Characterization of initial events of bacterial colonization at solid-water interfaces using image analysis
by Robert Franz Mueller
A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in
Environmental Engineering
Montana State University
© Copyright by Robert Franz Mueller (1990)
Abstract:
The processes leading to bacterial accumulation on solid-water interfaces are adsorption, desorption,
growth, and erosion. These processes have been measured individually in-situ in a flowing system and
in real time using image analysis. The flow was laminar (Re = 1.4) and the shear stress was kept
constant during all experiments at 0.75 Nm^-2. Five different substrata -- copper, silicon, 316 stainless
steel, glass, and polycarbonate --and three different bacterial strains -- Pseudomonas aeruginosa.
Pseudomonas fluorescens (mot+) and Pseudomonas fluorescens (mot-) -- were used in the experiments.
The surface roughness varied among the substrata from 25 Å for silicon as the smoothest substratum to
150 A for copper as the roughest substratum. The effects of nutrient limitation and starvation of cells
on sorption phenomena were also examined.
Sorption related processes were found to be influenced by the surface properties of the substrata and by
the nature of the bacterial strain. Adsorption velocity was positively correlated with the surface
roughness of the substratum and the surface hydrophobicity of the cells. While the sticking efficiency
of a non-motile strain was found to be much higher than the corresponding motile strain, the motile
cells adsorbed more rapidly due to superior transport properties arising from motility. Starvation of
Pseudomonas aeruginosa cells was observed to result in a lower sticking efficiency, but increased
motility resulting from dwarfing/fragmentation of the cells. Copper was found to inhibit cell growth on
its surface, but the metal oxide layer on the surface as well as an already formed biofilm protected the
cells and replication was observed after a longer exposure time to cell-containing bulk water. A
mathematical model was developed to predict cell accumulation on a substratum, based on kinetic data
for adsorption, desorption, growth, and erosion, for each of the tested substrata and organisms. O K m C I E R I Z A T I O N O F INITIAL EVENTS O F BACTERIAL COLONIZATION
A T SOLID-WATER INTERFACES USING IMAGE ANALYSIS
by
Robert Franz Mueller
A thesis submitted in partial fulfillment
of t h e requirements for t he degree
Cf
Master of Science
in
Environmental Engineering
MONTANA STATE UNIVERSITY
Bozeman, Montana
June 1990
ii
APPROVAL
of a thesis submitted by
Robert Franz Mueller
This thesis has been read by each member of the thesis committee and
has been found t o be satisfactory regarding content, English usage,
format, citations, bibliographic style, and consistency, and is ready for
submission t o the College of Graduate Studies.
Date
Approved for the Major Department
Approved for the College of Graduate Studies
si
/?,
/??o
7
Hi
STATEMENT OF PERMISSION TO USE
In presenting this thesis in partial fulfillment of t he requirements
for a master's
degree at Montana
State University,
I
agree that
the
Library shall m a k e it available t o borrowers under rules of th e Library.
Brief
quotations
from
this
thesis
are
allowable
without
special
permission, provided that accurate acknowledgement of source is made.
Permission for extensive quotation from or reproduction of this
thesis m a y b e granted b y m y major professor,
D e a n of Libraries when,
or in hi s absence, b y the
in the opinion of either,
t he proposed u s e of
material is for scholarly purposes. A n y copying or u se of t he material in
this thesis for financial gain shall not b e allowed without m y written
permission;
Date
£ / / £ / /?9(y
7
7
iv
ACKNOWLEDGEMENTS
I wo u l d like to thank all XPA/ERC members for providing a n enjoyable
environment and helping m e with this study.
I especially would like t o
thank Bill Characklis for his advice and guidance, Warren Jones and A n n
Camper for their help and numerous discussions, John Sears for making this
study possible, K a y Smith for her help, and Gordon Williams, John R o m p e l ,
a n d Paul Stoodley are thanked for their technical and analytical support,
also. Diane, Linda, Wendy, and Jamie provided assistance a nd moral support
during m y stay in Bozeman.
In particular
I would
like to
express m y
Stover, Roger Colberg, and Marvin B e m t at
appreciation to
John
TM A Technologies, Inc. Bozeman
helped m e w i t h their light scatter measurements and Doug Caldwell from
Vancouver, British Columbia for contributing his Pseudomonas fluorescens.
Last n o t least I would like to thank m y Bozeman climbing friends w h o
gave m e joy, challenge, and a good escape from the laboratory w ho
tired of playing on those "stones" with me.
/
never
V
TABLE OF CONTENTS
Page
L I S T O F T A B L E S .............
v iii
LIST O F FIGURES ......................................................
...................................
I N T R O D U C T I O N .............
Goal of Research
. .
Objectives of Research
xi
U W H
ABSTRACT
BACKGROUND
........................................................ ..
Overview .............................
4
Previous Research
............................................. 4
Surface Free E n e r g y ....................................... 5
Surface R o u g h n e s s ............. ■ ...................... 6
Hydrodynamic F a c t o r s ..........................
7
Chemical Factors
....................
7
. ...........
g
Microbiological Factors ...............
Nutrient Availability - Starvation
. ................... 8
T H E O R Y .................................
..10
Process Analysis a t the Solid-Liquid Interface
. . . . . . 10
Rate Coefficients of
Early Surface C o l o n i z a t i o n ............................ 12
Growth on S u r f a c e s .................... ........... .. . 13
Cell Balance at the Substratum...................... . . . . . 14
Bulk Water Interface
Process Analysis for Chemostat O p e r a t i o n ................... 15
Modified Chemostat Operation for S t a r v a t i o n .......... 16
Transport
........................
18
Diffusivity of Motile C e l l s ............................ 18
Diffusivity of Non-Motile C e l l s ........................ 19
Transport to the Solid-Liquid I n t e r f a c e ............... 20
Adsorption V e l o c i t y ........................
22
Sticking E f f i c i e n c y .............
23
Fluid Dynamics
.................................................24
Reynolds Number . . .............. . . . . . .......... 25
METHODS ...................... . . . . . .
......................... 26
Experimental System
....................................... . 2 6
Met h o d of Analysis
......................................... 29
Total Cell C o u n t s ...............................
29
Viable Cell Counts
.................................... 29
Cell Surface H y d r o p h o b i c i t y .................
29
Total Organic C a r b o n ...............
30
Dissolved Organic C a r b o n .............................. 30
Surface Roughness ...........................
31
vi
TABLE OF CONTENTS (Continued)
Page
Estimation of Cell Motility................ ..
31
Determination of Rate of Adsorption
and Desorption on the Substratum.................. 32
Determination of the Rate of Growth and
Erosion on the Substratum.......................... ..
Image Analysis.......................
33
R E S U L T S ............................... .............. ..........34
Chemostat Operation
.............................. ! ! ! 34
Image P r o c e s s i n g ............ ........................... ..
Cell Surface Hydrophobicity.................... * . . . . 35
Pseudomonas aeruginosa. Kinetic R e s u l t s ..................! 3 6
Surface Colonization ofPseudomonas aeruginosa . . . 36
C o p p e r .................................
36
S i l i c o n ...........................
39
316 Stainless S t e e l ............................
41
Surface roughness...........................
43
Pseudomonas fluorescens (mot+), R e s u l t s .............. .* .* 43
Surface colonization of
Pseudomonas fluorescens
(mot+) . .................... 43
316 Stainless S t e e l ................................. 44
Polycarbonate............. ....................
45
G l a s s ...............................
45
Surface Roughness.................... ..............47
Results with Pseudomonas fluorescens (mot-) ...............47
Surface Colonization of
Pseudomonas fluorescens (mot+)............ ; . . . 47
316 Stainless Steel ...................................
Polycarbonate........ ......... ...................48
G l a s s ........ ............................. . . . . . ’ 48
Surface Colonization and Dislocations
on Stainless S t e e l .............. ........................50
Pseudomonas aeruginosa. Starvation
and Adsorption to 316 Stainless Steel .....................51
Dwarfing and Fragmentation . .......................52
Surface Colonization of Starved
Pseudomonas aeruginosa on 316 Stainless Steel . . . . 54
Cellular Motility.................................. ..
SIMULATION OF CELL ACCUMULATION................................ 57
Bulk Water Cell Concentration............................ 57
Fluid Shear S t r e s s ................. ......................57
Substrata and Bulk Water Nutrient Concentration.......... 58
Mixed Populations.................................. ..
58
vii
TABLE O F CONTENTS (Continued)
Page
D I S C U S S I O N ..............................
Comparison of Pseudomonas
aeruginosa on Various Substrata
............................
Comparison of Pseudomonas aeruginosa
a n d Pseudomonas fluorescens (mot+)
. ........................
Comparison of M o t + and Mot- Pseudomonas fluorescens
. ..
Influence of Starvation on Colonization of
Pseudomonas aeruginosa on 316 Stainless Steel
.............
72
C O N C L U S I O N S .................
75
62
62
67
70
N O M E N C L A T U R E / S Y M B O L S .......................................... .. . 77
N o m e n c l a t u r e ................................................. ..
Symbols
..................................................... ..
LITERATURE C I T E D ..............................................
81
A P P E N D I X ...................................
Cell Accumulation at Various
Durations of Starvation
...........
86
87
viii
LIST O F TABLES
Table
1.
2.
3.
4.
5.
6.
7.
Page
Chemical Conposition of Nutrient
.................... 27
System Specific P a r a m e t e r s ........................ ..
27
Cell surface Hydrophobicity
. '........................... 35
Pseudomonas aeruginosa.Summary of the R e s u l t s .......... 43
Pseudomonas flourescens (mot+), Summary of the Results . . 47
Pseudomonas flourescens (mot-), Summary of the Results . . 50
Starvation of Pseudomonas aeruginosa.
Summary of the Results
........... ........................ 73
ix
LIST OF FIGURES
Figure
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Page
Definition of processes during early
colonization of a s u b s t r a t u m ................................. ..
Experimental system
.......................................... 26
Results from a laser light scatter measurement o n copper. . . 37
Cell accumulation of Pseudomonas aeruginosa on copper . . . . 38
Cell accumulation of Pseudomonas aeruginosa o n silicon. . . . 40
Results from a laser light scatter measurement o n stainless . 41
Cell accumulation of Pseudomonas aeruginosa o n stainless. . . 42
Adsorption velocity of Pseudomonas
aeruginosa versus surface roughness .........................44
Cell accumulation of Pseudomonas
fluorescens (mot+) on stainless ............................ 45
Cell accumulation of Pseudomonas
fluorescens (mot+) on polycarbonate........................ 46
Cell accumulation of Pseudomonas
fluorescens (mot+) on g l a s s ................................ ..
Cell accumulation of Pseudomonas
49
fluorescens (mot-) on stainless...................... ..
Cell accumulation of Pseudomonas
fluorescens (mot-) on polycarbonate .......... ..............49
Cell accumulation of Pseudomonas
fluorescens (mot-) on g l a s s ................................ ..
Photograph, plate count on R2A agar ........................ 52
Cell Concentration Versus Duration of Starvation............ 53
Cell Size Distribution Versus Duration of Starvation........ 53
Adsorption Velocity Versus Duration of Starvation.......... 55
19. Probability of Desorption versus Duration of Starvation . . . 55
20. Simulation of cell accumulation of Pseudomonas aeruginosa
o n stainless a t various bulk water cell concentrations. . . . 59
21. Simulation of cell accumulation of Pseudomonas aeruginosa
o n stainless at various shear stress........................... 59
22. Simulation of cell accumulation of Pseudomonas aeruginosa
o n stainless, copper, and silicon at rich n u t r i e n t ......... 60
23. Simulation of cell accumulation of Pseudomonas aeruginosa
o n stainless, copper, and silicon at low n u t r i e n t ........... 60
24. Simulation of cell accumulation of a mixed population . . . . 61
25. Comparison of adsorption velocity of
Pseudomonas aeruginosa on various substrata . . .............. 63
26. Comparison of reversible
a n d irreversible adsorption v e l o c i t y ........................ 68
27. Cell surface h y d r o p h o b i c i t y ................................... 68
28. Cell accumulation,
chemostat effluent D = 0.2 h r '1 ........... .. . .............. 88
29. Cell Accumulation,
48 Hours of Starvation
88
X
LIST OF FIGURES (Continued)
Figure
Page
30. Cell Accumulation,
72 Hours of Starvation. ........................................ 89
31. Cell Accumulation,
120 Hours of S t a r v a t i o n ......................................... 89
32. Cell Accumulation,
168 Hours of S t a r v a t i o n ........... ................. -.......... 90
33. cell Accumulation,
216 Hours of S t a r v a t i o n ...................
90
34. Cell Accumulation,
348 Hours of S t a r v a t i o n ......................................... 91
35. Cell Accumulation,
504 Hours of S t a r v a t i o n ......................................... 91
36. Cell Accumulation,
648 Hours of Starvation ........................................ 92
37. Cell Accumulation,
816 Hours of S t a r v a t i o n ......................................... 92
38. Cell Accumulation,
936 Hours of S t a r v a t i o n ..........................
93
39. Cell Accumulation,
1224 Hours of S t a r v a t i o n ...............
93
xi
ABSTRACT
T h e processes leading t o bacterial accumulation o n solid-water
interfaces are adsorption,
desorption,
growth,
a nd erosion.
These
processes h a v e been measured individually in-situ in a flowing system a nd
in real time using image analysis. The flow was laminar (Re = 1.4) a nd th e
shear stress w a s k e p t constant during all experiments a t 0.75 Mm"2. Five
different substrata — copper, silicon, 316 stainless steel, glass, and
polycarbonate — and three different bacterial strains —
Pseudomonas
a e r uginosa. Pseudomonas fluorescens (mot+) and Pseudomonas fluorescens
(mot-) — w e r e used in the experiments. The surface roughness varied among
t h e substrata from 25
for silicon as the smoothest substratum t o 150 A
for copper as t h e roughest substratum. The effects of nutrient limitation
a n d starvation of cells on sorption phenomena were also examined.
Sorption related processes were found to b e influenced b y the
surface properties of the substrata and b y the nature of t he bacterial
strain.
Adsorption velocity was positively correlated w i t h the surface
roughness of the substratum and the surface hydrophobicity of the cells.
While t h e sticking efficiency of a non-motile strain w a s found t o b e m u c h
higher than the corresponding motile strain, the motile cells adsorbed
m o r e rapidly due to superior transport properties arising from motility.
Starvation of Pseudomonas aeruginosa cells was observed t o result in a
lower
sticking efficiency,
but increased motility resulting from
dwarfing/fragmentation of the cells. Copper was found t o inhibit cell
growth o n its surface, but the metal oxide layer o n t h e surface as well
as a n already formed biofilm protected the cells an d replication w as
observed after a longer exposure tine t o cell-containing bulk water.
A
mathematical model was developed t o predict cell accumulation o n a
substratum, based o n kinetic data for adsorption, desorption, growth, and
erosion, for each of the tested substrata and organisms.
A
1
INTRODUCTION
T h e individual processes contributing t o biofilm accumulation and
activity m u s t b e understood and quantified in order t o provide rational
methodologies for their control or enhancement. A crucial initial event
in
biofilm
accumulation
substratum.
If
conditions,
these
is
the
adsorbed
they will replicate,
thickness of that
adsorption
cells
find
bacteria
suitable
cells
at
a
environmental
grow, and finally form, a biofilm. T he
film can vary from a
single cell
h u n dred millimeters as found in algae mats
only cause biocorrosion but can also
exchangers,
of
layer to
(Escher 1986).
Biofilms not
increase the resistance
reduce flow rate in pipes,
several
in heat
increase t he d r a g of a ship, or
cause pathogenic problems in municipal water supplies a nd food processing
(Charadklis and Marshall, 1990). In waste water treatment, biofilms are
u s e d as mat r i x for immobilizing microorganisms
1990).
Cell
concentrations
magnitude higher
than
in
those
biofilms
are
for planktonic
(Stevens arid Crawford,
three
cells.
to
four
orders
IO 12 cells
of
c m "3 is
within biofilms, whereas IO 8 to IO 9 cells m l '1 are the
commonly found
m a x i m u m concentration for cells in suspension (Characklis a nd Marshall,
1990). Hence,
biofilm systems allow not only higher dilution rates but
also higher loading rates than conventional water or waste water treatment
systems.
Bacterial
availability
of
growth
in
carbon
many
or
natural
other
substrates
oligotrophia environments,
the flux of
importance
al.
(Kjelleberg et
1983).
habitats
is
limited
(Marshall
by
th e
1985).
In
substrates t o t h e cells
An
oligotrophia
is of
environment w as
2
(1981)
as a habitat with flux of
I g C m "3 d a y "1
In these
low nutrient environments,
attachment of
defined b y Pointexter
substrate o r
less.
organisms t o surfaces m a y b e of particular advantage d ue t o increased flux
of nutrients to immobilized cells in comparison t o planktonic organisms
(Kjelleberg
&
Hermansson
1984).
If
nutrient
sufficiently low, starvation m ay occur (Morita 1982).
b e asked are:
concentrations
are
Questions that m a y
Is adsorption to surfaces of any significance to starved
bacteria? H o w is adsorption of a specific bacterium influenced b y the
duration of starvation? Is adsorption used as a strategic response t o
starvation for increasing the chances of survival b y increasing nutrient
flux?
Measuring only the accumulation of cells does not provide sufficient
information t o control, predict, or simulate a specific system. Therefore,
bacterial colonization on a substratum has to b e distinguished between t he
single contributing processes, such as adsorption, desorption, growth, and
erosion. W i t h a knowledge of the rate coefficient and reaction order for
each of these processes, colonization at the solid-liquid interface can
b e characterized and simulated. Then,
predict
cell
accumulation
environmental conditions.
on
simulation models c an b e used to
different
substrata
under
various
These models for initial colonization can b e
combined wi t h existing models for mature biofilms, so that the complete
development of a biofilm could be predicted and simulated.
This study continued the work of Escher (1986). Escher used image
analysis t o determine the influence of shear stress o n early bacterial
surface colonization in a laminar flow system at various bulk water cell
concentrations. Pseudomonas aeruginosa
was the test organism and a smooth
I
3
glass surface was the substratum.
Goal of Research
Characterize early bacterial colonization in a continuous flow
system for various solid-water interfaces, environmental conditions, and
microbial species.
Objectives of Research
I.
Determine t h e rate of early bacterial colonization of Pwehdrrnionaf=;
aeruginosa
for
various
solid-water
interfaces,
such
as
copper,
silicon, 316 stainless steel, glass, and polycarbonate.
2.
Conpare the rate of colonization of two different bacterial species
(Pseudomonas
aeruginosa and Pseudomonas
flagellated
cells
(mot+)
and
fluorescens) as well
unflagellated
cells
(mot-)
as
of
Pseudomonas fluorescens.
3.
Determine the relative effects of rich nutrient conditions w i t h
conditions of starvation on colonization rates.
4.
Mathematically simulate early surface colonization a t various solid
interfaces, shear stress, and nutrient conditions for pure cultures
a n d m i x e d populations.
4
BACKGROUND
Overview
T h e number of cells accumulating at the substratum will depend on
m a n y variables whi c h can b e categorized as surface properties a t t he
solid-water
interface,
microbiological.
liquid,
the
hydrodynamics,
Temperature,
presence
of
pH,
trace
water
chemistry,
substrate concentration
elements
and
other
and
in the bulk
nutrients
such
as
phosphate and nitrogen, and their ratio to carbon, the ionic strength of
t h e b u l k liquid, the coating of the surface, the substratum material, the
substratum free surface energy (as reflected b y hydrophobicity and surface
roughness), as well as cellular properties might all influence the rates
processes leading t o bacterial accumulation on a solid-liquid interface
(Abbott e t al.
(1983), Hermansson and Marshall
(1986), Paul and Jeffrey (1985), Fletcher,
(1985), Pedersen e t al.
(1980)).
Previous Research
T h e results reported b y various researchers related t o early events
of bacterial surface colonization are not always consistent. Almost all
research found in literature on early bacterial colonization o n solid liquid
interfaces was
done very carefully with respect t o microbial,
chemical, biochemical,
and physico-chemical aspects. Unfortunately most
of
the
reported
research .did
not
differentiate
between
adsorption,
desorption, or growth o n surfaces, but measured only cell accumulation.
5
V e r y little w o r k can b e found on hydrodynamic influences o n harrt-e>rial
colonization.
Bowen, Iavine and Epstein (1976) proposed a n analysis t o describe
t h e cellular transport rate to the substratum for a continuous flowing
system a n d laminar flow conditions. Escher (1986), as well as Powell and
Slater (1984), u s e d image analysis t o study the colonization behavior of
bacteria cells o n a smooth glass substratum under laminar flow. Powell and
Slater u s e d
Bacillus
cereus
to
study
cellular
adsorption
in a
glass
capillary a t laminar flow and compared their experimental results with
theoretically calculated deposition rates for non motile cells when using
Bowen's model. Escher used Pseudomonas aeruginosa as a test organism and
var i e d b u l k water cell concentration and fluid shear stress and determined
the
kinetic
Bowen's
rates
model,
calculated.
responsible
sticking
Escher
for
efficiency
concluded
that
early
and
cell
colonization.
particle
sorption
capture
related
By
factor
processes
using
were
can
be
described as first order rates with respect to cell concentration in the
b u l k liquid and zero-order rates with respect to surface cell density.
Growth related processes were found to be first order w i t h respect t o
surface cell density.
Surface Free Energy
Inorganic solids with high melting points such a s glass, metal, o r
silicon h a v e h i g h surface free energies (500 - 5000 e rg cm'2) . Soft organic
solids w i t h low melting points usually have surface energies less than 100
e r g c m '2 (Zisman 1964). The m ain forces which contribute t o the surface
free
energy
are
dispersion,
dipole,
electrostatic,
and
metallic
6
interactions.
Electrostatic
adsorption
charge
(Fletcher,
may
1980).
also
be
of
importance
for
bacterial
The adsorption of organic substances m a y
reduce t h e surface free energy. Adsorption of any organic material o n
surfaces will b e determined b y the free surface energy of t he solid phase
and t h e surface tension of the liquid phase (Fletcher, 1980). Similarly,
adsorption and movement of bacteria at solid-liquid interfaces should be
influenced b y interfacial tension and bacterial surface properties. The
free energy of solids can b e determined b y measuring t he contact angle of
wa t e r
to
the
surface
(B) ,
also
called
hydrophobicity
of
the
wetted
surface.
Surface Roughness
Surface roughness is mainly a function of the surface history (the
way
the
surface
was
worked:
machine
worked,
mechanically
polished,
electrochemically polished, etched e t c ) . The cosine of t he contact angle
(0 ) of t h e solid interface to water (surface hydrophobicity)
proportional t o the
surface roughness
(Bikerman,
1970).
is directly
Every groove,
val l e y o r scratch on a solid interface acts as a capillary tube in wh i c h
t h e liquid rises as long as 0<9O° or descends when 0>9O°. Thus a rough
surface tends t o exaggerate the wetting behavior of a solid
I960). W h e n the contact angle of a smooth surface is small,
(Adamson,
it is even
smaller o n a rough solid of identical material.
Vanhaecke et al.
strains
in
a
stagnant
(1990) tested six different Pseudomonas aeruginosa
system
and
measured
cellular
accumulation
on
different polished stainless steel surfaces (electropolished and 120 grit
7
treated)
after 30 minutes of exposure.
The authors reported a drastic
influence of surface roughness to the rate of cellular adsorption for
Pseudomonas aeruginosa strains with a low cell surface hydrophobicity.
F o r t h e strains w i t h high hexadecane partitioning values th e influence of
surface roughness was reported to b e minimal.
Hydrodynamic Factors
Escher (1986) was one of the few studies found whi c h distinguished
v e r y carefully between adsorption,
desorption,
growth,
a nd erosion t o
describe cell accumulation. H e used Pseudomonas aeruginosa in a laminar
flow system and varied fluid shear stress while keeping all the other
parameters
constant.
Shear
stress
controlled the
process
of
cellular
colonization w i t h a negative exponential influence o n t h e rate of cell
adsorption and cell separation. The rate of desorption w a s found t o b e
independent of shear stress.
Chemical Factors
Calcium and magnesium ions positively influence bacterial adsorption
(Turakhia 1986).
Other complex organic compounds like EDTA or polymers
h a v e be e n shown t o inhibit microbial adhesion (Marshall, 1972) simply b y
converting a favorable hydrophobic surface into an unfavorable hydrophilic
surface. Stanley (1983) found an adsorption maximum at stainless steel for
viable cells, both motile and non-motile at p H 7 to 8 , whereas n on viable
cells adsorbed best at low p H values (pH 2 to p H 3). T h e same author found
a n increase in adsorption for Pseudomonas aeruginosa t o stainless steel
w i t h increasing
CaCl 2 or NaCl concentration in the bulk water. Varihaecke
8
et. al.
(1990) determined cell accumulation of Pseudomonas aeruginosa on
stainless
steel.
They
found
cell
surface
hydrophobicity
and
surface
roughness o n the substratum were the major parameters influencing t he rate
of cellular adsorption but found n o correlation between salt concentration
a n d adsorption.
Microbiological Factors
Vanhaedce et al.
(1990) reported a curvelinear positive correlation
between cell surface hydrophobicity and cellular adsorption t o 316-L and
304 stainless steel for 15 different Pseudomonas aeruginosa strains. Cell
surface hydrophobicity w a s measured as bacterial adherence t o hydrocarbons
(BATH-test). Also,
bacterial
available t o the organisms.
adsorption
is
affected
by
the
nutrient
Kjelleberg, Marshall a nd Hermansson
(1985)
reported a relatively higher rate of adsorption t o smooth glass substrata
for marine bacteria grown in a low nutrient environment than for those
grown in h i g h nutrient environments. However, n o significant differences
w e r e reported for the tendency of irreversible binding of the cells to
surfaces.
Bacterial adsorption m a y also depend on t he type of limiting
nutrient (e.g. carbon, oxygen, phosphate or nitrogen)
(Brown et al. 1977).
Nutrient Availability - Starvation
In low nutrient environments bacteria on surfaces have selective
advantages for growth and survival over planktonic cells. Cells need not
b e firmly attached to a substratum in order to benefit from the nutrient
enriched
state
at
the
solid-liquid
continuous flowing systems,
interface
(Marshall
1978).
In
a n additional benefit is a higher nutrient
9
flux for adsorbed bacteria because there is no diffusion limitation of
nutrient uptake. Marshall also reported regrowth to a normal size of small
dwarf like bacteria within 12 t o 20 hours after surface colonization.
Kjelleberg
(1983)
reported
fragmentation
a nd
continuous
size
reduction (dwarfing) as a response t o the initial phase of starvation. In
K j e l l e b e r g 1s experiment, the irreversible adsorption of Pseudomonas sp.
strain S9 (hydrophobic) and Vibrio sp. strain DWl (hydrophilic) increased
during 4 hours of starvation. Kjelleberg (1984) d id a similar experiment
w i t h seven marine isolates and found a small increase in irreversible
adhesion after 5 hours followed b y a small decrease after 22 hours of
starvation for
six strains.
remained constant.
For one strain,
the
irreversible binding
10
THEORY
Process Analysis at the Solid-Liquid Interface
Mathematical modeling attempts t o describe real system behavior w i t h
a set o f mathematical equations. Abiotic variables describe the physical
a n d chemical environment in which the microbial process
Process Analysis combines these variables in mass,
balance
equations
coefficients
to
with
stoichiometry,
form predictive models.
reaction
These
is occurring.
energy and momentum
rate,
a nd
transport
coefficients vary w i t h
environmental parameters, such as temperature, pH, concentration of trace
elements, the composition and balance of essential nutrients and its ratio
t o carbon, t h e ionic strength, the presence or absence o f specific ions,
such as calcium and magnesium.
There are several processes occurring during cellular colonization
on
a
solid-liquid
biofilms,
growth,
controlling
interface which must
be
erosion,
and
processes.
detachment,
Initial
events
in
distinguished.
sloughing
biofilm
For mature
are
the
rate
accumulation
are
substratum conditioning b y organic molecules and subsequent adsorption of
cells (Characklis and Marshall, 1990). Some of these adsorbed cells will
stay o n t h e surface (irreversibly adsorbed cells) while others m a y desorb
after some time (reversibly adsorbed cells). Kefford a nd Marshall (1984)
suggested a model for the irreversible cell adsorption of Leptospira,
wh e r e
reversible
substratum
after
mechanisms
were
adsorbed
a
cells
certain
found
to
be
time
become
irreversible
period.
chemically
The
cell
different
for
adsorbed
surface
to
a
binding
reversibly
a nd
11
irreversibly
adsorbed
cells.
Escher
(1986)
reported
a
time
interval
between initial adsorption and desorption of Pseudomonas aeruginosa cells
of 0.93 h r and a decreasing probability of desorption over the residence
time of adsorbed cells with a high probability of desorption during the
first 5 t o 10 minutes. If a cell remained longer than 0.93 hours adsorbed
o n a substratum it was termed irreversibly adsorbed. Irreversibly adsorbed
cells
will
replicate
under
suitable
environmental
conditions
and
the
daughter cells m a y either desorb - erosion - or also become irreversibly
adsorbed .
TRANSPORT
ADSORPTION
MULTIPLICATION
EROSION
-- -
C F U - SEPARATION
Figure I. Definition of
substratum (Escher 1986)
DESORPTION
processes
during
early
colonization
of
a
12
The
processes
occurring at
the
solid-liquid
interface
could b e
distinguished between (Figure I ) :
- Transport from the bulk water phase to the solid-liquid interfac
- Adsorption o n the substratum
- Desorption from the substratum t o the bulk wa t e r phase
- Replication of irreversible adsorbed cells
- Erosion of cells from adsorbed colonies t o t he b u l k water phase
R a t e Coefficients of Early Surface CnlnniyAtinn
The
equations
rate
to
coefficients
express
the
were
determined
quantitative,
by
using
individual
differential
relationships
of
adsorption, desorption growth, and erosion at the interface:
ra
= KaAXb'
(i)
rg
= K g A X 11
(2)
re
= Ke*X''
(3)
r a irr =
r r *1^
*^1
f,T r
= ( I " b)
f Irr
=
.
(4)
(5)
X1
T r"
-----X
A rev
vX''
Xrev' '
X irr"
r
Ka
Kg
Ke
^irr
b
(6)
11
cell concentration in the bulk water [cell# L'3]
cell surface density at time t [cell# L'2]
cell surface density of reversibly adsorbed cells at time t
[cell# L"2]
cell surface density of irreversibly adsorbed cells a t time
t [cell# IT2]
rate of adsorption, growth, or erosion, respectively
[cell# L '2 f 1]
adsorption velocity [L t'1]
coefficient of growth o n the surface [t"1]
coefficient of erosion [t'1]
fraction of irreversibly adsorbed cells [-]
probability of desorption [-]
13
t:
time [t]
Equations I, 2, 3, and 4 assume first order rates (Escher 1986) for
adsorption, desorption, growth, and erosion. For growth a nd erosion, the
rates are dependent of cell surface density (Xm ) a nd a re independent of
bulk
water
desorption
cell
are
concentration
dependent
of
(Xb').
The
bulk water
rates
cell
of
adsorption,
concentration
and
(Xb')
t ut
independent of cell surface density ( X " ) .
Growth o n Surfaces
Most
populations
in
uninhibited
systems
grow
exponentially
as
described b y Equation 2 (assuming n o adsorption or desorption):
X 1' = X 0 V 1Aeflt
X 11
X 0'1
P
An
:
:
:
(7 )
cell surface density at time t [cell# L'2]
initial cell surface density [cell# L"2]
specific growth rate [t'1]
important
constant
for
an
exponentially
growing
system
is
the
generation time or doubling time g ( X " = 2X 0 '').
. In 2
g
g
=
----
(8)
: time necessary to double the population [t]
F r o m experimental results with Pseudomonas aeruginosa in biofilms,
Bakke e t al.
(1984) suggested that the growth rates of organisms in the
biofilm
in
and
the
planktonic
state
are
the
same
as
long
as
the
microenvironment of the cells is the same. This should also b e valid for
adsorbed cells during early surface colonization. T he specific growth rate
14
has
a
hyperbolic dependency on
substrate concentration wh i c h
is m o s t
commonly described b y the Monod equation.
AtUiax S
Ks + S
Pmax
Ks
S
:
:
m a x imum growth rate
[t"1]
cellular growth saturation constant
substrate concentration
[M L'3]
[M L'3]
K s = 2.0 g m "3 for Pseudomonas aeruginosa grown o n glucose (Robinson
1984). W h e n the maximum growth rate on a surface is known.
Equation 9
could b e u s e d t o calculate the specific growth rate for various nutrient
conditions.
Cell Balance a t the Substratum-Bulk Water Interface
T h e observed accumulation of cells on a substratum is dependent o n
adsorption rate, desorption rate, generation rate, an d erosion rate:
accumulation
rate
=
adsorption - desorption + generation - erosion
rate
rate
rate
rate
(10 )
T h e rate of cell accumulation on a substratum c an b e expressed w i t h
t h e rate coefficients of sorption, growth and erosion:
dX' '/dt
=
fjrr*Ka*Xb'
+
Kg*X"
-
(11 )
Ke*X"
Equation 11 does not have individual terms for death a nd decay of
t h e adsorbed cells. These rates are included in growth a n d erosion. The
rates
describing
bacterial
colonization
are
measured
at
constant
15
environmental
conditions as rate of growth,
rate of
erosion,
rate of
adsorption, and desorption. The kinetic coefficients w e r e determined byregression analysis.
The rate coefficients determined are not maximum
rates b u t dependent o n temperature,
substratum, nutrient concentration,
hydrodynamics in a system, e t c . . Equation 11 was u s e d a s basic equation
for a ll t h e surface colonization simulations presented in this study.
Process Analysis for Chemostat Operation
In a chemostat, mixing is assumed to b e ideal a nd biofilm growth on
t h e walls
condition,
is neglected.
When operating a chemostat under steady state
the number of
cells
in the
effluent
is
constant.
Thus,
a
chemostat w a s chosen t o control cell concentration for all experiments in
this study. W i t h dilution rate smaller than the maximum growth rate, a n
ideal chemostat can b e described b y the following equations, assuming a
sterile influent (X1
- = 0) and steady state conditions:
n
= Q/V
X = Y* (S1
- — S)
>H X CO CO > a
i
:
:
:
:
:
:
(12)
(13)
yield [M M'1]
biomass concentration in chemostat [M L'3]
substrate concentration in chemostat (M L'3]
influent substrate concentration [M L'3]
volume of chemostat [L3]
influent/effluent flowrate
[L3 t'1]
T o convert biomass into cell numbers a conversion,
size a nd cell density has to be used:
X ~
X c' Pcell V ac
including cell
(14)
16
X c'
Pceii
V ac
: chemostat effluent cell concentration [cell# L'3]
: average cellular density [M L'3]
: average cell volume [L3]
T h e cellular density is fairly constant for most o f the pseudomonads
with
a
wet
cell
density
of
pceU
=
l.004*10"3 g
m m '3
(Characklis
a nd
Marshall, 1990). T h e average cell volume wa s determined b y image analysis
for Pseudomonas aeruginosa and Pseudomonas fluorescens
V c(av) =
For
starved
cells
of
Pseudomonas
Pseudomonas fluorescens;
0.675 ± 0.105 /zm3
aeruginosa, a nd
v c(av) =
(p = 0.2 hr'1) :
starved
cells
of
0.355 + 0.095 p m '3
Modified Chemostat Operation for Starvation
T o study starvation of microbial cells, the chemostat influent flow
w a s terminated. Therefore, the chemostat w as changed into a batch reactor.
Inactive microbial mass was generated because some cells died and lysed
as a result of endogenous metabolism and did not contribute t o substrate
removal.
Death and lyse rate are generally treated as first order w i t h
respect t o viable and dead cell biomass, respectively. Processes occurring
in a batch reactor w i t h only one bacteria species can b e described b y t he
following equations, assuming substrate removal is attributed only t o the
active, viable cells and also assuming that only dead cells will lyse:
dS
P Xv
Substrate balance:
(15)
dt
Y
dXt
Biomass balance:
(total cells)
P
X v - k t (Xt " X v)
(16)
dt
dXv
Biomass balance:
(viable cells)
(p
dt
- kj) X v
(17)
17
d ( X t - X v)
Biomass balance:
----------(non-viable cells) .
dt
Xt
Xv
kd
kL
:
:
:
:
=
Jcd X v
- R 1 (Xt - X v)
(18)
total biomass concentration [M L'3]
viable biomass concentration [M L'3]
death rate w i t h respect to viable cell concentration [t"1]
lyse rate w i t h respect to non-viable cell concentration [ f 1]
The
specific
growth
rate
(fi)
in
biochemical
systems
is
best
described b y the empirical Monod equation (Equation 9). A t low substrate
concentrations, the specific growth rate is directly
proportional t o the
substrate concentration but independent of it at high concentrations. In
t h e case of starvation ( S « K s) , a first order expression results:
/^max S
M = -------Ks
(19)
A f t e r an even a short starvation period,
lysed cells would b e th e
only available carbon source, other additional carbon sources having been
depleted in the system. W i t h the assumption that all t he carbon from lysed
cells
is
available
as
nutrient
for
the
remaining
viable
cells
the
substrate balance (Equation 15) can b e modified as follows:
dS
---- -dt
n Xv
-------- + F Iq (Xt - Xv)
Y
(20)
F : gravimetric cell factor, carbon content per biomass [M M'1]
Assuming that the change in substrate concentration will approach
zero after a long duration of starvation, Equation 20 c an b e solved for
lyse rate (kt) b y substituting Equation 19 for
16 c a n be solved for death rate. (Icd) :
/i. Simultaneously,
Equation
18
/ i Itiax
S
(21)
kI=
Ks Y F
AiHiax
(Xt - X v)
S
I
dXv
Xv
dt
(22)
k Ci =
Ks
W i t h Equation 21 and 22, lyse rate and death rate c an b e determiner!
b y following only three parameters versus duration of starvation: viable
a n d total cell concentration and the substrate concentration (dissolved
organic
carbon)
in
the
water
phase.
With
Equation
14,
concentration can b e converted into cell concentration.
Transport
Diffusivitv for Motile Cells
Diffusion
is
the
main
transport
process
perpendicular
to
the
direction of flow for small particles such as bacteria cells under laminar
flow conditions.
Jang
and Yen
(1985)
used the
following equation
to
calculate diffusivity for various organisms:
vr * dr
D c -- -----------------------------------------
(23)
3 * ( I - C O Sa)
Dc
vr
dr
a.
:
:
:
:
diffusivity [L2 t'1]
velocity of motility [L f 1]
free length of random run [L]
m a i n angle of turn; assuming n o chemotaxis cosa = o
Equation 23 enables the calculation of non-Brownian diffusivity from
cellular motility data for a specific species. Escher (1986) pointed out
that diffusion b y Brownian motion can b e neglected relative t o motility.
19
H e calculated n o n Brownian diffusivity of Pseudomonas aeruginosa
data obtained b y Vaituzi and Doetsch (1969)
(vr = 55.8
/m
from
s'1;
d r = 50-85 /mi) as IO "3 m m 2 s'1.
Kim
(1990)
determined
diffusivity
of
Pseudomonas
experimentally, b y using a capillary tube method.
w h i c h w a s derived b y Segel et al.
capillary
tube
with
a
small
aeruginosa
H e u s e d a n equation
(1977) assuming that for a long enough
cross-sectional
area,
diffusion
c an
be
described b y a material balance in one dimension:
TT N 2
Dc
(24)
4 X 0'2 A 2 t
N
X 0'
A
: total number of cells in the tube [cell#]
: suspended cell concentration [cell# L"3]
: cross-sectional area of the tube [L2]
K i m reported a value of 2.1*10 '3 m m 2 s '1 and a standard deviation of
1.3*10 3 m m 2 s 1 for n o n Brownian diffusion of Pseudomonas aeruginosa.
Therefore, the values obtained from the calculation using Jang an d
Yen's equation match well with the experimental results from K i m using
Segel's equation for a capillary tube.
Diffusivity of Non-Motile Cells
T h e Stokes-Einstein equation yields good estimates for diffusion
coefficient of small particles or non motile bacteria cells b ut cannot b e
u s e d for motile cells:
Icb T
Icb T
D
f.
(25)
f
6 TT /Ab R a
20
Jcb
T
Vb
Ra
f
: Boltzman constant, 1.38062*10'23 J K '1
: Temperature
[T]
: Viscosity of suspended medium [N t L'2]
: Cell radius
[L]
: cell factor [N t L'2]
For r o d shaped non motile cells with an average volume of 0.675 /rni3
a n d a diameter t o length ratio of 0.45
(for Pseudomonas species),
t he
diffusivity calculated by the Stokes-Einstein Equation a t 22°C is
3.9*10 "7 m m 2s"1. Thus, calculated diffusivity for motile cells is almost 4
orders of magnitude higher than the calculated diffusivity for non-motile
cells of t h e same size and shape.
Transport t o the Solid-Liquid Interface
Bowen
et
al.
(1976)
proposed
an
analysis
to
describe particle
deposition from a suspension at fully developed laminar flow in a parallel
plate channel. They assumed a first order reaction rate a t the wall and
calculated the transport of small particles t o a solid-liquid interface.
T h e particle deposition rate is expressed fcy the following equation which
converges well for large Peclet numbers:
X b' D c
K a' • ---------h
K a"
X b'
Dc
h
K1
e
r
(2/9Ki)1/3
----------------------------[F(4/3) + (I/ e) *(2/9K,)1/3]
: flux of particles adsorbing t o the substratum [# L "2 t]
: particle concentration in the bulk water [# L"3]
: diffusivity [L2 t'1]
: half-thickness of channel [L]
: dimensionless distance [-]
: surface particle capture factor [-]
: gamma function [r(4/3) = 0.89338]
(28)
21
T he authors limited their analytical solution t o particles which are
characterized b y small Brownian diffusivities and large Peclet numbers.
However,
Equation 28 w a s used
(Escher 1986) t o describe cell transport
f r o m t h e bulk water t o the solid interface for motile cells with h i g h
diffusivity values. Bowen's proposed analysis will, therefore, b e u s e d in
this study t o calculate sticking efficiency and particle capture factor,
a n d tested fcy a ccmparison of motile and non motile cells of th e same
Pseudomonas strain.
The
surface particle capture
factor
(e)
in Equation
27 can b e
described as a first order rate coefficient with respect t o the bulk water
cell
concentration.
adsorption,
Low
values
w h e n the value
of
of
e indicate
a
low
e approaches infinity
probability
[ (1/e)
=
of
0 ] the
surface acts like a perfect sink and Equation 27 becomes:
X b' D c
N tM = ---- ----
(2/9X0 1/3
— ------------
h
N t"
(27)
r(4/3)
: total flux of particles t o the substratum [# L "2 t"1]
Equation 27 describes the total flux of particles t o t he substratum.
W h e n measuring the rate of adsorption, the surface particle capture factor
(e) can b e calculated from Equation 26 knowing the dimensionless distance
K1
- as:
K1
- = P e '1
Pe
x
h
8 x
----3 h
: Peclet number [-]
: Length from channel inlet [L]
: half thickness of channel [L]
(28)
22
T h e Peclet number is defined as follows:
Pe =
4 Vm h
--------
(29)
Dc
vm
: m e a n bulk velocity [L t'1]
Adsorption Velocity
T h e adsorption velocity can b e calculated b y normalizing the flux
of particles/cells t o the substratum t o the particle/cell concentration
in t h e b u l k water.
K a"
-----
Ka =
(30)
X b'
K a : adsorption velocity [Lt"1]
Adsorption
velocity
is
independent
of
bulk wat e r
particle/cell
concentration b u t dependent on geometry and hydrodynamics of the flow
system a n d diffusivity and adsorption ability of the particles/cells. K a
c a n b e calculated using the viable cell counts (Kav) as w e l l as using t he
total cell counts (Kat) . The adsorption velocity of irreversibly adsorbed
cells c a n b e expressed by:
K a irr = f irr Ka
(31)
K a irr: adsorption velocity of irreversibly adsorbed cells [L t"1]
Adsorption velocity could b e used t o compare bacterial adsorption
for various
substrata
or
organisms
geometry in t h e flow system.
for
constant
flow
conditions
and
23
Sticking Efficiency
Sticking efficiency is defined (Oiaracklis and Marshal, 1990) b y the
ratio
of
cells
adsorbing
to
the
substratum
to
t he
number
of
neiie
colliding w i t h the substratum. Sticking efficiency can b e u s e d as a system
independent
parameter
to
compare
bacterial
adsorption
at
various
interfaces o r environments. The sticking efficiency c an b e expressed as
follows:
Ka"
Sticking efficiency
------
(32)
Nt "
K a'1 a n d N t'1 are described in Equation 26 and 27, respectively. Sticking
efficiency could b e expressed in the ratio of these t wo Equations:
r(3/4)
$ = -------------------r(3/4) + (1/e) (2/9Ki)1/3
0
(33)
: sticking efficiency [-]
Sticking efficiency is a function of the particle capture factor a nd t he
Peclet number but independent of cell concentration. T he values for $ will
v a r y between 0 a n d I. High values for $ indicate a h i g h tendency for the
cells t o adsorb t o a substratum, while low values of $ indicate a low
tendency of cells t o adsorb t o a substratum. It should b e emphasized that
a h i g h value for sticking efficiency dos not necessarily correspond t o a
h i g h adsorption velocity, because sticking efficiency is independent of
the transport properties of the cells.
24
Fluid Dynamics
For fully developed laminar flow in a Newtonian fluid, shear stress
c a n b e calculated b y Newton's law of viscosity. W i t h a direction of flow
in y-direction and a n even flow distribution in x-direction,
the shear
stress is given by:
Tzy = - Pw ---
(34)
dz
Tzy
: shear stress [N L'2]
: fluid viscosity [N t L'2]
: bulk liquid velocity [L t'1]
: length coordinate in axial direction [L]
: height coordinate normal to the flow [L]
Pw
Vy
y
z
A t t h e center of the flow channel z = 0;
V y = V max .= 3/2 v m and
d v y/dz = 0; a t the walls z = +h and v y = 0. In a fully developed laminar
flow between two parallel plates the velocity profile h as a parabolic
shape (Bowen et al. 1976):
V y (Z) =
3/2
Vm
(I -
Z 2IV2)
(35)
shear stress at the wall:
rWall —
rwaii
h
vm
~ 3/2
fiy
Vm h ^
(36)
: shear stress a t the wall [N L"2]
: half-thickness of Channel [L]
: m e a n bulk velocity [L t'1]
W h e n t h e m e a n bulk water velocity is much
larger than cellular
motility, t h e hydraulic detention time of cells above t h e substratum can
I I I
25
b e determined as a function of their location:
21g
HRT(Z) -----------------3vm [I -(Z2M"2)]
Is
HRT
:
:
The
(37)
length, of substratum [L]
hydraulic detention time of cells above the substratum [t]
cellular detention time is a function of b u l k water velocity
a n d t h e cell location at the entrance perpendicular t o t he substratum.
Reynolds Number
T h e Reynolds number is commonly used to categorize flow conditions
in closed conduits as laminar or turbulent.
If t he Reynolds number is
smaller t h a n 2100 the flow is called laminar. For values greater than 2100
t h e flow is said to be turbulent. Reynolds number depends mainly o n th e
m e a n b u l k wa t e r velocity and the geometry of t he
flow system an d
defined as follows:
Re =
2 -Rh V m /)M
— --------
(38)
Pw
Rh
=
A
-----Z
Re
vm
Pw
Pw
R tl
A
Z
: Reynolds number [-]
: m e a n bulk water velocity [L t'1]
: density of water [M IT3]
: fluid viscosity [N t I/5]
:hydraulic radius [L]
:cross sectional area [L2]
: wetted perimeter [L]
(39)
is
26
METHODS
Experimental System
The
experimental
(Figure 2).
system was
used
as described b y
Bacterial cells were grown in a chemostat
Escher
(1986)
(330 ml volume)
until steady state conditions were reached. A phosphate buffered glucose
solution w a s used as nutrient
carbon limited
(Table I ) . Growth in all experiments was
(40 g m"3 as TOC)
and buffered at p H 6.8.
relatively high cell concentration in the chemostat,
Because of a
the effluent w as
diluted w i t h dilution water before entering the flow cell. The dilution
water h a d the same chemical composition as the nutrient except that no
organic carbon was present.
M A G B ANALYZER
NUTRIENT
DILUTION
MIXING CHAMBER
EPI LIGHT
MICROSCOPE
FLOW IN
NOMARSKI LENS
IEEEIiEiiisa
METAL COUPON
TocTum
FLOW OUT
Figure 2.
Experimental system
27
Table I.
Chemical composition of nutrient.
Chemical
Concentration in g m"3
Glucose
NH4C1
Na2HP04
KH2P04
CaGOS
100
36
568
544
90
Micronutrients were added according t o Trulear (1983).
T h e diluted chemostat effluent was pimped b y a non-pulsing gear pump
at a constant flow rate through a flow cell
(Table 2). A l l experiments
w e r e conducted in laminar flow at a shear stress of
corresponds t o a
flow rate of
0.75 N
Hf2
whi c h
1.2*10'4 m 3 hr'1 and a m e a n bulk water
velocity of 2.75*l0"2 m s ' 1. For a shear stress of 0.75 N m'2 a t t he solidwa t e r interface, the Reynolds number in the flow chamber w a s 1.4.
Table 2.
System specific parameters
experiments.
for the
length of flow cell
length of metal coupon
w i d t h of flow cell
w i d t h of metal coupon
height of flow channel
cross sectional area
b u l k liquid flow rate
b u l k liquid m e a n velocity
corresponding shear stress
m e a n cellular detention time
40.00
26.50
12.10
6.70
0.10
1.21
0.12
0.0275
0.75
0.96
Coupons
inserted
at
the
bottom
of
the
flow cell used
in t he
mm
mm
mm
mm
mm
mm2
I hr"
m s'1
N m"2
S
flow
cell
were
used
as
substrata. T h e reflective coupon surfaces (copper, silicon, 316 stainless
steel)
w e r e monitored continuously using reflective light from a n EPI
light microscope equipped with a Nomarski lens. Magnification w as between
28
1000 a n d 1500.
surface
was
For lower magnifications, the light reflection from the
too
colonization.
disperse
Transmitted
to
obtain
light
was
quantitative
used
for
information
monitoring
of
cell
transparent
surfaces, such as glass or polycarbonate. The microscope output was linked
t o a vi d e o camera.
The video signal was transferred t o a n image analysis
system (IAS) which converted the grey image from the video camera into a
binary image. The IAS counted the actual number of cells o n the surface
w i t h respect t o size and location.
The cell measurement data could b e
stored t o disk during experiments and retrieved for analysis at a later
time.
T h e u s e of the IAS to measure bacterial surface colonization is
described in m o r e detail b y Escher (1986).
The
metal
coupons
were
mechanically
polished
so
that
light
reflection enabled the observation of individual bacteria cells. "Buehler
Alp h a Micropolish II deagglcmerated Alumina 1.0 and 0.3" were used in
distilled water
for polishing.
This polisher has
a
hexagonal
crystal
structure and a particle size of 1.0 or 0.3 /m, respectively.
Pseudomonas aeruginosa and Pseudomonas fluorescence w e r e the test
organisms. Both Pseudomonas strains are strict aerobic, rod-shaped, polar
flagellated, gram-negative cells, 1.0 t o 2.5 yum Iorg a n d 0.4 t o 0.6
/m
in
diameter. T h e same Pseudomonas fluorescence strain was u s e d in a modified
m o t - version (provided from D. E. Caldwell, Vancouver, British Columbia).
T h e m o t - m utant w e r e found to lack a flagellum b y transmission electron
microscopy
analysis
difference
between
(Korber
the
et.al.
growth
rate
1989).
of
There
was
Pseudomonas
no
significant
aeruginosa
and
Pseudomonas fluorescens of the mot+ parent and the mot- mutant in batch
cultures (/z = 0.45 hr'1),
(Korber et. al. 1989).
29
Method of Analysis
Total Cell Count
T h e water sairples (bulk water or dhemostat effluent) were fixed in
2 % Formalin solution and stored at 4 0C. A ll containers w e r e autoclaved
before u s e for sampling. For staining, the prepared sample w a s homogenized
a n d suspended in 2 m l 0.01% acridine orange for 30 min. T h e stained sample
w a s filtered through a 0.2
black filter and the cells w e r e counted b y
image analysis using fluorescence light at 1000 x magnification. 10 fields
o n each filter w e r e averaged t o determine the cell concentration in th e
sample.
Replicates
of several filters with the same sample yielded a
standard deviation of less than 10%.
Viable Cell Counts
A dilution series was made from the water sample.
0.1 m l from the
dilutions was equally distributed o n three replicate plates w i t h R2A-agar
as growth media. Replicates of several plates of the same sample showed
a standard deviation between 7 and 15 %.
Cell Surface Hvdronhobicitv
This method
hydrocarbons"
is based
(BATH)
on the
degree of
"bacterial
adherence
to
and has found application in t he study of a w i d e
variety of bacteria and bacterial mixtures
(Rosenberg 1984). The method
u s e d (BATH-test) w a s developed b y Rosenberg and co-workers (Rosenberg et.
al.
1980)
hydrocarbon
and
was
adapted
for
use
(n-octane or n-hexadecane)
with
image
analysis.
Two
ml
of
were added t o round bottom test
30
tubes containing 5 m l phosphate-buffered sample. P-xylene w as originally
suggested for use
in the test,
but was abandoned because it produced
unreliable results. Others also reported problems w i t h th e u s e of p-xylene
for
measuring
Following
10
cell
surface
minutes
of
hydrophobicity
preincubation
at
(Vanhaecke,
30
0C,
the
Pijck,
1988).
mixtures
were
homogenized for 2 minutes. The hydrocarbon phase w as allowed t o separate
for 15 minutes, and the aqueous phase w as removed w i t h a pasteur pipette.
Total cell concentration was performed for the initial bacteria mixture
a n d t h e aqueous phase after hydrocarbon extraction.
T he decrease in cell
concentration of the aqueous phase w as used as a measure of cell surface
hydrophobicity.
Cell surface hydrophobicity is reported as t he fraction
of cells removed b y hexadecane or octane,
respectively,
a nd is called
hexadecane or octane partitioning value, respectively.
Total Organic Carbon CTOCl
T O C w a s measured using a Dohrman model D C 80 total organic carbon
analyzer. The p H of the water sample was lowered t o p h 2 w i t h phosphoric
acid, a n d aerated t o strip all carbon dioxide from the sample. A ll samples
w e r e homogenized and analyzed b y direct injection.
Dissolved Organic Carbon (DOC)
T h e measurements for filtered water samples were v e r y inconsistent.
U p t o 10 p p m carbon w a s leached from the filter. The carbon contamination
also
varied
considerably
between
filters.
Especially
for
low
carbon
concentrations, it was difficult to obtain consistent results. Th e u se of
PTFE membranes w i t h a pore size of 0.2
/m.
finally gave satisfying results
31
after rinsing for several
(about 8 to 10) times w i t h diluted phosphoric
acid.
Surface Roughness
Surface
roughness
was
obtained
from
reflected
light
distribution b y using a Raleigh perturbation relationship.
power spectral density function
(PSD)
as
a
function
of
the
A reflector
can b e calculated from the light
scatter a t various reflection angles.
behaves
scatter
surface
In general,
topology.
t he
calculated PSD
Integrating
the
P SD
function gives the root mean square (RMS) surface roughness value (Stover,
Ser a t i , Gillespie 1984).
Estimation of Cell Motility
T o estimate cellular motility,
the substratum w a s exposed to the
cell containing bulk water at a shear stress of 0.75 N m"2 until a cell
surface density of approximately4 000 CFU mm'2 was reached. After initial
adsorption occurred, the flow over the substratum w as terminated. Within
5 t o 10 minutes, m o s t of the initially adsorbed cells became motile o n t he
surface.
T h e surface was monitored continuously a t a magnification of
1500x an d recorded b y a high resolution V C R on video tape. Th e T V screen
w a s calibrated b y recording a calibration slide at the same magnification
in horizontal and vertical direction. After replaying t h e tape several
times, t h e random free runs could be measured b y marking traveling phase
(free run)
and twiddling phase
(change
in traveling direction). W h e n
replaying t h e tape in slow motion with time control, the cellular swimming
speed for a free ru n could be estimated (Korber et a l . , 1989).
32
T h e same procedure was used to test mot- cells for motility after
cellular colonization.
Determination of Rate of Adsorption
a n d Desorption o n the Substratum
T h e fate of adsorption and desorption was measured b y exposing the
solid interface t o a bulk water flow of constant cell concentration and
shear stress
(0.75 N m"2) . Individual cells were, counted an d memorized
w i t h respect t o their location and size on the substratum over time. W h e n
plotting t h e number of cells adsorbing or desorbing versus tine linear
regression analysis could give the values for adsorption velocity and the
probability of desorption (Escher, 1986).
Determination of Rate of Growth
and Erosion on the Substratum
T h e substrata w e r e exposed to a bulk water flow (shear stress 0.75
N m"2) of approximately 3*10* cells ml'1 until the cell surface density
reached a value between 2000 and 3000 CFU mm"2. These cells were adsorbed
at
an
irreversible
state.
After
this
initial
colonization
period,
approximately 30 minutes for copper to 2 hours for silicon, t he bulk water
flow
was
nutrient
switched
from
solution
of
the
diluted
identical
dhemostat
chemical
effluent
composition
to
as
a
fed
sterile
to
the
dhemostat.
The rate of growth on surfaces was determined b y following
individual
cells
replicating
over
time.
Simultaneously, t he
rate
of
erosion could b e determined fcy following individual cells desorbing from
t h e surface. After plotting the number of cells replicated versus time,
the
growth
rate
and
doubling time
could be
determined b y
regression
33
analysis.
Image Analysis
The
Cambridge Q l O , equipped with a n
image analyzer unit
(68000
processor) w a s used in the experimental system. In addition t o the display
a n d output of the measurements,
the image analyzer c an perform limited
statistical analysis. U p t o eight different distribution histograms can
b e calculated at the same time. The instrument has two different routines,
one in compiled C and one in M-Basic. Images were manipulated b y a digipad
connected t o t h e image analyzer.
34
RESULTS
Results were obtained for five different materials a n d two different
bacteria
species
- Pseudomonas aeruginosa and Pseudomnnag
fluoresens.
Pseudomonas fluoresens was used as a mot+ strain as well as a mot- mutant.
Substrata
included
copper,
316
stainless
steel,
silicon,
polycarbonate. The surface roughness varied with various
25
A
(2.7*10-3 ^m) for silicon t o 150
A
glass
mm+erW
a nd
from
(15.6 IO"3 /rni) for Copper as the
roughest surface used in the experiments. 316 stainless steel was used t o
p e r form a long duration starvation study (as long as 1224 h o u r s ) .
Chemostat Operation
T h e chemostat was operated at a dilution rate of 0.2 hr"1 ,and t he
chemostat effluent concentration varied between 3.8 t o 4.5 IO7 cells ml"1
(measured as total count). The effluent dissolved organic carbon varied
between 3.2 and 5.0 g m"3. Before entering the flow chamber, th e chemostat
effluent w a s diluted t o a factor of approximately 10 w i t h carbon free and
cell free dilution water. Hence, the dissolved organic carbon in the bulk
w a t e r varied between 0.1 and 0.6 g m'3 and cell concentration in the bulk
wat e r w a s kept between IO6 and IO7 cells ml"1. Temperature wa s controlled
in t h e chemostat at 30 0C and the temperature at the solid—water interface
w a s estimated to b e constant at 22 ± 2°C. Shear stress w a s kept constant
during all experiments at 0.75 N m"2 in the flow chamber.
35
Image Processing
Images w e r e taken at 10 minute intervals and manipulated b y using
t h e digipad drawing device. Regression analysis was applied t o determine
t h e r a t e coefficients for adsorption, desorption, growth a nd erosion.
Cell Surface Hvdroohobicity
Cell surface hydrophobicity was measured as "bacterial adherence t o
hydrocarbons"
cells,
the
(BAIH-test,
hexadecane
Rosenberg,
partitioning
et al,
1980).
values
F or all the tested
were
higher
than
t he
partitioning values for octane (Table 3). Mean hydrocarbon partitioning
(cell surface hydrophobicity)
values were calculated b y averaging the
hexadecane a n d octane partitioning values for each species an d strain. Th e
lowest
carbonhydrate
fluorescens
partitioning
(mot+). The
values
Pseudomonas
were
found
fluorescens
for
(mot-)
Pseudomonas
strain
had
a
significantly higher value for cell surface hydrophobicity than the motile
strain. Pseudomonas aeruginosa was found to be highly hydrophobic.
Table 3.
Cell surface hydrophobicity.
Strain
octane part,
value (-)
hexadecane part,
value (-)
m e a n H G part,
value (-)
Ps aeruginosa
0.95+0.01
0.99+0.01
0.97+0.02
Ps fluorescens (mot+)
0.54+0.11
0.80+0.04
0.67±0.13
Ps fluorescens (mot-)
0.91+0.04
0.93+0.03
0.92+0.04
36
Pseudomonas aeruginosa. Kinetic Results
Surface Colonization
of Pseudomonas aeruginosa
Pseudomonas
aeruginosa
cells
were
transported
to
t he
interface
either a s single cells or in a divided cell state (divided cells but not
separated). Colonization occurred in the same manner. Clustering was not
observed o n t h e substratum and desorption occurred as single or divided
cells from t h e substratum to the bulk water.
Copper
Copper corrosion results in a copper oxide layer within hours of
exposure t o air. The copper oxide layer has a protective function for the
metal underneath. T h e presence of an oxide layer can b e easily detected
b y visual observation. A
reflective,
light.
A
freshly polished copper surface is shiny and
whereas a n oxidized copper surface looks dull and absorbs
fresh
polished
m e asurements. Copper
surface
is a
was
soft metal
important
for
reflective
and polished v e r y
light
fast but wa s
difficult t o keep clean. The orientation of polishing w as determined from
a light scatter surface roughness measurement (Figure 3). T h e RMS surface
roughness of a freshly polished copper coupon, when measured in vertical
direction
(direction
of
flow), was
220 A,
whereas
t he
RMS
surface
roughness in horizontal direction (direction perpendicular t o the flow)
w a s 82
A.
where
mai n l y
Therefore, the orientation of dislocations
in
the
vertical
direction.
For
(polishing tracks)
comparison
reflective surfaces a n average value wa s calculated (150
with
A) . After
other
the
37
1.00E+09
PSD
2.00E+03
A
LOG
I.00E+07
Copper (oxydized): 1680
.50E+03
1.00E+05
-- 1.00E+03
1.00E+03
5.00E+02
/
L-—
Copper (vertical): 220 A _
—
'
I
I
Illl IA'-l— l— t—
— |— t-Copper (horizontal): 82 A 3""0v0BE+00
0
.3
.6
.9
1.2
1.5
0.6320pm Oi = 5.0° R =0.06 X = -13.9mm Y = -8.0mm 2Ws = 3.0mm
Figure 3.
Results from a laser light scatter measurement on the copper
surface. Power spectral density (PSD) is plotted versus frequency. The
integration of this curves provide values for RMS surface roughness. A
fresh polished copper surface, measured in vertcal and horizontal
direction is compared with a highly oxidized surface.
polished copper coupon was exposed to air for several weeks the RMS
surface roughness increased to 1680
A.
Cell accumulation on the copper substratum w as expressed as cell
surface density ( X " ) and observed for a bulk water cell concentration of
1.50*106 cells ml'1 (Figure 4 a ) . The steep slope indicates a high rate of
adsorption
(Ka = 2.23 m m hr'1). The rate of desorption w as rather low
(probability of desorption b = 0.10). There was no replication of cells
observed when using copper as substratum (Figure 4). Even for an extended
period
of
time
(24
hours)
no
replication
exposure time t o a cell suspension was
cells to accumulate,
was
seen.
However,
if
the
long enough for a monolayer of
replication of cells in the second and higher cell
layers was observed. Unfortunately, an accurate measurement of growth rate
38
COPPER, Ps aeruginosa, 1.40 E+OG CELLS/ML
20000
ACCUMULATION
b = 20.89 + /- 0.75
15000-
DESORPTION
-*r-
Ka = 2.00 + /- 0.21 mm/hr
ADSORPTION
GROWTH
z
10000-
5000-
0tltm ujm wfr6i6iLi!ir,iacictanjr]CJOL ifiiEtati) trirrti
OO
150
200
TIME IN MINUTES
COPPER, Ps aeruginosa
10000
ACCUMULATION
EROSION
8000-
ADSORPTION
GROWTH
6000-
4000-
100
150
TIME IN MINUTES
Figure 4.
Cell accumulation of Pseudomonas aeruginosa on copper using
a bulk water concentration of 1.4*10* cells ml"1 (Figure 4 a) and a sterile
nutrient (40 ppm C) after initial cell colonization of 2000 CFU mm"2
(Figure 4 b). Shear stress 0.75 N m"2.
39
in up p e r cell layers w a s technically not possible w i t h t h e image analysis
system. T h e kinetic results for Pseudomonas aeruginosa a re summarized in
Table 4.
Silicon
Silicon is used in industry as a semiconductor material mostly in
its monocrystalline form.
procedures,
To allow very accurate photographic etching
the silicon chips are thinly sliced and polished with high
precision t o very
low surface roughness values
and
a
h i g h degree
of
flatness.
T h e monocrystalline silicon substratum h ad a PMS surface roughness
o f 25
was
A
without regard to its orientation. The bulk wa t e r concentration
constant during the accumulation experiment a t
3.5
IO6 cells ml'1
(Figure 5 a ) . T h e adsorption velocity w as low (Ka = 0.47 m m hr'1) and th e
probability of desorption was high (b = 0.86). Most cells d id not remain
at their location for an extended period of time after initial adsorption.
T h e rate of erosion (Ke = 0.32 hr'1) w as slightly higher than the growth
rate
(Kg = 0.28 h r 1) (Figure 5 and Table 4), hence t he accumulation of
cells o n the silicon surface was slow. Even cells that w e r e adsorbed for
a long p eriod of time (over 200 minutes) desorbed w h e n additional shear
w a s applied. Replication of adsorbed cells often resulted in desorption
of both, mother and daughter cells .
40
SILICON, Ps aeruginosa, 3.50 E+06 CELLS/ML
20000
ACCUMULATION
b = 85.80 + /- 6.39
DESORPTION
Ka = 0.49 + /- 0.04 mm/hr
ADSORPTION
15000-
GROWTH
10000-
=>
5000-
DO
150
200
TIME IN MINUTES
SILICON, Ps aeruginosa
10000
Ke = 0.32 +/-0.01 1/hr
8000-
Kg = 0.28 + /- 0.02 1/hr
ACCUMULATION
EROSION
ADSORPTION
GROWTH
6000-
4000-
(Z)
2000-
)K
i
*
100
150
TIME IN MINUTES
Figure 5.
Cell accumulation of Pseudomonas aeruginosa on silicon using
a bulk water concentration of 3.5*10* cells ml'1 (Figure 5 a) and a sterile
nutrient (40 ppm C) after initial cell colonization of 2100 CFU mm'2
(Figure 5 b). Shear stress 0.75 N m"2.
41
316-Stainless Steel
A
freshly polished 316 stainless steel surface h ad a RMS surface
roughness of 77
A when
measured in vertical direction. After this surface
was exposed to air for several weeks, the RMS surface roughness increased
to 120
A
(Figure 6).
1.00E+10
PSD AzWii
LOG
1.00E+08
1.50E+02
.
— L20E+B2
316 stainless steel (oxydizcd): 120 A
1.00E+06
9.00E+01
316 sta nless steel (vertical): 77 A
1.00E+04
1.00E+02
3.0BE+01
1.00E+00
I I I I I 0.0BE+0B
1.2
1.5
-8.0mm 2Ws = 3.0mm
1/Wi
0
0.6328wi
Oi = 5.0° Ii =0.62
X = -13.9mm
Y
Figure 6.
Results from a laser light scatter measurement on the 316
stainless steel surface. Power spectral density (PSD) is plotted versus
frequency. The integration of this curves provide values for RMS surface
r o u g h n e s s . ^A fresh polished surface, measured in vertcal direction is
ccmP ared w i t h a highly oxidized surface, measured in vertical direction.
The bulk water concentration was constant during the accumulation
experiment at 1.10 IO6 cells ml'1 (Figure 7 a ) . The adsorption velocity was
K a = 1.82 m m h r 1, the probability of desorption was b =
0.30,
and the
growth rate on the substratum (Kg = 0.33 hr"1) was higher than the rate of
erosion ( K e = 0.09 hr'1) (Figure 7 and Table 4).
42
31C STrAlNIJiSS STEIiL, Ps aeruginosa, 1.10 E+06 CEIJS/ML
20000
ACCUMULATION
b = 29.68 + /- 0.32
DESORPTION
Ka = 1.82 + /- 0.22 mm/hr
ADSORPTION
GROWTH
Z
10000-
5000-
B a c i i iirjL i
DO
150
20
TIME IN MINUTES
310 STAINIJiSS STEEIt Ps aeruginosa
10000
9000-
Ke = 0.092 +/-0 .0 1 3 1/hr
ACCUMULATION
EROSION
8000Kg = 0.33 + /- 0.023 1/hr
U
ADSORPTION
7000-
-O
GROWTH
6000500040003000-
O
1000-
0 4^'k
T T T
Jc T ;k ;|; )!(= #
100
150
TIME IN MINUTES
Figure 7.
Cell accumulation of Pseudomonas aeruginosa on 316 stainless
steel using a bulk water concentration of 1.1*106 cells ml'1 (Figure 7 a)
and a sterile nutrient (40 ppm C) after initial cell colonization of
2000 CFU mm
(Figure 7 b). Shear stress 0.75 N m"2.
Y
43
Table 4.
Pseudomonas aeruginosa. Summary of the Results
(shear stress 0.75 Nm'1)
rate constant units
copper
silicon
316 stainless
glass*
RMS
[A]
Ka
[mm hr"1] 2.22+0.43
b
[-]
Ke
[hr'1]
e
["]
0.032+0.006
0.007+0.001
0.026+0.003
0.016
$
[-]
0.026+0.005
0.006+0.0004
0.022+0.002
0.018
9
[hr]
kg
.
150
27
0.47+0.04
0.21+0.02
0.86+0.06
0.32±0.01
-
[hr'1]
-
2.47±0.71
0.00
0.28+0.08
121
-
1.84+0.22
0.30+0.02
0.09+0.01
2.10+0.38
0.33+0.06
1.14
0.62+0.18
0.32+0.08
1.36+0.29
0.5 1 ± 0 .11
Results from Escher (1986)
Surface Roughness
R M S surface roughness was measured for all metal surfaces used in
t h e experiments and w a s correlated t o adsorption velocity
(Figure 8) .
Adsorption velocity increased with increasing roughness of t he substratum
and
the
probability
of
desorption
decreased
with
increasing
surface
roughness. Hence, the irreversible adsorption was influenced even stronger
b y t h e roughness of the substratum.
Pseudomonas fluorescens (mot+). Results
Surface Colonization of Pseudomonas
fluorescens fmot+'l
Cellular
colonization
of
Pseudomonas
fluorescens
(mot+)
different from the behavior of Pseudomonas aeruginosa. Pseudomonas
fluorescens (mot+) cells were transported as single cells t o the
w as
44
REVERSIBLE ADSORPTION
IRREVERSIBLE ADSORPTION
SILICON
<
316 STAINLESS
COPPER
0.5-/
'
77
I
RMS SURFACE ROUGHNESS (A)
Figure 8. Adsorption velocity of Pseudomonas aeruginosa versus surface
roughness of the substratum. Shear stress 0.75 N m %
fluorescens (mot+) cells were transported as single cells t o the
substratum but adsorption occurred more often in clusters. Desorption of
cells even from clusters, occurred as single cells in m o s t cases.
Growth
rate
and
the
rate
of
erosion
were
not
determined
for
Pseudomonas fluorescens.
316 Stainless Steel
Cell accumulation of Pseudomonas f luorescens (mot+) o n 316 Stainless
w a s observed at a cell concentration of 1.3*106 cells ml'1 (Figure 9).
Pseudomonas fluorescens
(mot+)
had an adsorption velocity on Stainless
Steel of K a = 0.30 m m hr'1 and a probability of desorption b = 0.61. The
kinetic results for adsorption and desorption of PseudnmnnmR fluorescens
(mot+) on stainless steel and other substrata are summarized in Table 5.
45
Polycarbonate
Cell accumulation of Pseudomonas fluorescens (mot+) on polycarbonate
was observed at a bulk water cell concentration of 3.85 IO6 cells ml'1
(Figure 10 and Table 5). The
adsorption velocity o n
polycarbonate was
K a = 0.23 m m hr'1. The probability of desorption (b = 0.03) w as relatively
low on polycarbonate.
Glass
Cell accumulation of Pseudomonas fluorescens
(mot+)
on glass was
observed at a bulk water cell concentration of 3.85 IO6 cells ml'1 (Figure
11 and Table 5). The adsorption velocity on glass was K a = 0.25 m m hr"1,
the probability of desorption was b = 0.41.
316 STAINLESS STEEL Ps fluorescens (mol+),
1.30 E+6 CELLS/ML
IGOOO1
ACCUMULATION
E 14000E
3 12000LU
O
10000 55
z
LU
DESORPTION
b = 61.00 + /-8 .1 4
ADSORPTION
-EH
Ka = 0.302 + /- 0.036 mm/hr
GROWTH
8000-
Q
LU
O
2
DC
3
CO
LU
O
60004000-
2000
-
t LHm t B u H i im L m iiL m i n ij u j i j j t l H l i i lJ L ljd iL l H lii lJ L iiL ljL i rl I r j U
100
150
200
TIME IN MINUTES
250
'
300
Figure 9.
Cell accumulation of Pseudomonas fluorescens (mot+) on 316
stainless steel using a bulk water concentration of 1.3*106 cells ml'1.
Shear stress 0.75 N m'2.
46
POLYCARBONATE, Ps fiuorcsccns (mot+),
3.05E+G CELLS'/ML
16000
ACCUMULATION
E
E 14000-
DESORPTION
I
LL
b = 20.07 -+/- 3.65
12000 -
ADSORPTION
’□
GROWTH
O
Z
Ka = 0.232 -+/- 0.017 mm/hr
10000
Tn
z
m
8000-
Q
LU
I
CC
6000>ieieE
=>
CO
4000-
UJ
2000
O
-
L
WLummiiimtHuiu itiii urn m ull mumuiu ii I
ifL16Jji 'l l Iii
50
100
150
200
TIME IN MINUTES
250
300
Figure 10.
Cell accumulation of Pseudomonas f luorescens
(mot+)
on
polycarbonate
using a bulk water concentration of 3.85*106 cells ml'1
Shear stress 0.75 N m'2.
GLASS, Ps fluoresceus (mol+),
3.05 E+OG CELLS/ML
16000
ACCUMULATION
14000-
DESORPTION
?
12000-
b = 40.85 + /- 5.59
10000-
Ka= 0.246 + /- 0.027 mm/hr
Z
8000-
U
6000-
3
4000-
ADSORPTION
GROWTH
2000.
ITl Cl 111 O
0
20
40
60
CTI □
111 Cl W □
I
80 100 120 140 160 180 200
TIME IN MINUTES
Figure 11. Cell accumulation of Pseudomonas fluorescens (mot+) on glass
using a kulk water concentration of 3 .8 5 *1 0 * cells ml"1. Shear stress
0.75 N m .
47
us i n g a b u l k wa t e r
concentration of 3.85*106 cells ml'1. Shear
stress
0.75 N m"2.
Table 5.
Pseudomonas fluorescens (mot+), Summary of th e Kinetic Results
(shear stress 0.75 Mm'2) .
rate constant
Unit
Ka
316 Stainless
[mm hr"1]
Polycarbonate
Glass
0.30+0.03
0.23+0.02
0.25+0.01
b
[-]
0.61+0.20
0.20+0.22
0.41+0.25
€
[-]
0.0042+0.0005
0.0032+0.0004
0.0034+0.0004
0
[-]
0.0036+0.0004
0.0028+0.0002
0.0029+0.0003
Surface Roughness
Numbers
for RMS
surface roughness
for the
transparent
surfaces
cannot b e given, because the power spectral density theory t o calculate
R M S roughness for transparent surfaces does not exist. However, a laser
light
scatter
measurement
on
these
surfaces
permitted
a
qualitative
estimate indicating that the polycarbonate surface wa s slightly rougher
t h a n t h e glass.
Results wi t h Pseudomonas fluorescens Cmot-I
Surface Colonization of
Pseudomonas fluorescens (’m o t-1
Surface colonization of
Pseudomonas fluorescens (mot-) cells w e r e
similar t o the motile parent. Cells d id get transported as single cells
t o t h e substratum but adsorption occurred often in clusters. Desorption
of cells even from clusters occurred as single cells in m o s t cases.
48
This excluded the possibility that these cells switched back t o their mot+
origin.
316 Stainless Steel
Cell accumulation of Pseudomonas fluorescens (mot-) o n 316 stainless
w a s observed a t a bulk water cell concentration of
(Figure
12
fluorescens
and
Table
(mot-)
6).
The
adsorption
3 .50 * 10 * reiip ml'1
velocity
on 316 Stainless Steel w as K a =
of
Psendnrnnn^c;
0.18 m m hr"1, the
probability of desorption from the metal substratum w a s b = 0.84.
Polycarbonate
Cell accumulation of Pseudomonas fluorescens (mot-) o n polycarbonate
w a s observed a t a bulk water cell concentration of 5.15*10* cells ml'1
(Figure 13 and Table 6). The adsorption velocity
was K a = 0.15 m m hr"1 and
t h e probability of desorption was b = 0.03.
Glass
Cell accumulation of motile Pseudomonas flourescens o n glass w as
observed a t a bulk water cell concentration of
8.90*10* cells ml"1 (Figure
14 and Table 6). The adsorption velocity on glass was K a = 0.09 m m hr'1 and
t h e probability of desorption was b = 0.23.
49
3 IG STAlNIJiSS STElil4 Ps lluorcsccns (mol-), 3.50 E+OG CEIiS/ML
20000
ACCUMULATION
18000-
DESORPTION
16000
O
b = 84,28 + /- 11.38
ADSORPTION
14000Ka = 0.183 + /- 0.033 mm/hr
12000-
GROWTH
*)K)K)K)K)KCK)K)^
10000 800060004000-
U
2000ILUUJIUUHItinei I-ILlliH JClClL Ir,IU O LOLImr]
00
150
20I
TIME IN MINUTES
Figure 12.
Cell accumulation of Pseudomonas fluorescens (mot-) on 316
stainless steel using a bulk water concentration of 3.5*106 cells ml'1.
Shear stress 0.75 N m"2.
POLYCARBONATE, Ps Iluorcsecns (m ol-),
5.15 E+OG CELLS/ML
16000
ACCUMULATION
14000-
1200010000-
DESORPTION
b = 3.24 + /-1 .5 6
ADSORPTION
Ka= 0.152 + /- 0.022 mm/hr
GROWTH
800060003
4000-
2000
-
IUUUl
50
100
150
200
250
TIME IN MINUTES
300
350
Figure 13. Cell accumulation of Pseudomonas fluorescens (mot-) on
polycarbonate using a bulk water concentration of 5.15*106 cells ml'1.
Shear stress 0.75 N m'2.
50
GLASS, Ps fluorcsccns (mol-),
0.90 E+OG CELLS'/ML
ACCUMULATION
DESORPTION
-M -
ADSORPTION
-o-
GROWTH
Figure 14.
Cell accumulation of Pseudomonas fluorescens (mot-) on glass
using
a bulk
water concentration of 8.9*10* cells ml"1. Shear stress
0.75 N m"2.
Table 6
Pseudomonas fluorescens (mot-), Summary of the Kinetic Results
(shear stress 0.75 Mm'2) .
; constant
Ka
Unit
[mmhr"1]
316 Stainless
Polycarbonate
0.18+0.04
0.15+0.02
Glass
0.09+0.02
b
[-]
0.84+0.11
0.03+0.01
0.23+0.03
€
["]
14.65+0.47
7.84+0.18
3.14+0.94
$
["]
0.48+0.03
0.33±0.01
0.17+0.06
Surface Colonization and Dislocations on Stainless Steel
A
higher
adsorption
rate
at
cracks,
interruptions
or
other
dislocations on the surface was not observed for any bacteria species in
51
any
of
the
experiments.
However,
all
dislocations
found
mechanically polished surfaces were in the micro range (<o.l
/m
on
the
in width)
a n d d i d not change hydrodynamics on the surface. A typical colonization
pattern o n 316 stainless steel for low cell surface densities showed a
random cell distribution. Dislocations were not colonized at any higher
level than other places on the surface.
Pseudomonas aeruginosa - Starvation and
Adsorption to 316 Stainless Steel
The
chemostat
nutrient
supply .was
cut
off
after
steady
state
conditions w e r e reached to study cellular colonization of Pseudomonas
aeruginosa in oligotrophia environments. Thus, the former chemostat w as
changed
into
maintained.
a
batch
reactor.
Aeration
and
agitation
were
still
The cell size distribution and the cell concentration w e r e
monitored over time. Adsorption/desorption of starved cells were observed
a t various durations of starvation.
A t defined intervals,
the diluted
effluent w a s pumped over the 316 stainless steel surface a t a constant
shear stress of 0.75 N m'2. The cell concentration w as constant during one
adsorption experiment w i t h values between IO6 and
IO7 cells ml"1. T he
cellular accumulation curves are presented in the Appendix. Contamination
was
observed
after
900
hours
of
starvation
and
the
experiment
w as
terminated after 1224 hours. The viable plate counts after 816 hours of
starvation showed no contamination (Figure 15).
52
Figure 15. Photograph, plate count on R 2A medium of Pseudomonas aeruginosa
starved for 816 hours.
Dwarfing and Fragmentation
Immediate responses of Pseudomonas aeruginosa t o starvation were
size
reduction
(dwarfing)
and
an
increase
in
cell
concentration
(fragmentation). Total and viable cell concentrations increased after 24
hours followed by a period of constant cell concentration until 200 hours
of starvation (Figure 16). From 200 t o 300 hours duration of starvation,
the cell concentration decreased rapidly. After 300 hours of starvation
the rate of decrease slowed. The average size distribution of Pseudomonas
aeruginosa w a s determined as cell surface area as a function of duration
of
starvation
determined
(Figure
at 0.91
17).
The
/m2. After
average
size
of
healthy
24 hours of starvation,
cells
was
the average size
decreased to 0.58 pm2. For longer durations of starvation the average size
stayed relatively constant between 0.45 and 0.58 pm2.
53
CELL CONCENTRATION VERSUS DURATION OF STARVATION
>100
□
O
125
250
375
500
625
750
VIABLE CELL COUNTS
TOTAL CELL COUNTS
875
1 0OO 1 1 2 5 1 2 5 0
DURATION OF STARVATION IN HOURS
Figure 16. Total and viable suspended cell concentration versus duration
of starvation, (curve fit with Axum software).
AVERAGE SIZE DISTRIBUTION VERSUS DURATION OF STARVATION
1 25
250
375
500
625
750
875
1 0 0 0 1 1 2 5 1 2 50
DURATION OF STARVATION IN HOURS
Figure 17. Cell size distribution (mean and standard deviation) of
Pseudomonas aeruginosa versus duration of starvation, (curve fit with Axum
software).
54
Surface Colonization of Starved Psgnrinrnnnag
aeruginosa on 316 Stainless Steel
For
v e lo c ity
Pseudomonas
aeruginosa
on
316
Stainless
Steel
adsorption
w a s a function of duration of starvation (Figure 18). Adsorption
velocity w a s determined from linear regression analysis o f a 100 minutes
cellular accumulation experiment (Appendix, Figure 28-39). T he adsorption
velocity (Ka) w a s calculated using the viable cell counts (Kav) as well as
t h e total cell counts (Kat) . Because viable counts were consistently lower
t h a n total counts the adsorption velocity was higher w h e n using viable
counts. In either case, the adsorption velocity
of
starvation
up
to
500
hours
in
a
linear
decreased for durations
fashion.
Past
500
hours
starvation, adsorption velocity remained at a relatively constant level.
Val u e s
for K a t
and K a v were
0.10-0.35 m m hr'1 a nd
0.4-0.6 m m hr"1,
respectively. The values of K a t and K a v for healthy cells
(/z = 0.2 hr"1)
w e r e measured a t 1.25 and 1.56 m m hr'1, respectively.
T h e probability of desorption (b) from the stainless steel surface
decreased during the first 300 hours of starvation a nd remained a t its
lowest values through 800 hours (Figure 19). The probability of desorption
increased slightly w i t h longer durations of starvation a nd remained a t a
relatively constant value of b = 0.07.
In a ll starvation experiments, n o growth or cell division was seen
o n the substratum during the time period measured .
Cellular Motility
T h e K a values were chosen t o present the adsorption data because
cellular motility is not required for its calculation. Cellular TnnH H t y
55
VIABLE COUNTS
TOTAL COUNTS
1 25
2 5 0 3 7 5 5 0 0 6 2 5 7 5 0 8 7 5 1 0 0 0 1 1 25 1 2 5 0
DURATION OF STARVATION IN HOURS
Figure 18. Adsorption
steel at different
determined using the
cell concentration in
velocity of Pseudomonas aeruginosa on 316 stainless
stages of starvation. Adsorption velocity w as
total number of cells as well as using the viable
the bulk water, (curve fit with A x u m software).
0 .5 4
0 .3 0 -
1 25
250 375 500 625 750 875 1000 1125 1250
DURATION OF STARVATION IN HOURS
Figure 19. Probability of desorption for adsorbed Pseudomonas aeruginosa
at different stages of starvation at a 316 stainless steel substratum,
(curve fit wi t h A x u m software).
56
during starvation w a s qualitatively determined after 816 hours duration
of starvation. The average random free runs were observed a t
from
47
observations.
For
the
cellular
estimate c a n b e given with v r = 100 to 200
swimming
/m
s"1.
speed
33 + 11 ^ m
only
a
rough
57
SIMULATION OF CELL ACCUMULATION
Csllulair colonization on 316 stainless steel, copper, and silicon
could b e simulated for Pseudomonas aeruginosa from experimental results
(Equation 12).
From the results with Pseudomonas fluorescens. the t wo
species could b e used for comparison of early surface colonization. Th e
growth rate on the substratum was calculated for low nutrient conditions
(Equation 9), whereas the value for the maximum growth rate a t 22 0C w as
determined
experimentally
for Pseudomonas
aeruginosa. For
fluorescens the value of the maximum growth rate
determined
by
Korber
et.
al.
(1989).
Pseudomonme
o n glass w as u s e d
Certain parameters
were
varied
individually, and cellular accumulation was simulated.
Bulk Water Cell Concentration
The
stainless
steel
surface
coverage
for
a
bulk
water
cell
concentration of IO6 cells ml"1 after 700 minutes w a s predicted t o b e 10
% of t h e total surface area (Figure 20). For a cell concentration of 4*106
t h e cell covered surface area was simulated to be 33 %.
Fluid Shear Stress
Cell accumulation of Pseudomonas aeruginosa
is simulated o n
316
stainless steel at various fluid shear stresses on the substratum (Figure
21). W i t h a n exponential influence of shear stress t o adsorption (E s cher,
1986), cellular accumulation decreased with higher shear o n the surface.
58
Substrata and RnlTc
Water Nutrient Concentration
Cell accumulation of Pseudomonas aeruginosa is presented for three
reflective substrata (Figure 22). Cell concentration w a s assumed constant
a t 2*106 cell ml'1 in the bulk water and nutrient w as assumed t o b e no t
limiting. T h e adsorbed bacteria grow at their maximum rate found o n each
surface. T h e Silicon surface showed a total area coverage (TAC) of less
than
1% after 700 minutes,
Copper ha d a TAC value of
3.4
% and
316
stainless steel showed the highest TAC value of 16 % after 700 minutes.
A
similar simulation of cell accumulation of Pseudomonas aeruginosa on
316 stainless steel, copper and silicon was performed for a low nutrient
environment (Figure 23). The total coverage on the silicon surface than
w a s still approximately I % . The TAC o n copper also d id n o t change because
there w a s no growth o n the copper surface. The TAC o n th e stainless steel
surface as a result of the slower growth process decreased t o 5 %.
Mixed Population
Cell accumulation of a mixed population of Pseudomonas aeruginosa
a n d Pseudomonas fluorescens (mot+) and Pseudomonas fluorescens (mot-) is
presented in Figure 24.
Growth rate for both species w e r e identical and
there w e r e no toxic or other chemical interactions between th e two species
assumed. T h e simulation was proceeded until TAC w as 100 %. T he fraction
of total area covered b y Pseudomonas aeruginosa was 94 %, th e fraction of
area covered b y Pseudomonas fluorescens (mot+) was 5 % a nd the fraction
of area covered by Pseudomonas fluorescens (mot-) w as about I %.
59
E
E
l[
Z)
LL
O
£
55
z
HI
Q
111
%
DC
3
(ZD
LU
O
TIME IN MINUTES
Figure 20.
Simulation of cell accumulation of Pseudomonas aeruginosa on
316 stainless steel at various bulk water cell concentrations. Shear
stress 0.75 N m"2.
600000
0.5 N/SQM
^
500000-
S
400000-
Z
300000-
1.0 N/SQM
1.5 N/SQM
200000 -
to 100000-
300
400
TIME IN MINUTES
Figure 21.
Simulation of cell accumulation of Pseudomonas aeruginosa on
316 stainless steel at various shear stress. Bulk water cell concentration
6*106 cells ml"1.
60
160000
140000-
O 120000
100000-
316 STAINLESS STEEL
COPPER
8000060000-
SILICON
4000020000 -
300
400
TIME IN MINUTES
Figure 22.
Simulation of cell accumulation of Pseudomonas aeruginosa on
316 stainless steel, copper,and silicon at rich nutrient conditions in the
bulk w ater (S >10 p p m C ) . Bulk water cell concentration 2*10* cells ml'1
and shear stress 0.75 N m"2.
50000
45000-
P
316 STAINLESS STEEL
4000035000-
COPPER
30000z
25000-
SILICON
200001500010000 5000300
400
TIME IN MINUTES
Figure 23.
Simulation of cell accumulation of Pseudomonas aeruginosa on
316 stainless steel, copper, and silicon at low nutrient conditions in the
bulk water (S =0.6 p p m C ) . Bulk water cell concentration 2*10* cells ml'1
and shear stress 0.75 N m"2.
61
900000
800000=) 700000Pseudomonas aeruginosa
z
600000500000-
Q
400000-
<
300000-
3
200000 -
Pseudomonas fluorescens mot+
Pseudomonas fluorescens mot -
100000-
TIME IN MINUTES
Figure 24.
Pseudomonas
fluorescens
4*106 cells
Simulation of cell accumulation of a mixed population of
aeruginosa. Pseudomonas fluorescens (mot+), and P s e u d o m o n a s
(mot-) on 316 stainless steel. Cell concentration
ml'1 and shear stress 0.75 N m'2.
I
\ fl
.62
DISCUSSION
Comparison of Pseudcanonas aeruginosa on Various Substrata
In all experiments, adsorption rate and desorption rate were found
t o b e independent of the accumulated cell surface density a nd were treated
as first order w i t h respect to bulk water cell concentration. The rate of
erosion and replication was found to be independent of t h e bulk water cell
concentration and first order with respect to the accumulated cell surface
density. This finding is consistent with the results obtained b y Escher
(1986).
Reversible
and
irreversible
adsorption
velocity
of
Rwndnmnnac;
aeruginosa w e r e dependent on the substratum (Figure 25). Th e lowest rates
of reversible and irreversible adsorption
were found o n th e very smooth
monocrystalline silicon surface. The irreversible adsorption velocity was
calculated a t 0.07 m m hr'1. Only 14 % of all reversibly adsorbed cells
became irreversibly adsorbed. The silicon surface also h a d the lowest RMS
surface roughness value of 25
roughness value of 77
A
A.
316 stainless steel h a d a n RMS surface
and an irreversible adsorption velocity of 1.33
m m hr'1 whi c h w a s 20 times higher than o n silicon. 70 % of all reversibly
adsorbed cells became irreversibly adsorbed over time o n t he stainless
steel
surface.
average R M S
Copper was
the
roughness value
of
roughest
150
A
surface
and
an
investigated with
irreversible
an
adsorption
velocity of 1.58 m m hr"1 which was 23 times higher than o n silicon. 79 %
of all reversibly adsorbed cells became irreversibly adsorbed over f i ™ =
on
copper.
Hence,
surface
roughness
strongly
influenced
bacterial
63
accumulation on a substratum without regard to the material. Vanhaecke et
al.
reported
an
stronger
influence
of
surface
roughness
for
less
hydrophobic cells but, overall a positive curvelinear correlation between
surface
roughness
of
stainless
Pseudomonas aeruginosa.
surface
roughness
are
steel
and
the
rate
of
Surface hydrophobicity of the
directly
related.
Therefore,
adsorption
of
substratum and
a
high
surface
hydrophobicity could exaggerate the effect of high surface roughness on
cell accumulation at solid-water interfaces. A low surface hydrophobicity
on the other hand could depress the effect of high surface roughness on
bacterial adsorption.
REVERSIBLE ADSORPTION
IRREVERSIBLE ADSORPTION
COPPER 316 STAINLESS GLASS
SILICON
Figure 25.
Comparison of reversible and irreversible adsorption velocity
of Pseudomonas aeruginosa on various substrata. Shear stress 0.75.
64
There w a s n o cell replication of Pseudomonas aeruginosa o n copper
(Figure 4), wh i c h indicates that the adsorbed cells w e r e metabolically
inactive,
dead o r severely inhibited. With sufficient exposure time,
a
cell monolayer accumulated on the copper substratum an d replication of
cells in the upper bacteria layers were observed. The greater the distance
of
the
cells
from
the
copper
surface
and,
consequently,
a
lower
concentration of copper ions in the cell microenvironment (concentration
gradient) m a y have reduced the inhibitory effect. T he reduced toxicity of
copper
ions
could
also
be
a
result
extracellular polymer in the biofihn.
of
copper
ccmplexation
In any case,
by
the
biofilms provide a
protection for the cells from the toxic effects of copper.
T h e rate of replication was found t o be lower for cells adsorbed to
t h e monocrystalline silicon or stainless steel surface than for cells
adsorbed t o glass. The generation time was 2.1 hours
t h e stainless steel surface and 2.5 hours
Escher
(1986)
(Kg = 0.33 hr'1) on
(Kg = 0.28 hr"1) o n silicon.
found a generation time for Pseudomonas aeruginosa o n a
smooth glass substratum of 1.36 hours (Kg = 0.551 hr"1) a nd a shear stress
dependent
separation
rate
of
0.067
hr'1.
The
investigated surfaces but copper were found to b e
growth
rates
higher than the
on
all
growth
(n =
0.2 hr'1) but lower than the m aximum growth rate
of Pseudomonas aeruginosa
(//max = 0.45 hr'1). The reason for the higher
rate in t h e chemostat
growth rate o n the silicon and 316 stainless steel substrata than in the
chemostat could be that the substrate concentration in t he flow cell bulk
wat e r
(40
chemostat
gm"3)
was
higher
than
the
substrate
concentration
in
the
(5 gm"3) . T h e growth rates for adsorbed cells were below the
m a x i m u m growth rate for planktonic cells and the lower temperature a t the
65
solid-water interface (20-24°C) m a y have been a contributing cause. Also,
t h e energy expended in initial adsorption of a cell t o t he metal should
n o t b e neglected. The energy of adsorption could explain t he slower growth
o n t h e smoother silicon surface where the cell adsorption strength w as
lower t h a n o n rougher surfaces.
Q n 316 stainless steel,
growth m a y b e
inhibited b y nickel and chromium ions which diffuse through the metal
oxide layer. 316 stainless steel contains nickel and chromium in a large
amount
(approximately 20 and 3 %, respectively). However,
diffusion of
these toxic metal ions m u s t occur at a lower rate o n stainless steel than
o n copper. The thickness of the metal oxide layer wa s n o t determined, b ut
the
oxide
layer
on
316
stainless
steel
was
found
to
be
much
more
homogeneous than the metal oxide layer on copper b y surface roughness
measurements of the oxidized surfaces (RMS surface roughness of 120
1680
A,
oxide
A
and
respectively). Hence, the composition and thickness of the mAtai
layer
provide
the
diffusion,
barrier
limiting,
cell
growth
on
metallic substrata.
For m o s t of the investigated materials the adsorption velocity h a d
a slightly higher value during the first hour of exposure. There m a y b e
preferred sites on the surface where cells adsorbed faster. After most of
these sites are occupied, adsorption becomes more energy intensive for t he
cells, hence the rate of adsorption w as lower. W h y specific sites w e r e
preferred t o others cannot be answered from these experiments.
Dislocations smaller than 0.1 ^m
on the polished surfaces d id not
influence the rate of adsorption as determined b y visual observation. T he
general
surface
roughness,
adsorption and desorption.
however,
exerted
a
large
influence
on
it
66
Surface colonization on the test substrata can b e simulated w i t h the
rate coefficients (Table 4).
Adsorption and desorption control the rate
of accumulation in the very early stages of surface colonization (linear
cell
accumulation
over
time), but
become
less
important
over
time.
Subsequently, growth and erosion became the rate determining processes
(exponential cell accumulation over ti m e ) . Growth a nd erosion are both
first order rates w i t h respect to cell surface density, whereas adsorption
a n d desorption are zero order with respect to cell surface density and
first
order
with
respect
to
bulk
liquid
cell
concentration.
With
a
constant cell concentration in the bulk water the Cell surface density
increases
linear
for the
first
200 minutes.
For
longer time periods
cumulative growth and erosion of cells lead to exponential increase and
dominate t h e accumulation of cells at the surface.
T he simulation of early surface colonization for various nutrient
conditions can be simulated with the Monod equation (Equation 9). Esther's
model contributes the effect of shear stress. The simulation model w a s
tested for accuracy o n 316 stainless steel and resulted in a 14 % faster
total surface colonization than observed experimentally. T he differences
m a y b e explained b y the heterogeneous cellular colonization observed as
compared t o the homogeneous colonization as assumed i n t he model. Escher
(1986)
hypothesized
that
field" around themselves,
Pseudomonas
aeruginosa
produce
a
"protected
blocking further adsorption in the vicinity.
This wo u l d slow the process of adsorption at high cell surface densities
a n d a less homogeneous colonization pattern results.
67
Comparison of Pseudomonas aeruginosa
and Pseudomonas fluorescens fmot+l
Cellular
accumulation of
Pseudomonas
aeruginosa
a nd
Pseudom onas
fluorescens w a s compared on two different substrata: 316 stainless steel
and glass.
T h e results
Escher (1986)
for Pseudomonas aeruginosa
o n glass were
from
(Figure 26). Sticking efficiency, particle capture factor,
a n d adsorption rate we r e dependent on the substrata in a similar manner
for both species. O n both substrata, adsorption was approximately 5 -H m g g
slower
for
Pseudomonas
fluorescens
than
for
aeruginosa.
Pseudom onag
Consequently the sticking efficiencies for Pseudomonas fluorescens showed
a n 84 % lower value on the stainless steel surface and a n 79 % lower value
o n t h e glass surface than the values obtained for Pseudomonas aeruginosa.
The
probability
fluorescens
on
of
desorption
stainless
steel
was
but
slightly
lower
on
higher
glass.
for
Pseudomonas
Hence
cellular
accumulation of Pseudomonas fluorescens was a much slower process than for
Pseudomonas aerruoinosa.
Pseudomonas aeruginosa is highly hydrophobic, whereas
Psendnm onag
fluorescens h a d a n intermediate cell surface hydrophobicity (Table 3 and
Figure
27).
Therefore
the
cells
with
the
higher
cell
surface
hydrophobicity resulted in a higher adsorption velocity o n the tested
substrata.
To
draw
a
general
conclusion,
however,
more
strains
with
different cell surface hydrophobicity values should b e tested. Varihaedce
et al.
(1989) measured cell surface hydrophobicity and rate of adsorption
of 15 Pseudomonas aeruginosa strains t o stainless steel in a stagnant
system and found higher number of adsorbing cells within 30 minutes
68
cc
I
5
S
z
316 STAINLESS STEEL
O
S
LU
>
Z
GLASS
o
£
CC
O
POLYCARBONATE
CO
Q
<
r
Ps aeruginosa
r
Ps fluorescens+
Ps fluorescens-
Figure 26. Conparison of reversible and irreversible adsorption velocities
t o 316 stainless steel,glass and polycarbonate of Pseudomonas aeruginosa.
Pseudomonas fluorescens (mot+), and Pseudomonas fluorescens (mot-). Shear
stress 0.75 N m'2.
Ps fluor. (mot+)
Ps floor, (mot-)
Ps aeruginosa
Figure 27.
Cell
surface
hydrophobicity
of
the
tested
organisms:
Pseudomonas aeruginosa. Pseudomonas fluorescens (mot+), and Pseudomonas
fluorescens (mot-). Ffeasured as hydrocarbon partitioning factor.
69
contact time for increasing hexadecane partitioning values of the cells.
Combining t h e results from this work with Vanhaecke1s findings, increasing
cell
surface
hydrophobicity
increases
adsorption
velocity,
for
equal
transport conditions.
T h e cellular colonization behavior of Pseudomonas fluorescens
w as
different from Pseudomonas aeruginosa. There were m o r e clusters observed
for t h e colonization of the former, possibly as a result of the lower
sticking
efficiency
(surface
hydrophobicity). W i t h
one
cell
already
adsorbed t o t h e substratum, less energy was required for other cells to
cluster together, than to form n e w cell-surface bonds. Simultaneously, this
m i g h t also b e a reason for a higher probability of desorption. The cell
metal b o n d might have been strong enough t o resist the shear for one cell,
whereas f or two or more, it could b e too w e a k and result in desorption of
t h e v4iole cluster.
Cell
accumulation
was
simulated
for
a
mi x e d
population
of
Pseudomonas aeruginosa and Pseudomonas fluorescens as m o t + and mot- strain
(Figure
24).
fluorescens
Due
to
strains
lower
adsorption
accumulate
much
velocities,
more
slowly
t he
Pspudrmnnag
than
Pspndnmnnag
aeruginosa a t the stainless steel surface. This is true even for longer
exposure times.
In the simulation,
the surface w as totally covered b y
cells after 950 minutes. The area coverage b y both
Pseudom onas
fluorescens
strains w a s about 6 % whereas the faster adsorbing cells of Pseudomonas
aeruginosa covered 94 % of the total surface area. This w i l l also b e true
for t h e other surfaces investigated because of similar differences
adsorption velocity.
cells will
Hence,
in mixed populations,
in
t he faster adsorbing
dominate during early surface colonizations,
even
if both
70
species g r o w a t a similar rate. The difference in adsorption of different
bacterial
species has been used b y others to control mixed microbial
cultures via specific cell adsorption (Roos and Hjortso, 1989 a) a nd Roos
a n d Hjortso,
1989 b ) . A mixture of two Escherichia coli
separated and monitored.
adsorption
between
strains were
The proposed method uses differences
competing
populations
and
maintains
in r ^ n
a
desired
population ratio between even closely related populations.
Comparison of Mot+ and
Mot- Pseudomonas fluorescens
Cellular colonization of motile and non-motile cells w as compared
b y us i n g Pseudomonas
different surfaces
fluorescens as
a mot+ and mot-
strain o n three
(316 stainless steel, glass, and polycarbonate). The
m o t - mut a n t d i d not posses a flagellum as the parent m o t + strain did. Th e
behavior of mot- cells was very similar to the behavior of mot+ cells of
t h e same strain. The mot- cells also h ad a high tendency t o accumulated
in clusters w h e n adsorbing to a substratum. The probability of desorption
was
almost
identical
for
the
two
strains
on
all
the
investigated
substrata. The observed adsorption velocity of motile cells was twice as
h i g h as for non-motile cells (Figure 26). Korber et. al.
(1989) reported
a four fold increase in recolonization rate of motile cells of PseudnmnnAR
fluorescens
than
for
the
mot-
mutants.
However,
there
is
a
large
difference between the transport properties of motile cells a nd non-motile
cells. T h e difference in diffusivity is almost four orders of magnitude,
this results in a high value for sticking efficiency of n o n motile cells
71
and
indicates
a
transport
limitation
rather
than
a
limitation
in
adsorption. N o n motile cells did stick t o the substratum at a relatively
higher
rate
than
motile
cells
and
desorbed
at
a
lower
rate
from
polycarbonate a n d glass, but at a higher rate from stainless steel. This
findings exclude the possibilities that the cell flagellum is of major
importance
for
adsorption
on
the
substratum
as
described
by
others
(Marshall 1982). However, the cellular adsorption velocity and, therefore,
t h e rate of cell accumulation on a substratum is largely dependent on the
cell transport rate from the bulk water to the solid-water interface.
Hence, flagellated cells adsorb and accumulate on a substratum at a higher
rate th a n unflagellated cells.
As
indicated previously,
a
significant difference
in adsorption
velocity wo u l d b e expected for motile and non motile cells of the same
strain.
The
cell
surface
hydrophobicity
of
the
mot-
mutant
was
statistically significantly higher than for the mot+ parent and slightly
lower t h a n for Pseudomonas aeruginosa (Table 3 and Figure 35). Hence, one
w o u l d assume similar values for sticking efficiencies
aeruginosa and Pseudomonas fluorescens
for Pseudomonas
(mot-) to b e in a similar range
between Pseudomonas aeruginosa and Pseudomonas fluorescens (mot+). Using
Bowen's model, motile cells are predicted t o adsorb about IO3 times faster
t h a n m o t - cells, as reflected from the difference in diffusivities. A s t he
observed difference
in adsorption velocity was
small,
th e
calculated
sticking efficiency for non motile cells was 2 orders of magnitude higher
t h a n for t he motile cells. The reason could either b e that mot- cells of
Pseudomonas
fluorescens.
are
much
stickier
despite
cell
surface
hydrophobicity values, or there is an inconsistency in Bowen's model. The
72
mod e l
was
originally
designed
for
small
particles
in
suspension
in
Brownian moti o n and are characterized b y a low diffusivity value. Thus,
Bowen
limited
the
characterized b y
model
to
high
Peclet
h i g h diffusivity values
numbers.
M o tile
and relatively
cells
small
are
Peclet
numbers. Therefore, the model does not apply for motile cells and m a y not
h a v e a h i g h accuracy for non motile cells.
Influence of Starvation on Colonization of
Pseudomonas aeruginosa on 316 Stainless Steel
Adsorption velocity decreased continuously for t h e first 500 hours
of starvation. For
longer durations, the adsorption velocity was about
50
measured
%
of
the
value
concentration in the
for healthy
cells
when
bulk water was used as a basis
t he
viable
cell
for calculation
(Table 7). The adsorption velocity decreased in the same manner t o 30 %
of t h e value found for healthy cells w h e n the total cell concentration wa s
u s e d for calculation. The probability of desorption decreased t o a value
of 27 % of the value for healthy cells.
Starved cells of Pseudomonas aeruginosa were estimated t o swim a t
about 2 - 4
times the speed of healthy cells. The random free runs appear
t o b e shorter b y about half. Hence, the calculated diffusivity for starved
cells (value of D) h a d a slightly higher value than that for healthy cells
b u t h a s a large standard deviation. T he mean diffusivity a t 816 hours of
starvation w a s D c = (1.3+0.8) *10"3 m m 2s'1 and is not significantly different
than
the
diffusivity value
for healthy
cells
(Dc =
IO'3 m m 2s"1) . The
increase in swimming speed could be explained b y the size reduction the
73
cells undergo during the first 24 hours of starvation. T h e shorter free
runs m a y be a response to the low nutrient conditions.
Table 7
Summary of the Kinetic Results from the Starvation Experiment
(shear stress 0.75 N m"2) .
t st [hr]
K a v [mm hr"1]
b [-]
$ [-]
e ["]
0
I .5 6 ± 0 .34
28.1+1.1
0.019+0.005
0.022+0.006
72
1.21+0.29
17.0+2.7
0.010+0.002
0.013±0.003
216
0.90±0.17
10.5+3.3
0.007±0.001
0.010+0.002
348
0.95+0.16
6.7+3.6
0.008+0.002
0.010+0.003
504
0.54±0.16
0
0.004+0.002
0.006+0.003
0.47+0.15
7.7+5.4
0.004±0.001
0.005+0.002
(648-1224)*
t h e v alues for 648, 816, 936 and 1226 hours of duration of starvation
w e r e averaged
Particle capture factor and sticking efficiency c a n b e calculated
w h e n t h e transport properties and the adsorption velocity for the species
are known. T h e particle capture factor (e), sticking efficiency ($), the
adsorption velocity
(Kav) ,. and the probability of
presented in Table 6. Sticking efficiency
desorption
(b)
are
and particle capture factor
decreased 80 % after 500 hours of starvation,
sticking efficiency an d
particle capture factor decreased almost proportionally t o t he adsorption
velocity,
since
diffusivity
(i.
e.
transport
rates)
d id
not
change
significantly.
When
adsorbed t o the
substratum,
the
probability
of
desorption
decreased significantly for starved cells. This can b e explained b y a n
increasing
substrate
flux
for
adsorbed
cells.
Therefore
it wo u l d
be
74
energetically inefficient for the adsorbed cells t o leave the substratum
after t hey adsorbed (energetically unfavorable reaction).
75
CONCLUSIONS
A technique w a s developed t o measure initial cell colonization o n
reflective metal surfaces based o n previous work b y Escher
transparent
substrata:
surfaces,
surfaces.
copper,
and
Experiments
silicon,
polycarbonate
and
were
316
and
conducted
stainless
glass
as
on
(1986)
various
steel
for
solid
as
reflective
transparent
materials.
Experiments w e r e conducted w i th Pseudomonas aeruginosa, a nd Pseudomonas
fluorescens as m o t + and a mot- strain. The results were consistent w i t h
t h e results of Esdher (1986). The following conclusions have been derived
from this work:
1.
Image analysis can b e used t o measure early surface colonization b y
bacteria cells o n reflective surfaces.
2.
Bacterial growth on surfaces depends on the nutrient supplied b y th e
b u l k water as well as properties of the substratum.
Growth w as
initially inhibited on copper surfaces but w a s observed to occur
once a n intervening layer of cells h ad formed.
3.
Surface
roughness,
influence
sorption
in
the
related
micro
range
processes.
(27-150
A),
Adsorption
appears
velocity
to
w as
positively correlated to surface roughness.
4.
Sorption related processes depend on the bacteria species a nd the
cell surface hydrophobicity. Pseudomonas fluorescens (intermediate
76
hydrophobic)
adsorbed at a rate 5 times slower than Pseudomonas
aeruginosa m o t + (highly hydrophobic).
5.
Flagellated cells of Pseudomonas fluorescens adsorb a nd accumulate
o n t h e substratum at a higher rate than unflagellated cells.
6.
Nutrient conditions in the bulk water not only influence t he growth
of
adsorbed cells but also
influence cellular transport t o t he
solid-water interface and the rate of adsorption a nd desorption on
the substratum.
Starved cells of Pseudomonas aeruginosa exhibited
increased cell motility b y size reduction but di d n ot adsorb a t an
increased rate on 316 stainless steel over healthy cells. T he rate
of desorption for starved cells was significantly lower than for
healthy cells.
77
NCMENCIATURE/ SYMBOLS
78
Nomenclature
Adsorption:
substratum.
Adsorption
Brownian diffusivity:
is defined as the
linking of
a
C FU w i t h the
Diffusivity due t o Brownian motion.
C F U : Colony forming unit. It compromises one to several cells and forms
a single colony.
Desorption: Desorption is defined as the breaking of t he linkage of a C FU
and its complete removal from the substratum.
Erosion:
t h e CFU.
Cells within a CFU can detach and reduce th e number of cells of
Growth: Cell separation due to metabolic activity of t he cells. A CF U w as
called separated w h e n the spatial distance between t he t w o separating C F U
exceeded 0.1 ^m.
Non-Brownian diffusivity:
Offset time:
Diffusivity due t o cell motility.
Time difference between initial adsorption and desorption.
79
Symbols
cross-sectional area [L2]
m a i n angle of turn for cellular motility
probability of desorption [-]
diffusivity [L2t"1]
free length of random run [L]
surface particle capture factor [-]
gravimetric cell factor, carbon content p er biomass [MM"1]
fraction of irreversibly adsorbed cells [-]
cell factor [NtL'2]
gamma function [r(4/3) = 0.89338
half-thickness of channel [L]
rate coefficient [t'1]
flux of cells adsorbing to the interface [cell/L'^t]
Boltzman constant, 1.38062*10"23 JK"1
dimensionless distance [-]
adsorption velocity [Lt'1]
coefficient of growth on substratum [t'1]
coefficient of erosion on substratum [t"']
cellular growth saturation constant [ML'3]
length of substratum [L]
growth rate [t"1]
fluid viscosity [NtL'2]
Peclet number [-]
sticking efficiency [-]
flowrate
[L3t'1]
radius [L]
Reynolds number [-]
density [ML"3]
hydraulic radius [L]
substrate concentration [ML'3]
time
[t]
Temperature [K]
shear stress [NL'2]
volume [L3]
velocity [Lt"1]
cellular motility [Lt'1]
cell surface density [cell#L"2]
suspended cell concentration [cell#L'3]
suspended biomass concentration [ML"3]
width-coordinate [L]
length coordinate in axial direction [L]
yield [MM'1]
height coordinate normal to the flow [L]
wetted perimeter [L]
Subscripts
A O
eff
d
I
i
m
t
v
:
:
:
:
:
:
:
bu l k water
cell
effluent
death
lysis
influent
m e a n value
total
viable
81
LITERATURE CITED
82
LITERATURE CITED
Abbott, A . , P. R. Rutter, R. C. W. Berkeley
1983
T h e Influence of Ionic Strength, pH, and a Protein Leyer
Interaction between Streptococcus mutans and Glass Surfaces,
p.439-445. In Journal of General Microbiology, Vol.129
on
the
Adamson, A. W.
1960
Physical Chemistry of Surfaces.
Interscience Publishers, Inc., N e w York
B a k k e , R.
1986
Biofilm Detachment,
Ph.D. Thesis, Montana State University
Biker m a n , J. J.
1970
Physical Surfaces.
Academic Press, N e w York and London
Bowen, B. D., S. Levine, N. Epstein
1976
Fine Particle Deposition in Laminar Flow Through Parallel-Plate
Cylindrical Channels,
p.375-390. In Journal of Colloid Interface Science, Vol.54, No.3
and
Brown, C. M., D. C. Ellwood, J. R. Hunter
1977
p.163—166. In FEMS Microbial. Letters, Vol.l
Characklis, W. G., J. D. Bryers, M. G. Trulear, N. Zelver 1980
Biofouling Film Development and its Effect oh Energy Losses,
p. 49-76. In J. F. Garey (ed.), Condensor biofouling control. A n n Arbor
Science I n c . , A n n Arbor, Mich.
Characklis, W. G . , K. C. Marshall
Biofilms.
J o h n Wil e y & Sons, I n c . , N e w York
1990
E s c h e r , A. R
1986
Colonization of a Smooth Surface b y Pseudomonas aeruginosa: Image Analysis
Methods,
PH.D. Thesis, Montana State University
Fletcher, M.
1980
Adherence of Marine Micro-Organisms t o Smooth Surfaces,
p . 347-374.
In
Beachey
(ed.)
Bacterial
Adherence
(Receptors
and
R e c ognition. Series B, Volume 6) published b y Chapman an d Hall, London.
Jang, L. K., T. F. Y e n
1985
A Theoretical Model of Convective Diffusion of Motile an d Non-Motile
Bacteria Toward Solid Surfaces,
p.226-246.
In Microbes
and Oil
Recovery,
V o l . I,
International
Bioresources Journal
83
K e f f o r d , B., K. C. Marshall
1984
A d h esion of Leptospira at Solid-Liquid Interface: a Model,
p.84-88. In Archives of Microbiology (138), Springer Verlag
Kim, Y. C.
1990
Diffusion
Coefficients
for
Pseudomonas
pnomoniiae.
Mas t e r Thesis, Montana State University
aeruginosa
a nd
Klabsiella
Kjelleberg, S . , B. A. Humphrey, K. C. Marshall
1982
Effect of Interfaces on Small, Starved Marine Bacteria,
p. 1166-1172. In Applied & Environmental Microbiology Vol.43, No.5
Kjelleberg, S. M. Hennansson
1984
Starvation Induced Effects on Bacteria Surface Characteristics,
p.497-503. In Applied and Environmental Microbiology, Vol.48, No3
Kjelleberg, S . , K. C. Marshall, M. Hemansson
1985
Oligotrophic and Copiotrophic Marine Bacteria - Observations Related t o
Attachment,
in FEMS Microbiology Letters 0, E l s e w i e r Biomedical Press
K o r b e r , D. R . , J. R. Lawrence, B. Sutton, D. E. Caldwell
.1989
Effect of Laminar Flow Velocity on the Kinetics of Surface Recolonization
b y M o t + a nd Mot- Pseudomonas fluorescens.
p.1-19. Microbial Ecology, Springer Verlag N e w York Inc.
Marshall, K.
C.
1979
Growth a t Interfaces,
p. 281-290. Berlin: Dahlem Konferenzen
Marshall, K.
C.
1980
Microorganisms and Interfaces,
p.246-249. In BioScience Vol.30, No.4
Marshall, K.
C.
1985
Bacterial Adhesion in Oligotrophic Habitats,
p.321-326. In Microbiological Sciences, V o l . 2, No.11
M o r i t a , R. J.
1982
Starvation of Heterotrophs in the Marine Environment,
p. 171-198. In K.C. Marshall (ed.), Advances in Microbial Ecology. Vol.6,
Plenum Publishing C o r p . N e w York
M o r i t a , R. J.
1988
Bioviability of Energy and its Relationship t o Growth a nd
Survival in Nature,
p.436-441. In Canadian Journal of Microbiology, Vol.34, N o.4
Starvation
Paul, J. H . , W. H. Jeffrey
1985
Evidence for Separate Adhesion Mechanisms of Hydrophilic a nd Hydrophobic
Surfaces in Vibrio proteolvica.
84
P * 431-437. In Applied and Environmental Microbiology, Vol.50, No.2
Pedersen, K., C. Holmstroem, A.-K. Olsson, A. Pedersen
1986
Statistic Evaluation of the Influence of Species Variation, Culture
Conditions, Surface Wettability and Fluid Shear on Attachment and Biofilm
Development of Marine Bacteria,
in Arch, of Microbiology, Springer V e r l a g .
Poindexter, J. S.
1981
Fast a n d Famine Existence,
p.63-89. I n Advances in Microbial Ecology, Vol.5
Powell, M. S., N. K. H. Slater
1983
T h e Deposition of Bacterial Cells from Laminar Flows onto Solid Surfaces,
p.891-900. In Biotechnology and Bioengineering, V o l . 25
R o o s , J. W., M. A. Hjortso
1989(a)
Control of Mixed Microbial Cultures via Specific Cell Adhesion,
p.638-649. In Biotechnology and Bioengineering, V o l . 33
R o o s , J. W., M. A. Hjortso
1989(b)
Dtermination of Population Balances in a Mixed Culture b y Specific Cell
Adhesion,
p.7-12. In Biotechnology Techniques, V o l . 3, No. I
Rosenberg, M., D. Gutnick, E. Rosenberg
1980
Adherence of Bacteria t o Hydrocarbons: a Simple Method for Measuring Cell
Surface Hydrophobicity,
p.29-33. In FEMS Microbiology Letters 9
Robinson, J. A . , M. G. Tralear, W. G. Characklis 1984
Cellular reproduction and extracellular polymer formation b y Pseudomonas
aeruginosa in Continuous Culture,
p . 1409-1417. In Biotech, and Bioeng. Vol.26
Rosenberg, M.
1984
Bacterial Adherence t o Hydrocarbons: a useful Technique for Studying Cell
Surface Hydrophobicity,
p.289-295. In FEMS Microbiology Letters 22
S e g e l , L. A., I. Chet, Y. Henis
1977
A Simple Quantitative Assay for Bacterial Motility,
p.329-337. In Journal of General Microbiology, Vol.98,
Stanley, P. M.
1983
Factors affecting the Irreversible Attachment of Pseudomonas aeruginosa
t o Stainless Steel,
p.1493-1499. In Canadian Journal of Microbiology, V o l .29
Stevens, T. O., R. L. Crowford, D. L. Crowford
1990
Biodegradation of Dinoseb in Several Idaho Soils w i t h Various Dinoseb
Exposure Histories,
85
p.133-139. In Applied and Environmental Microbiology, Vol.56, No.I
Stover, J. C. S. A. Serati
1984
Calculation of Surface Statistics From Light Scatter,
p.406-412. In Optical Engineering, Vol.23, No.4
T r u l e a r , M. G.
1983
Cellular Reproduction and Extracellular
Developments of Biofilms,
Ph.D. Thesis, Montana State University
Polymer
Formation
in
the
T u r a k h i a , M. H.
1986
T h e Influence of Calcium o n Biofilm Processes,
PH.D. Thesis, Montana State university
V a i t u z i , Z., R. N. Doetsch
1969
Motility Tracks: a quantitative Study of Bacterial Movement,
p.584—588. In Applied and Environmental Microbiology, Vol.17
Varihaecke, E . , and J. Pijck
1988
Bioluminescence As s a y for Measuring the Number of Bacteria Adhering t o the
Hydrocarbon Phase in t h e BATH Test,
p.1436-1439. In Applied and Environmental Microbiology, Vol.54
V a n h a e k e , E., J. P. Remon, M. Moors, F. R a e s , D. D e Rudder, A. V a n
Petighem
1990
Kinetics of Pseudomonas aeruginosa Adhesion to 304 a n d 316-L Stainless
Steel: R o l e of Cell Surface Hydrophobicity,
p.788-795. In Applied and Environmental Microbiology, Vol.56, No.3
Zisman, W. A.
1964
p.1-51. In A d v . C h e m . S e r . Vol.43,
86
APPENDIX
Cell Accumulation of Pseudcanonas
aeruginosa on 316 Stainless Steel
a t Various Durations of Starvation
88
FIGURE 28. CHEMOSTAT EFFLUENT D = 0.2 1/HR
0.5 N/SQM; 1.0 E +06 CELLS/ML
16000
ACCUMULATION
14000-
3
DESORPTION
-9KADSORPTION
1200010000 8000-
O
6000-
3
4000-
2000
-
10
20
FIGURE 29.
30
40 50 60 70
TIME IN MINUTES
80
90
100
DURATION OF STARVATION: 48 HOURS
4.0 E +06 CELLS/ML
ACCUMULATION
DESORPTION
— *r—
ADSORPTION
Q
O
0
10
20
30 40 50 60 70
TIME IN MINUTES
80
90
100
89
FIGURE 30.
DURATION OF STARVATION: 72 HOURS
5.0*E6 CELLS/ML
CELL SURFACE DENSITY (CFU/SQMM)
16000
ACCUMULATION
14000-
DESORPTION
12000 -
— we- ■
ADSORPTION
1000080006000
4000-
2000
-
0
10
20
FIGURE 31.
30 40 50 60 70
TIME IN MINUTES
80
90
100
DURATION OF STARVATION: 120 HOURS
5.0*E6 CELLS/ML
CELL SURFACE DENSITY (CFU/SQMM)
16000
ACCUMULATION
14000-
DESORPTION
12000 -
ADSORPTION
10000
800060004000-
2000
-
0
10
20
30
40
50
60
70
TIME IN MINUTES
S
80
90
100
90
FIGURE 32.
DURATION OF STARVATION: 168 HOURS
1.5*E6 CELLS/ML
CELL SURFACE DENSITY (CFU/SQMM)
16000
ACCUMULATION
14000-
DESORPTION
12000-
ADSORPTION
10000 800060004000-
2000
10
20
FIGURE 33.
30 40 50 60 70
TIME IN MINUTES
80
90
100
DURATION OF STARVATION: 216 HOURS
3.0*E6 CELLS/ML
CELL SURFACE DENSITY (CFU/SQMM)
16000
ACCUMULATION
14000-
DESORPTION
12000-
- x -
ADSORPTION
10000 800060004000-
2000
-
10
20
30 40 50 60 70
TIME IN MINUTES
80
90
100
91
FIGURE 34.
DURATION OF STARVATION: 348 HOURS
1.75*E6 CELLS/ML
CELL SURFACE DENSITY (CFU/SQMM)
16000
14000-
ACCUMULATIOh
12000-
DESORPTION
10000-
ADSORPTION
800060004000-
2000
-
10
20
FIGURE 35.
30
40
50
60
70
TIME IN MINUTES
80
90
100
DURATION OF STARVATION: 504 HOURS
2.0*E6 CELLS/ML
CELL SURFACE DENSITY (CFU/SQMM)
16000
ACCUMULATION
14000-
DESORPTION
12000 -
— Me—
ADSORPTION
10000 800060004000-
2000
-
0
10
20
30
40
50
60
70
TIME IN MINUTES
80
90
100
92
FIGURE 36.
DURATION OF STARVATION: 648 HOURS
1.7*E6 CELLS/ML
CELL SURFACE DENSITY (CFU/SQMM)
16000
ACCUMULATION
14000-
DESORPTION
12000 -
— )K —
ADSORPTION
10000800060004000-
2000
-
0
10
20
FIGURE 37.
30 40 50 60 70
TIME IN MINUTES
80
90
100
DURATION OF STARVATION: 816 HOURS
1.7*E6 CELLS/ML
CELL SURFACE DENSITY (CFU/SQMM)
16000
ACCUMULATION
14000-
DESORPTION
12000-
—>K -
ADSORPTION
10000 800060004000-
2000
-
10
20
30 40 50 60 70
TIME IN MINUTES
j
80
90
100
93
CELL SURFACE DENSITY (CFU/SQMM)
FIGURE 38.
DURATION OF STARVATION: 936 HOURS
1.8*E6 CELLS/ML
16000
14000-
ACCUMULATIOI'
12000
-
DESORPTION
10000 -
ADSORPTION
800060004000-
2000
-
10
20
FIGURE 39.
30
40
50
60
70
TIME IN MINUTES
80
90
100
DURATION OF STARVATION: 1224 HOURS
1.7*E6 CELLS/ML
CELL SURFACE DENSITY (CFU/SQMM)
16000
ACCUMULATION
14000-
DESORPTION
12000ADSORPTION
10000 800060004000-
2000
-
TIME IN MINUTES
S
MONTANA STATE UNIVERSITY LIBRARIES
111I II 11111IIIIII
3
713 2
IO C 1 6 9 4 OO
7