the
figures
represent
the
means
of
three
replicates.
Drying
in
the
desiccated
treatment
group
began
immediately,
and
by
day
2,
the
water
potentials
in
the
cultures
averaged
-0.47
MPa.
On
day
3,
the
desiccated
cultures
had
reached
a
water
potential
of
-1.46
MPa,
which
0.25
._
a
0.15
N
Z
0.05
0)
240
"0
a
e)
0)
-
0.
200
160
120
Electron
Transport
Activity
P.0.43
Protein
,
140
120
/
/
Control
\
100
Desiccated
P
0.01
.I,
.
,
80
0
Day
2
to
0
(D.
.
0
I
c
3
FIG.
2.
Changes
in
the
amount
of
protein
and
electron
transport
activity
as
measured
by
the
reduction
of
INT.
P
values
are
the
levels
of
significance
for
differences
between
desiccated
and
control
treat-
ment
groups
from
days
1
to
3.
O0
indicates
that
treatment
groups
were
not
separated
until
after
day
0.
was
close
to
their
water
potential
after
the
preexperiment
drying
cycle
on
day
0.
The
water
potentials
in
the
control
treatment
group
remained
relatively
constant
after
initial
wetting.
Figure
2
shows
the
change
in
activity
and
protein
amount
during
the
experiment.
Electron
transport
activity
decreased
in
both
treatment
groups
over
the
course
of
the
experiment.
Although
the
treatments
did
not
significantly
affect
activity,
the
level
of
activity
appeared
slightly
higher
in
the
desiccated
treatment
than
in
the
control
group
on
days
1
and
2
but
fell
below
that
of
the
control
group
by
day
3.
Desiccation
significantly
reduced
the
amount
of
protein
in
the
cultures.
The
amount
of
protein
in
the
control
cultures
increased
between
days
0
and
3,
but
in
the
desiccated
treatment
group,
protein
concentration
increased
for
only
1
day
after
wetting
and
then
declined.
The
number
of
CFU
was
also
significantly
lower
in
the
desiccated
treatment
group
than
in
the
control
group
(data
not
shown).
The
changes
in
glucose
and
ammonium
concentrations
in
the
cultures
are
shown
in
Fig.
3.
These
are
expressed
as
the
percentage
of
the
total
amount
supplied
(initial
culture
medium
plus
0.25
x
medium
added
on
day
0)
that
was
found
in
the
cultures
each
day.
Almost
no
glucose
was
detected
in
the
cultures
during
the
experiment,
and
the
amount
of
glucose
was
not
affected
by
the
treatments.
Significantly
more
residual
ammonium
was
present
in
the
control
cultures
than
in
the
desiccated
treatment
group.
After
the
ammonium
added
in
the
diluted
medium
on
day
0
was
used
by
the
cultures,
the
amount
of
ammonium
changed
little
during
the
experiment
in
either
treatment
group.
Figure
4
shows
the
amounts
of
neutral,
uronic
acid,
and
total
polysaccharide
in
the
cultures
each
day.
There
was
a
Control
PS0.01
Desiccated
APPL.
ENVIRON.
MICROBIOL.
1

DESICCATION
AND
BACTERIAL
POLYSACCHARIDE
PRODUCTION
1287
.cm
.Ea
C
E
0,
z
I
z
12
8
4
0
0.08
I'
co
0*
0
8
c
0.06
0.04
P
<0.61
0
1
2
3
Day
FIG.
3.
Availability
of
glucose
and
ammonium,
expressed
as
the
percentage
of
the
total
amount
supplied,
that
was
present
in
the
cultures
each
day.
P
values
are
the
levels
of
significance
for
differences
between
desiccated
and
control
treatment
groups
from
days
1
to
3.
E1
indicates
that
treatment
groups
were
not
separated
until
after
day
0
measurements.
sharp
decrease
in
these
amounts
after
wetting
on
day
0
for
both
treatment
groups,
possibly
the
result
of
EPS
consump-
tion
by
the
growing
cultures.
The
amount
of
polysaccharides
in
the
control
cultures
remained
low
throughout
the
experi-
ment.
The
amount
in
the
desiccated
treatment
group,
how-
ever,
began
to
increase
immediately
as
the
water
potential
decreased
and
continued
to
increase
until
day
3.
As
was
found
for
protein,
the
quantity
of
polysaccharides
in
the
desiccated
cultures
at
the
end
of
the
experiment
was
similar
to
that
on
day
0,
at
the
end
of
the
preexperiment
desiccation.
There
was
no
significant
difference
between
the
proportion
of
uronic
acid
in
the
EPS
in
the
desiccated
and
control
treatment
groups
(data
not
shown).
Micrographs.
The
micrographs
(Fig.
5)
show
bacteria
and
EPS
on
sand
after
wetting
on
day
1,
when
the
water
potential
in
the
cultures
was
-0.025
MPa
(Fig.
5A
through
C),
and
after
the
cultures
had
dried
to
approximately
-1.0
MPa
(D
through
F).
Fibers
of
EPS
are
visible
in
Fig.
5B
and
C,
but
bacteria
without
obvious
EPS
can
also
be
seen
in
Fig.
SA.
Panels
D
through
F
show
that
the
amount
of
EPS
in
the
cultures
substantially
increased
after
exposure
to
desicca-
tion.
Thick
layers
of
EPS
covered
bacteria
and
sand
sur-
faces.
Before
desiccation,
the
edges
of
the
bacteria
appear
sharp
(Fig.
5A
through
C),
while
after
desiccation,
the
bacteria
appear
to
be
partially
embedded
in
EPS.
EPS
water
retention
characteristics.
(i)
Polysaccharide
moisture
release
curve.
The
moisture
release
curve
(Fig.
6)
shows
the
relationship
between
water
potential
and
water
content
for
the
purified
EPS.
The
EPS
showed
a
high
affinity
for
water
at
all
water
potentials.
At
-1.5
MPa,
the
EPS
held
approximately
five
times
its
weight
in
water.
At
-0.5
MPa,
1300
[
1100
.
900
g
850
co
0
as
2
750
10
3
Im
650
X
2100
co
(a)
c
1900
0
1700
F
go
520
ca.
c
6
440
'-
360
Uronic
Acids
PSO
O7
.315
278
*
Control
241
O
Desiccated
I
Total
Polysaccharides
-
850
750
-
650
PS0.02
I
1~~~~~~~~550
0
2
3
e
Ca
0
co
o
0
0c
6
co
Day
FIG.
4.
Neutral,
uronic
acid,
and
total
polysaccharides
in
the
cultures.
Uronic
acids
and
neutral
polysaccharides
are
expressed
as
glucuronic
acid
and
glucose
equivalents,
respectively.
P
values
are
the
levels
of
significance
for
differences
between
desiccated
and
control
treatment
groups
from
days
1
to
3.
O0
indicates
that
treat-
ment
groups
were
not
separated
until
after
day
0
measurements.
it
contained
10
times
its
weight
in
water.
These
values
are
similar
to
those
reported
for
the
neutral
fungal
polysaccha-
ride
scleroglucan
(8).
For
comparison,
a
medium-textured
soil
holds
between
0.04
and
0.1
g
of
water
per
g
of
soil
at
-1.0
MPa.
Figure
6
also
shows
the
effect
of
EPS
on
the
moisture
release
curve
of
quartz
sand.
The
addition
of
a
small
amount
of
EPS
greatly
increased
the
amount
of
water
held
by
the
sand
at
all
water
potentials.
The
points
fit
a
curvilinear
function
overall,
but
to
further
analyze
the
effect
of
EPS,
they
can
be
split
at
-0.9
MPa
into
two
linear
portions.
Each
linear
portion
can
be
analyzed
separately
by
linear
regres-
sion.
This
approach
has
the
advantage
of
allowing
a
test
of
the
statistical
significance
of
the
effect
of
EPS
by
using
an
analysis
of
covariance
(Table
1)
(34).
Table
1
shows
the
slopes
of
the
regressions
between
water
potential
and
water
content
for
the
control
and
EPS-amended
sand.
Table
1
also
shows
the
significance
level
(P
value)
for
the
hypothesis
that
the
slopes
differ
between
the
control
and
EPS-amended
Neutral
Polysaccharldes
PS0.02
VOL.
58,
1992
1

1288
ROBERSON
AND
FIRESTONE
FIG.
5.
SEM
micrographs
showing
changes
in
the
amount
and
hydrated
structure
of
EPS
in
Pseudomonas
cultures
growing
on
quartz
sand
at
-0.025
MPa
(A
through
C)
and
after
desiccation
to
-1.0
MPa
(D
through
F).
sand.
Over
both
water
potential
ranges,
the
slope
of
the
regression
line
was
significantly
(P
<
0.1)
greater
for
the
control
sand
than
for
the
EPS-amended
sand.
This
means
that
water
potential
decreased
more
with
decreasing
water
content
in
the
control
sand.
In
other
words,
an
identical
decrease
in
water
content
caused
a
smaller
decrease
in
water
potential
in
the
EPS-amended
sand
than
in
the
control
sand.
(ii)
Effect
of
EPS
on
drying
rate.
The
effect
of
EPS
on
the
APPL.
ENVIRON.
MICROBIOL.

DESICCATION
AND
BACTERIAL
POLYSACCHARIDE
PRODUCTION
1289
0
a-
I..
-4
-1
0
0~
-12
0
a-
2
-2.0
'a
c
0
w-
-4.0
-6.0
*
Pure
EPS
0
200
400
600
800
1000
Water
Content
(%)
B
CDr~t
Control
@00
0
0
EPS
Added
00
Sand
0
2
4
Water
Content
(%)
6
8
FIG.
6.
Relationship
between
water
potential
and
water
content
in
purified
EPS
(A)
and
in
sand
(B)
with
(O)
and
without
(-)
EPS
added.
decrease
of
water
content
and
water
potential
with
time
of
drying
is
shown
in
Fig.
7.
EPS
had
no
effect
on
the
rate
of
decrease
of
water
content
in
the
sand.
This
rate
was
constant
throughout
the
experiment
in
both
the
EPS-amended
and
the
control
sand.
However,
EPS
did
have
a
significant
effect
on
the
rate
of
decrease
of
water
potential
values
of
>
-0.9
MPa
(Table
1).
The
slope
of
the
regression
of
water
potential
against
time
was
significantly
smaller
in
the
EPS-amended
sand.
The
rate
of
water
potential
decrease
in
the
sand
was
therefore
significantly
slowed
by
the
addition
of
EPS
at
values
of
>
-0.9
MPa,
although
the
rate
of
water
content
decrease
was
not
significantly
slowed
(Fig.
7).
The
effect
of
TABLE
1.
Linear
relationship
of
water
potential
with
time
and
water
content
in
drying
sand,
with
and
without
EPS
added,
over
two
water
potential
ranges
Water
potential
Slope'
Variable
Ma
(MPa)
Control
EPS
added
P
value"
Water
content
>-0.9
0.28
0.11
0.07
<-0.9
3.55
1.25
<0.01
Time
>
-0.9
-0.09
-0.04
0.03
<
-0.9
-1.15
-1.30
0.74
"The
slopes
of
the
regression
of
water
potential
against
time
(measured
in
hours)
or
water
content
(measured
as
percentage)
in
the
given
water
potential
range
using
the
data
in
Fig.
6
and
7.
'
Significance
levels
for
the
differences
between
the
slopes
of
the
control
and
those
of
the
EPS-amended
sand.
-1.C
-2.0
-3.0
0L
-4.C
-5.C
-6.C
0.01
c
0.0(
co
@0
0.0;
0
2
4
6
8
Time
(h)
10
12
14
FIG.
7.
Changes
in
water
potential
(A)
and
water
content
(B)
during
14
h
of
drying
in
sand
with
and
without-
EPS
added.
this
is
that
the
control
sand
dried
to
a
water
potential
of
-0.9
MPa
in
approximately
5
h,
whereas
roughly
double
that
length
of
time
was
required
for
EPS-amended
sand.
At
values
of
<-0.9
MPa,
the
slopes
for
the
two
treatments
were
not
significantly
different.
DISCUSSION
Metabolic
response
of
bacteria
to
the
wet-dry
cycle.
Water
availability
strongly
controlled
the
production
and
consump-
tion
of
protein
and
polysaccharide
by
the
bacteria.
Wetting
caused
an
initial
decrease
in
the
amount
of
polysaccharide
in
all
cultures
between
days
0
and
1
(Fig.
4).
This
may
have
been
the
result
of
consumption
of
polysaccharides
by
bac-
teria
growing
in
response
to
the
increase
in
water
availabil-
ity.
The
concurrent
increase
in
protein
concentration
(Fig.
2)
suggests
that
some
polysaccharide
carbon
may
have
been
used
for
protein
production.
Conversely,
in
the
desiccated
treatment
group,
the
amount
of
polysaccharide
after
day
1
increased
while
that
of
protein
decreased,
implying
that
protein
and
possibly
other
cellular
components
as
well
are
used
for
polysaccharide
production
in
response
to
desicca-
tion.
It
should
be
noted
that
some
bacterial
carbohydrate
other
than
EPS
may
have
been
measured
as
EPS
because
of
the
technique
used
to
extract
the
sand
cultures.
The
most
likely
contaminating
carbohydrates
are
compatible
solutes.
Bacte-
ria
have
been
found
to
accumulate
the
sugars
trehalose
and
sucrose
as
compatible
solutes
in
response
to
low
water
potential
(17,
22).
The
amounts
reported
to
be
synthesized
are
too
low
to
affect
the
results
of
the
present
study,
however.
It is
also
possible,
although
less
likely,
that
intra-
cellular
storage
polysaccharides
were
measured
in
the
cul-
tures.
Pseudomonas
species
do
not
generally
accumulate
intracellular
polysaccharides
(28,
35),
but
the
amounts
syn-
thesized
by
organisms
that
do,
such
as
Escherichia
coli,
are
'A
g
|
°*
2B
A0
0
*
o~~~o
Water
Potential
B
o
*
Control
T
8o
B
a
O
EPS
Added
000
0
00
00
0
0
0
4
3
S
o
D
3~~~~~~~~~E
2
Water
Content
of
.
-
I
VOL.
58,
1992
I

1290
ROBERSON
AND
FIRESTONE
also
too
small
to
significantly
affect
the
present
experiment
(31).
Almost
no
glucose
was
detectable
in
the
cultures
during
the
experiment
(Fig.
3).
The
major
sources
of
carbon,
therefore,
were
cell
contents
and
EPS.
This
supports
the
conclusion
that
C
was
shuttled
between
protein
and
polysac-
charides
as
the
water
status
of
the
cultures
changed.
This
lack
of
an
external
source
of
available
C
may
be
similar
to
the
situation
in
soil,
in
which
the
pool
of
available
C
is
often
small
and
microbial
biomass
released
after
wetting
of
dry
soil
can
be
an
important
portion
of
the
C
available
to
the
microbial
community
(20,
40).
The
ammonium
concentration
did
not
change
significantly
after
day
1
for
either
treatment
group,
although
more
ammonium
was
present
in
the
control
cultures
than
in
the
desiccated
cultures
until
day
3.
This
may
reflect
elevated
use
of
ammonium
by
the
bacteria
in
response
to
desiccation
stress.
Electron
transport
activity
does
not
appear
to
have
been
as
strongly
controlled
by
water
potential
as
were
protein
and
EPS.
Activity
was
not
significantly
affected
by
the
treat-
ments
and
showed
a
pattern
of
steady
decrease
in
both
treatment
groups
throughout
the
experiment
(Fig.
2).
This
pattern
may
reflect
the
development
of
a
limitation
of
a
nutrient,
most
likely
C,
in
the
cultures.
Nutrients,
in
the
form
of
dead
cells
and
EPS,
were
relatively
abundant
in
the
cultures
on
day
0
after
the
preexperiment
desiccation
and
day
0
wetting
with
nutrient
solution.
This
finite
supply
was
used
by
the
bacteria,
particularly
between
days
0
and
1,
as
shown
by
the
increase
in
protein
amount
(Fig.
2),
and
limitation
developed.
Additional
limitation
may
have
devel-
oped
in
the
desiccated
treatment
group
as
low
water
content
reduced
solute
diffusion
to
the
bacteria.
This
additional
nutrient
limitation
in
the
desiccated
treatment
may
partially
explain
the
lower
viable
counts
and
protein
contents
mea-
sured
for
those
cultures.
Different
time
periods
are
represented
by
each
of
the
measurement
techniques
used
in
this
study.
The
tests
for
carbohydrates,
glucose,
and
protein
measure
the
cumulative
synthesis
and
consumption
of
these
compounds
before
each
measurement.
The
activity
assay,
on
the
other
hand,
indi-
cates
the
physiological
state
of
the
cultures
at
the
time
of
measurement.
The
tests
must
therefore
be
interpreted
some-
what
differently.
On
day
0,
for
example,
large
amounts
of
polysaccharide
and
low
levels
of
protein
(Fig.
2
and
4)
were
measured.
These
results
reflect
the
drying
cycle
prior
to
the
experiment.
A
high
level
of
water
availability
was
not
shown
in
these
measurements
because
the
cultures
had
not
yet
responded
to
it.
By
day
1,
however,
the
increase
in
protein
and
decrease
in
polysaccharides
showed
the
effects
of
the
wetting.
On
the
other
hand,
on
day
0,
the
level
of
activity
was
high,
reflecting
the
high
nutrient
and
moisture
availabil-
ity
after
wetting
the
dry
cultures
with
nutrient
solution
(Fig.
2).
EPS
water
retention
characteristics.
(i)
Polysaccharide
moisture
release
curve.
The
EPS
produced
by
this
bacterium
held
several
times
its
weight
in
water
at
low
water
potential
(Fig.
6).
The
EPS
was
an
effective
competitor,
relative
to
the
sand,
for
the
limited
supply
of
water
at
low
water
potential.
The
EPS
therefore
has
the
ability
to
retain
water
and
possibly
to
concentrate
dissolved
nutrients
in
the
bacterial
microenvironment
during
desiccation.
By
maintaining
a
high
water
content,
the
EPS
may
also
increase
diffusional
avail-
ability
of
nutrients
to
the
bacterium.
The
effects
of
the
water-holding
capacity
of
the
EPS
can
be
seen
in
the
effect
of
EPS
on
the
moisture
release
curve
of
sand
(Fig.
6).
The
EPS-amended
sand
held
more
water
at
all
water
potentials
than
the
unamended
sand.
The
analysis
of
covariance
results
in
Table
1
show
that
in
addition
to
quantitatively
holding
more
water,
the
EPS-amended
sand
also
lost
significantly
more
water
than
the
control
over
the
same
decrease
in
water
potential.
EPS,
therefore,
can
pro-
tect
bacteria
against
drying
in
two
ways.
It
holds
a
reservoir
of
water
in
the
microenvironment
surrounding
bacteria,
and
it
can
lose
substantial
amounts
of
water
from
this
reservoir
during
desiccation
with
relatively
little
change
to
the
internal
water
potential
in
the
microenvironment.
An
EPS-rich
mi-
croenvironment
surrounds
bacteria
like
a
protective
sponge,
buffering
them
against
external
changes
in
water
potential.
(ii)
Effect
of
EPS
on
the
drying
rate.
The
presence
of
EPS
substantially
slowed
the
rate of
water
potential
decrease
in
the
sand
(Fig.
7;
Table
1).
The
drying
rate
within
a
matrix
of
pure
polysaccharide,
such
as
that
in
a
bacterial
colony
in
soil,
would
be
lower
than
that
measured
in
the
EPS-sand
mixture
shown
in
Fig.
7.
Bacteria
must
maintain
equilibrium
with
the
water
potential
of
their
surroundings.
The
produc-
tion
of
EPS
in
response
to
desiccation
may
provide
signifi-
cant
additional
time
in
which
to
make
metabolic
adjustments
that
allow
bacteria
to
survive
this
environmental
stress.
In
this
experiment,
we
did
not
test
the
effect
of
EPS
on
the
rate
of
wetting.
However,
EPS
may
slow
the
rate
of
water
potential
change
during
wetting
as
well
as
drying
within
the
microenvironment
surrounding
bacteria.
Wetting
occurs
more
rapidly
than
drying
in
soil,
and
sudden
increase
in
water
potential
can
be
an
important
stress
for
soil
microor-
ganisms
(20).
The
production
of
EPS
may
be
as
important
to
bacterial
survival
during
wetting
as
during
the
desiccation
that
precedes
it.
Conclusions.
Many
researchers
have
hypothesized
that
microbial
EPS
protect
microorganisms
from
desiccation.
Although
this
study
does
not
confirm
that
hypothesis,
it
does
provide
evidence
supporting
it.
EPS
can
provide
a
microen-
vironment
that
holds
water
and
dries
more
slowly
than
its
surroundings.
This
property
may
at
least
partially
explain
our
observation
that
bacteria
respond
to
desiccation
by
channeling
energy
and
nutrients
into
polysaccharide
produc-
tion.
Soil
is
an
extremely
heterogeneous
environment,
and
wetting
and
drying
do
not
proceed
uniformly
throughout
it.
Many
microbial
processes
in
soil
depend
on
this
heteroge-
neity.
This
work
shows
that
in
addition
to
passively
taking
advantage
of
the
heterogeneity
in
soil,
bacteria
may
also
intervene
during
desiccation
to
create
a
controlled
microen-
vironment
that
enhances
their
survival.
ACKNOWLEDGMENTS
We
thank
Claire
Chenu
of
the
Station
de
Science
du
Sol,
Institut
National
de
la
Recherche
Agronomique,
Versailles,
France,
for
the
use
of
the
cryoscan
SEM
and
for
her
help
and
advice
during
this
project.
We
also
thank
Anne-Marie
Jaunet,
also
of
the
Station
de
Science
du
Sol,
who
took
the
electron
micrographs
and
whose
expertise
on
the
SEM
was
invaluable.
This
research
was
supported
by
the
Kearney
Foundation
of
Soil
Science
and
by
a
Hatch
project
of
the
University
of
California
Experiment
Station.
REFERENCES
1.
Abacus
Concepts.
1987.
StatView
II.
Abacus
Concepts,
Inc.,
Berkeley,
Calif.
2.
Abacus
Concepts.
1989.
SuperANOVA.
Abacus
Concepts,
Inc.,
Berkeley,
Calif.
3.
Blumenkrantz,
H.,
and
G.
Asboe-Hansen.
1973.
New
method
for
quantitative
determination
of
uronic
acids.
Anal.
Biochem.
54:484-489.
4.
Brink,
R.
H.,
P.
Dubach,
and
D.
L.
Lynch.
1960.
Measurement
APPL.
ENVIRON.
MICROBIOL.

DESICCATION
AND
BACTERIAL
POLYSACCHARIDE
PRODUCTION
1291
of
carbohydrates
in soil
hydrolysates
with
anthrone.
Soil
Sci.
89:157-166.
5.
Bushby,
H. V.
A.,
and
K.
C.
Marshall.
1977.
Some
factors
affecting
the
survival
of
root-nodule
bacteria
on
desiccation.
Soil
Biol.
Biochem.
9:143-147.
6.
Bushby,
H.
V.
A.,
and
K.
C.
Marshall.
1977.
Water
status
of
rhizobia
in
relation
to
their
susceptibility
to
desiccation
and
to
their
protection
by
montmorillonite.
J.
Gen.
Microbiol.
99:19-
27.
7.
Campbell,
G.
S.
1977.
An
introduction
to
environmental
bio-
physics.
Springer-Verlag,
New
York.
8.
Chenu,
C.
1989.
Influence
of
a
fungal
polysaccharide,
scleroglu-
can,
on
clay
microstructures.
Soil
Biol.
Biochem.
21:299-305.
9.
Crowe,
J.
H.,
L.
M.
Crowe,
J.
F.
Carpenter,
and
C.
A.
Wistrom.
1987.
Stabilization
of
dry
phospholipid
bilayers
and
proteins
by
sugars.
Biochem.
J.
242:1-10.
10.
Foster,
R. C.
1981.
Polysaccharides
in
soil
fabrics.
Science
214:665-667.
11.
Foster,
R.
C.,
A.
D.
Rovira,
and
T.
W.
Cook.
1983.
Ultrastruc-
ture
of
the
root-soil
interface.
American
Phytopathological
Society,
St.
Paul,
Minn.
12.
Griffiths,
B.
S.
1989.
Improved
extraction
of
iodonitrotetra-
zolium
formazan
from
soil
with
dimethylformamide.
Soil
Biol.
Biochem.
21:179-180.
13.
Guirard,
B.
M.,
and
E.
E.
Snell.
1981.
Biochemical
factors
in
growth,
p.
79-111.
In
P.
Gerhardt,
R.
G.
E.
Murray,
R.
N.
Costilow,
E.
W.
Nester,
W.
A.
Wood,
N.
R.
Krieg,
and
G.
B.
Phillips
(ed.),
Manual
of
methods
for
general
bacteriology.
American
Society
for
Microbiology,
Washington,
D.C.
14.
Harris,
R.
F.
1981.
Effect
of
water
potential
on
microbial
growth
and
activity,
p.
23-97.
In
J.
F.
Parr,
W.
R.
Gardner,
and
L.
F.
Elliott
(ed.),
Water
potential
relations
in
soil
microbiology.
Soil
Science
Society
of
America,
Madison,
Wis.
15.
Hartel,
P.
G.,
and
M.
Alexander.
1986.
Role
of
extracellular
polysaccharide
production
and
clays
in
the
desiccation
toler-
ance
of
cowpea
bradyrhizobia.
Soil
Sci.
Soc.
Am.
J.
50:1193-
1198.
16.
Hepper,
C.
M.
1975.
Extracellular
polysaccharides
of
soil
bac-
teria,
p.
93-111.
In
N.
Walker
(ed.),
Soil
microbiology,
a
critical
review.
John
Wiley
&
Sons,
New
York.
17.
Hershkovitz,
N.,
A.
Oren,
and
Y.
Cohen.
1991.
Accumulation
of
trehalose
and
sucrose
in
cyanobacteria
exposed
to
matric
water
stress.
Appl.
Environ.
Microbiol.
57:645-648.
18.
Keeney,
D.
R.,
and
D.
W.
Nelson.
1982.
Nitrogen-inorganic
forms,
p.
643-699.
In
A.
L.
Page,
R.
H.
Miller,
and
D.
R.
Keeney
(ed.),
Methods
of
soil
analysis,
part
2,
2nd
ed.
Ameri-
can
Society
of
Agronomy,
Madison,
Wis.
19.
Keston,
A.
S.
1956.
Specific
colorimetric
enzymatic
analytical
reagents
for
glucose,
abstr.
129,
p.
310.
Abstr.
Amer.
Soc.
Chem.
Soc.
Meet.
20.
Kieft,
T.
L.,
E.
Soroker,
and
M.
K.
Firestone.
1987.
Microbial
biomass
response
to a
rapid
increase
in
water
potential
when
dry
soil
is
wetted.
Soil
Biol.
Biochem.
19:119-126.
21.
Killham,
K.,
and
M.
K.
Firestone.
1984.
Salt
stress
control
of
intracellular
solutes
in
streptomycetes
indigenous
to
saline
soils.
Appl.
Environ.
Microbiol.
47:301-306.
22.
Larsen,
P.
I.,
L.
K.
Sydnes,
B.
Landfald,
and
A.
R.
Strom.
1987.
Osmoregulation
in
Escherichia
coli
by
accumulation
of
organic
osmolytes:
betaines,
glutamic
acid,
and
trehalose.
23.
McAneney,
K.
J.,
R.
F.
Harris,
and
W.
R.
Gardner.
1982.
Bacterial
water
relations
using
polyethylene
glycol
4000.
Soil
Sci.
Soc.
Am.
J.
46:542-547.
24.
Measures,
J.
C.
1975.
Role
of
amino
acids
in
osmoregulation
of
nonhalophilic
baceria.
Nature
(London)
257:398-400.
25.
Morse,
S.
R.
1990.
Water
balance
in
Hemizonia
luzulifolia:
the
role
of
extracellular
polysaccharides.
Plant
Cell
Environ.
13:39-
48.
26.
Oades,
J.
M.
1984.
Soil
organic
matter
and
structural
stability:
mechanisms
and
implications
for
management.
Plant
Soil
76:
319-337.
27.
Oades,
J.
M.,
M.
A.
Kirkman,
and
G.
H.
Wagner.
1970.
The
use
of
gas-liquid
chromatography
for
the
determination
of
sugars
extracted
from
soils
by
sulfuric
acid.
Soil
Sci.
Soc.
Am.
Proc.
34:230-235.
28.
O'Leary,
W.
M.
1989.
Practical
handbook
of
microbiology.
CRC
Press,
Boca
Raton,
Fla.
29.
Osa-Afiana,
L.
O.,
and
M.
Alexander.
1982.
Clays
and
the
survival
of
Rhizobium
in
soil
during
desiccation.
Soil
Sci.
Soc.
Am.
J.
46:285-288.
30.
Osa-Afiana,
L.
O.,
and
M.
Alexander.
1982.
Differences
among
cowpea
rhizobia
in
tolerance
to
high
temperature
and
desicca-
tion
in
soil.
Appl.
Environ.
Microbiol.
43:435-439.
31.
Preiss,
J.,
and
D.
A.
Walsh.
1981.
The
comparative
biochemistry
of
glycogen
and
starch,
p.
25-60.
In
V.
Ginsburg
and
P.
Robbins
(ed.),
Biology
of
carbohydrates,
vol.
1.
John
Wiley
&
Sons,
New
York.
32.
Rosacker,
L.
L.,
and
T.
L.
Kieft.
1990.
Biomass
and
adenylate
energy
charge
of
a
grassland
soil
during
desiccation.
Soil
Biol.
Biochem.
22:1121-1127.
33.
Schmidt,
E. L.
1974.
Quantitative
autecological
study
of
micro-
organisms
in
soil
by
immunofluorescence.
Soil
Sci.
118:141-149.
34.
Snedecor,
W.
G.,
and
G.
W.
Cochran.
1980.
Statistical
methods,
7th
ed.
Iowa
State
University
Press,
Ames.
34a.Soroker,
E.
F.
1990.
Ph.D.
thesis.
University
of
California,
Berkeley.
35.
Stanier,
R.
Y.,
J.
L.
Ingraham,
M.
L.
Wheelis,
and
P.
R.
Painter.
1986.
The
microbial
world,
5th
ed.
Prentice-Hall,
Englewood
Cliffs,
N.J.
36.
Trevors,
J.
T.,
C.
I.
Mayfield,
and
W.
E.
Innis.
1982.
Measure-
ment
of
electron
transport
system
(ETS)
activity
in
soil.
Microb.
Ecol.
8:163-169.
37.
van
Doorn,
W.
G.,
F.
Thiel,
and
A.
Boekestein.
1990.
Cryoscan-
ning
electron
microscopy
of
a
layer
of
extracellular
polysaccha-
rides
produced
by
bacterial
colonies.
Scanning
12:297-299.
38.
Wiebe,
H.
H.
1966.
Matric
potential
of
several
plant
tissues
and
biocolloids.
Plant
Physiol.
(Bethesda)
41:1439-1442.
39.
Wilkinson,
J.
F.
1958.
The
extracellular
polysaccharides
of
bacteria.
Bacteriol.
Rev.
22:46-73.
40.
Williams,
S.
T.
1985.
Oligotrophy
in
soil:
fact
or
fiction?,
p.
81-111.
In
M. M.
Fletcher
and
G. D.
Floodgate
(ed.),
Bacteria
in
their
natural
environments.
Academic
Press,
London.
VOL.
58,
1992