PB91-164277

Movement of Bacteria through Soil and Aquifer Sand

Cornell Univ., Ithaca, NY
Prepared for:

Robert S. Kerr Environmental Research Lab., Ada, OK
Mar 91

U.S. DEPARTMENT OF COMMERCE

National Technical Information Service

NTIS


-------
EPA/600/2-91/010
March 1991

MOVEMENT OF BACTERIA THROUGH SOIL AND AQUIFER SAND

by

M. Alexander, R. J. Wagenet, P. C. Baveye,
J. T. Gannon, U. Mingelgrin, and Y. Tan
Cornell University
Ithaca, New York 14853

Number CR-814487

Project Officer

John T. Wilson
Processes and Systems Research Division
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820

ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ADA, OKLAHOMA 7482 0


-------
TECHNICAL REPORT DATA

(Please read Instructions on the reverse before complet'

1. REPORT NO. 2.

EPA/600/2-91/010

3- PB91- 1 64277

4. TITLE AND SUBTITLE

MOVEMENT OF BACTERIA THROUGH SOIL AND AQUIFER SAND

5. REPORT DATE

March 1991

6. PERFORMING ORGANIZATION CODE

7. AUTHORIS)

M. Alexander, R.J. Wagenet, P.C. Baveye, J.T. Gannon,
U. Mingelgrin, and Y. Tan

8. PERFORMING ORGANIZATION REPORT NO.

9. PERFORMING ORGANIZATION NAME AND ADDRESS

Cornell University
Ithaca, NY 14853

10.	PROGRAM ELEMENT NO.

CEWD1A

11.	CONTRACT/GRANT NO.

CR-814487

12. SPONSORING AGENCY NAME AND ADDRESS

Robert S. Kerr Environmental Research Lab. -Ada, OK
U.S. Environmental Protection Agency
P.O. Box 1198
Ada, OK 74820

13. TYPE OF REPORT AN D PE R IOD COVE RED

Proiec.f. Report 10/01/87-12/31,

14. SPONSORING AGENCY CODE

EPA/600/15

15, SUPPLEMENTARY NOTES

Project Officer: John T. Wilson FTS: 743-2259

16. ABSTRACT

The transport of microorganisms in soils is of- major importance for
bioremediation of subsurface polluted zones. A procedure for
evaluating the relative mobility and recovery of bacteria in the
soil matrix was developed. Nineteen bacterial strains were
selected that differed in their ability to be transported through
soils. Measurements were made of sorption partition coefficient,
hydrophobicity, net surface electrostatic charge, zeta potential,
cell size, encapsulation, and flagellation of the cells. Only
sorption and cell length were correlated with transport of the
bacteria through soil. The breakthrouqh curves for Pseudomonas so.
KL2 moving through a column packed with a sandy aquifer material
were determined. Ionic strength of the inflowing solution,
bacterial density, and velocity of water flow were found to have an
effect on breakthrough.

17 KEY WORDS AND DOCUMENT ANALYSIS

a DESCRIPTORS

b. 1 DENTI FlERS/OPEN ENDED TERMS

e, COS ATI Field/Group







18. DISTRIBUTION STATEMENT

RELEASE TO THE PUBLIC
	—				—		

19. SECURITY CLASS (This Report)

UNCLASSIFIED

21. NO. OF PAGES

43

20. SECURITY CLASS (This page)

UNCLASSIFIED

22. PRICE

EPA Form 2220-1 (Re*. 4-77) previous eoition isobsolete


-------
DISCLAIMER

The information in this document has been funded wholly or in
part by the U.S. Environmental Protection Agency under assistance
agreement CR-814487 to Cornell University- It has been subjected
to the Agency's peer and administrative review, and it has been
approved for publication as an EPA document.

ii


-------
FOREWORD

EPA is charged by Congress to protect the Nation's land, air,
and water systems. Under a mandate of national environmental laws
focused on air and water quality, solid waste management and the
control of toxic substances, pesticides, noise and radiation, the
Agency strives to formulate and implement actions which lead to a
compatible balance between human activities and the ability of
natural systems to support and nurture life.

The Robert S. Kerr Environmental Research Laboratory is the
Agency's center of expertise for investigation of the soil and
subsurface environment. Personnel at the Laboratory are
responsible for management of research programs to: (a) determine
the fate, transport and transformation rates of pollutants in the
soil, the unsaturated and the saturated zones of the subsurface
environment; (b) define the processes to be used in characterizing
the soil and subsurface environment as a receptor of pollutants;
(c) develop techniques for predicting the effect of pollutants on
ground water, soil, and indigenous organisms; and (d) define and
demonstrate the applicability and limitations of using natural
processes, indigenous to the soil and subsurface environment, for
the protection of this resource.

There is a widely held perception that bacteria cannot move
appreciable distances through geological materials. This view is
based on experience with human enteric microbes such as E^_ coli.
which are effectively sorbed and inactivated in the subsurface.
This view has inhibited research on inocula to degrade organic
wastes in the deeper subsurface. This research establishes that
there are organisms that can be readily transported in the
subsurface and opens up a new strategy for the development and use
of inocula for hazardous waste in the deeper subsurface.

Clinton W. Hall
Director

Robert S. Kerr Environmental

Research Laboratory

iii


-------
ABSTRACT

The transport of microorganisms in soils is of major
importance for bioremediation of subsurface polluted zones. A
procedure for evaluating the relative mobility and recovery of
bacteria in the soil matrix was developed. In the method devised,
movement of bacteria along the walls of the column of soil and
channeling were prevented. Changes in population size during the
test period were minimal because temperatures of 2 to 5°C were
maintained during the test period and predators and parasites were
eliminated by 60Co irradiation. The 19 strains of bacteria tested
had markedly different degrees of transport. From 0.01 to 15% of
the added cells passed through a 5-cm long column of Kendaia loam
with four pore volumes of water. From 4.3% to essentially all of
the added bacteria were recovered. The marked differences in the
mobilities of the various bacteria and the high recoveries of most
of the isolates suggested that the procedure developed is a useful
means for selecting bacteria according to their mobilities in
soils, aquifer materials, and other porous media.

Measurements were made of hydrophobicity, net surface
electrostatic charge, zeta potential, cell size, encapsulation, and
flagellation of cells of 19 bacterial strains that differed in
their ability to be transported through soil. A wide range of
hydrophobicities was noted among the bacterial strains.
Determinations by electrostatic-interaction chromatography showed
that 5 to 47% of the cells were retained by a cation-exchange
resin. The zeta potentials of the various isolates ranged from -8
to -3 6 mV. Eight of the bacteria had capsules, and the cell
lengths varied from 0.7 to 2.6 /Jm. Some of the mobile strains were
flagellated. Sorption and cell length were correlated with
transport of the bacteria through soil.

The breakthrough curves for Pseudomonas sp. KL2 moving through
a column packed with a sandy aquifer material were determined.
Ionic strength of the inflowing solution, bacterial density, and
velocity of water flow were found to have an effect on
breakthrough.

iv


-------
CONTENTS

Pages

Disclaimer	ii

Foreword	iii

Abstract	iv

Figures	vi

Tables	vii

INTRODUCTION 		1-3

MATERIALS AND METHODS 		3-10

RESULTS	10-23

DISCUSSION	23-30

References	31-35

v


-------
FIGURES	Page

Figure 1. Experimental setup.	4

Figure 2. Breakthrough of Pseudomonas sp. KL2 with
0.01 M NaCl (flow rate 1.0 X 10"4 m s"1 equivalent to 3.5
pore volume h"1).	19

Figure 3. Breakthrough of Pseudomonas sp. KL2 with
deionized water (flow rate 1.0 X 10~4 m s"1, equivalent to
3.5 pore volumes h"1).	2 0

Figure 4. Breakthrough of Pseudomonas sp. KL2 with
0.01 M NaCl solution and flow rate of 2.0 X 10"4 m s"1
equivalent to 6.9 pore volumes h'1.	21

Figure 5. Breakthrough of Pseudomonas sp. KL2 with
deionized water at flow rate of 2.0 X 10"4 m s~1
equivalent to 6.9 pore volumes h"1.	22

Figure 6. Breakthrough of Pseudomonas sp. KL2 with
0.01 M NaCl at 1.0 X 109 cells mL"1 (flow rate 1.0 X 10-4
m s"1 equivalent to 3.5 pore volumes h"1).	2 4

Figure 7. Breakthrough of Pseudomonas sp. KL2 with
deionized water at 1.0 X 109 cells mL"1 (flow rate 1.0 x
10-4 m s"1 equivalent to 3.5 pore volumes h*1) .	25

Figure 8. Breakthrough of Pseudomonas sp. KL2 at
1.0 X 109 cells mL"1 using 0.01 M NaCl for 1 h followed by
deionized water (flow rate 1.0 X 104 m s"1 equivalent to
3.5 pore volumes h"1) .	2 6

vi


-------
TABLES

Table 1. Distribution of Pseudomonas KL2 in columns
of soil after passage of four pore volumes of water.

Table 2. The numbers and percentages of bacteria
transported and recovered.

Table 3. Adsorption coefficients of bacteria.

Table 4. Relationship between characteristics of
bacteria and their transport through soil.

vii


-------
INTRODUCTION

Approximately 50% of the United States population depends upon
ground water as a source of drinking water (Bitton and Gerba,
1984). Consequently, the contamination of ground water by organic
chemicals is widely recognized as a critical environmental problem.
The transport and mobility of bacteria in soil and subsurface
materials has been a subject of interest for the past few decades
because of the environmental importance of these organisms. The
bioremediation of underground waste-disposal sites by the use of
introduced bacteria requires that the microorganisms move from the
point of their introduction to the site of contamination. Such
inoculation is necessary if microorganisms degrading the chemical
contaminants are not present in the hazardous waste site or
adjacent groundwaters.

Many toxic organic chemicals persist at underground hazardous
waste sites despite being readily biodegradable under laboratory
conditions. When this occurs, introduced bacteria selected for
their capacity to degrade target contaminants, and capable of
surviving and proliferating after injection in the aquifer, might
be used to promote biodegradation. The introduction of bacteria to
degrade toxic wastes in soil and subsurface has generated renewed
interest in the transport and mobility of bacteria (Baveye and
Valocchi, 1989; Borden and Bedient, 1986; Gannon et al. , 1991).
However, the mobility of the added bacteria may determine their
effectiveness for in situ bioremediation. If biodegradative
bacteria are not present at the site of contamination, then
bioremediation will only be effective if the added bacteria have
both the capacity to reach the contaminated zone and to move
through porous materials with the contaminant plume.

Considerable attention has been given to the mobility of
bacteria and other microorganisms in soil and subsurface materials.
These studies were conducted primarily because of concern with the
dissemination of pathogens from land spreading operations,
groundwater recharge, or the disposal of manure or municipal sludge
(e.g., Gerba et al., 1975; Brown et al. , 1979; Bell and Bole,
1978) . Several studies have shown poor mobility of the
investigated species of bacteria through soil (Bitton et al., 1974;
Dazzo et al., 1973; Madsen and Alexander, 1982). However,
considerable movement of some bacteria was observed in field
studies (Schaub and Sorber, 1977; Viraraghavan, 1978).

It is unclear whether the movement of bacteria that has been
observed occurred through the soil matrix or through the macropores
or channels that afford the organisms a relatively unhindered
passage (Hagedorn et al., 1981; Rahe et al., 1978). It is,
however, generally acknowledged that the movement of bacteria in a
homogeneous porous medium is poor because of both adsorption
(Hattori and Hattori, 1976; Marshall, 1980) and mechanical
filtration (Pekdeger and Matthess, 1983; Smith et al., 1985) of

1


-------
bacterial cells, which have been suggested as mechanisms for their
retention in soils. In a study by Wollum and Casell (1978) ,
streptomycete conidia and a bacterium were displaced through a sand
column; only 0.2% of the bacterial cells and 6% of the conidia were
recovered in the effluent after passage of four pore volumes of
water, but over 90% of the organisms were recovered in the sand
within 3 cm of the inlet surface. When bacteria were injected into
a sandy aquifer using a forced gradient, the relative breakthrough
of cells into a sampling well located 1.7 m from the injection well
was less than 1% (Harvey et al., 1989). Smith et al. (1985) found
nearly a 16-fold greater breakthrough of Escherichia coli through
intact cores compared to disturbed cores of a sandy loam. They
concluded that soil structure and the velocity of water flow are
critical in determining movement of bacteria through soil.

Baveye and Valocchi (1989) reviewed a number of mathematical
models describing the concurrent growth of bacteria and transport
of biodegradable substrates in saturated porous media; these models
assume that bacteria form either continuous biofilms or discrete
microcolonies on surfaces of solid particles and hence are immobile
(Rittmann et al., 1980; McCarty et al., 1984; Molz et al., 1986).
However, many environmental factors greatly affect bacterial
transport. Ionic strength is particularly important. Krone et al.
(1958) and Tan (1989) showed that infiltrating solutions with low
ionic strength decrease retention of bacteria in sand. In acid-
treated sand, the efficiency of coliform retention was higher when
the bacteria were suspended in tap water than in distilled water,
and no retention occurred when the bacteria were suspended in
triple-distilled water (Goldshmid et al., 1973). On the basis of
the electrical double layer developing in the vicinity of charged
surfaces, Marshall (1975) concluded that the number of bacteria
attracted to surfaces lessened with decreased electrolyte
concentrations, whereas adsorption of cells to surfaces increased
with higher electrolyte concentrations. Duboise et al., (1976),
Gerba and Lance (1978), and Sosbey et al. (1980) reported that the
application of rain water or distilled water will result in the
desorption of viruses from soil particles. Vinten and Nye (1985)
reported that complete blockage of pores due to soil dispersion
does not occur in sodic soils leached with water of low electrolyte
content, and that dispersed clay may be transported considerable
distances before deposition. Despite these findings on the effects
of ionic strength, recent models for bacterial transport in porous
media do not include ionic strength as a factor in determining the
movement of cells (Corapcioglu and Haridas, 1984, 1985; Germann et
al., 1987; Taylor and Jaffe, 1990).

The effect of ionic strength on movement of bacteria through
subsurface earth materials seems to have been largely ignored in
terms of promoting bioremediation with introduced bacteria.

The objectives of this study were to develop a reproducible
procedure that would yield consistent measurements of relative

2


-------
mobility of bacteria in soil by avoiding uncontrolled variations in
bacterial behavior and to relate transport to efficiency of
recovery and adsorption of the cells. In the procedure that was
developed, flow through macropores did not occur. In addition, a
study was conducted to determine the influence of certain
properties of bacterial cells on their movement through soil. The
traits investigated were net surface electrostatic charge,
hydrophobicity, cell size, and presence of capsules. Disturbed
columns of aquifer sand were used to study the breakthrough of
bacteria and of a chloride tracer. The effect of variations in the
ionic strength of an inflowing solution, bacterial cell density,
and flow velocity on the transport of bacteria through the earth
materials was evaluated.

MATERIALS AND METHODS

TRANSPORT THROUGH SOIL

Bacteria able to degrade benzene, chlorobenzene, or toluene
were isolated by enrichment culture in solutions containing 100 mg
of the organic compound, 1.6 g K2HP04, 0.4 g KH2P04, 0.5 g (NH4)2S04,
25 mg CaS04-2H20, 0.2 g MgSO^- 7H20, and 2.3 mg FeSO^-7H^O per L of
deionized water. The pH was adjusted to 7.0. Each enrichment was
transferred at least four times into fresh medium of the same
composition before plating the bacteria on enrichment medium
supplemented with 1.5% agar. Biodegradation was determined by
spectrophotometric measurement of the loss of UV absorbance of the
added organic compound.

To obtain bacteria by a method that presumably favored mobile
organisms, diesel-fuel contaminated soil, Kendaia loam, and liquid
from the primary settling tank of the Ithaca, NY sewage treatment
plant were placed on the soil column, and bacteria that passed
through the soil were isolated on Trypticase soy agar (BBL
Microbiology Systems, Cockeysville, MD). Bacillus sp. CU 4519 was
obtained from S.A. Zahler, and Arthrobacter sp. Lula D from an
underground aquifer was provided by J. L. Sinclair, both from
Cornell University.

The bacteria were grown in Trypticase soy broth at 30°C for 24
to 48 h, the cells were harvested by centrifugation and washed
twice, and the organisms were suspended and diluted in 0.9% NaCl
solution. The cell suspensions were cooled to 0° for 1 h prior to
their addition to columns of soil. Bacterial counts were made on
Trypticase soy agar using triplicate samples from the effluent of
each soil column and triplicate plates per dilution. The plates
were incubated at 30°C for 24 to 48 h.

The soil was air-dried, ground, passed through a 2-mm sieve,
and then sterilized by 60Co irradiation (2.5 Mrad). Aseptic
conditions were maintained during the experiments.	The

experimental setup (Fig. 1) was as follows. The soil was packed in

3


-------
10 c /r,-

Deionized

h2o

Buchner
Funnel
600 ml

. Petrolatum
Coating

Soil

Fritted Disc
(40-60 /^m)

Hg Column
Row Regulator
Hg

500 ml
Erlenmeyer
Flask

Collected
Effluent

Figure I. Experimental Setup


-------
600-mL Buchner funnels, 10-cm diam, containing a fritted-glass disc
with pore sizes of 4 0 to 60 jL/m. Preliminary tests indicated that
the pore sizes in the disc permitted passage essentially only of
bacteria and water-dispersible clay, and during experiments, the
flow rate remained constant for each of the four pore volumes
passed through the soil column. Thus, the filters did not become
clogged, and this allowed for free passage of bacteria. The funnel
was attached to a 500-mL Erlenmeyer flask fitted with a sidearm
containing Hg in a bulb to control the flow rate. The Hg in the
bulk restricted the rate of flow by controlling the release of air
displaced as drops of effluent passed from the column into the
attached Erlenmeyer flask. The walls of the funnel were coated
with a thin layer of sterile petrolatum added in liquid form at
>60°C. The petrolatum sealed the interface between the funnel
walls and the soil. The soil was pressed down with a fitted brass
compacting plate to a depth of 5 cm to give a bulk density of
1.08±0.062 g cm"3 and porosity of 59.2%. The pore volume was 241.75
cm3.

The soil columns and sterile deionized water to be used for
mobility tests were cooled to 3.5±1.5°C, the temperature at which
experiments were conducted, and the soil was then again compressed
with the compacting plate to compensate for any disturbance of the
soil column in the period between packing and inoculation. The
soils were moistened from below by connecting a bottle of water to
the stem of the Buchner funnel. Sterile deionized water was added
at a rate of approximately 10 cm h"1 until the water level was
approximately 2 cm above the soil surface. The water was then
drained to the surface of the soil column, and a circular mound
with a radius of 1 cm of dry sterile soil was placed in the center
of that mound. A 1.0 mL inoculum of 1.0 x 108 cells was placed in
the center of that mound. An additional 1-cm layer of sterile soil
was added to the top of the column, thereby covering the inoculated
mound and making the column surface approximately level to prevent
horizontal movement of the inoculum. Deionized water was added to
the soil surface. Physical disturbance of the surface of the soil
column was avoided by pouring the water onto A1 strips placed above
the soil surface. A total of four pore volumes of effluent was
collected in the flask. The rate of water flow through the column
was maintained at approximately 0.8 pore volumes per h.

To determine the extent of lateral movement of the
microorganisms and the possible transport of bacteria along the
walls of the column, the inoculum was introduced at the center of
the soil surface in a circle of approximately 1-cm diameter. After
the column was leached with four pore volumes and allowed to drain,
five vertical cores (5 cm in length) were taken from the soil
column using a 5-mL syringe (1.4 cm i.d., 1.7 cm. o.d.) from which
the luer end was cut. A core taken from the center of the column
was divided into sections taken from depths of 0.0 to 1.7, 1.7 to
3.4, and 3.4 to 5.0 cm. Cores were also taken at distances of 2 cm
and 4 cm between the core center and the column center. Two cores

5


-------
were taken at each such distance at diametrically opposed points
per column. Tests of lateral movement were conducted with
duplicate columns. The total number of bacteria in each core and
core section was then determined.

Determinations were made of the percentage of the cells added
to the soil that appeared in the effluent (percentage transported)
and the percentage that could be recovered from both the effluent
and the soil column at the end of the test period (percentage
recovery). Triplicate columns were inoculated with 1 mL of
bacterial suspension. The bacteria in the effluent were counted
after passage of four pore volumes of deionized water, and the
number of bacteria remaining in the soil of two of the three
inoculated columns was determined after the effluent was collected.
For this purpose, all of the soil was removed from the column, the
soil was shaken with deionized water for 5 min on a rotary shaker
operating at 12 0 rpm, and counts were made after the larger
particles were allowed to settle for 5 min.

Adsorption of bacteria to soil was determined by a procedure
based on the difference in gravity sedimentation rates of the
bacterial cells and sand, silt, and clay size fractions of the
soil. At the experimental temperature of 3°C, the time required
for the size fraction >2 jL/m of the soil to settle below an 8-mm
depth is predicted by Stokes' law (Jackson, 1974) to be 65 min. To
determine the amount of water-dispersible clay (smaller than 2 jL/m)
that remained suspended with bacteria at the sampling depth, a 2 0-g
portion of Kendaia loam was shaken for 1 h at 3.5±1.5°C with 100 mL
of deionized water. The suspension was then diluted to 1 L with
deionized water and allowed to settle for 4.5 h (the settling time
was calculated to collect particles <2 jL/m diameter) . A 25-mL
portion was then taken from the 5-cm depth of a 1-L graduated
cylinder and dried overnight at 105°C. The water-dispersible clay
constituted 2.6% of the total soil, or 12% of the total clay
content.

To measure bacterial adsorption at 3°C, 2 0-g samples of
Kendaia loam and approximately 10® cells were shaken for 1 h with
100 mL of deionized water. The suspension was allowed to settle
for 65 min in a sterile 250-mL graduated cylinder. Determinations
were made of the number of bacteria at the 8-mm depth of the soil
suspension (S8) and of a control (C8) to which no soil was added.
The soil and control suspensions were again thoroughly mixed, and
the numbers of bacteria in the soil suspension (St) and the control
(Cf) were determined from a depth of approximately 1 to 2 cm. The
drop plate method (Reed and Reed, 1948) was used for bacterial
counts.

The data were expressed as adsorption coefficients (Kd) where
Kd = F/D (Ct - F)"1

6


-------
and F is the number of bacteria adsorbed, and D is the
concentration of soil in the suspension. The value of D is 0.2 (20
g/100 mL). The adsorption coefficients are expressed as the lower
or higher bound of bacterial adsorption. The lower-bound
adsorption value assumes that bacteria retained by clay particles
of equal or smaller size are not considered to be sorbed, and the
higher-bound adsorption value considers all bacteria-clay
interactions as adsorption.

To determine F, the equation takes into account bacterial
settling and the interaction between bacteria and water-dispersible
clay. The lower-bound of adsorption is expressed by the following
equation, which does not consider the interaction between bacteria
and water-dispersible clay in the top 8-mm sampling depth,

Flow —	~ (^8 ^ ^/Cg) *

The higher-bound of adsorption is expressed by the following
equation, which considers the interaction between bacteria and
water-dispersible clay in the top 8-mm sampling depth,

Fhigh = St - (S8 Ct/C8) d - Wd X Ct/C8)"1

where Wd is the water-dispersible fraction of the clay (0.12).

CELL PROPERTIES

All assays described below were conducted at 3°C, which was
the temperature used in tests of bacterial transport.

Hydrophobicity was determined by measuring bacterial adherence
to hydrocarbons (BATH) and hydrophobic-interaction chromatography
(HIC) . For the BATH procedure, a modification of the method of
Rosenberg et al. (1980) was used with octane as the assay
hydrocarbon. Bacteria grown in Trypticase-soy broth for 24 h at
3 0°C were washed and resuspended in a buffer-salts solution at pH
7.1 containing 22.2 g of K2HPO^ -3H20, 7.3 g of KH2P04, 1.8 g of urea,
and 0.2 g of MgS04-7H20 per liter to an optical density of 0.2 at
550 nm. A mixture of 4.0 ml of washed cells (108/ntl) and 0.00,
0.25, 0.5, 0.75, or 1.0 ml of octane was added to test tubes and
mixed with a vortex mixer for 60 sec. The contents of each test
tube were placed in a 5-ml syringe with a luer-lock to avoid
possible mixing of the organic layer with the aqueous phase. The
phases were allowed to separate for 30 min, and then the light
absorbance at 400 nm of the aqueous phase was measured with a
Hitachi U-2 000 spectrophotometer. The percentage of adherence to
octane is expressed as 100 times the ratio of (absorbance of the
octane-free bacterial suspension minus that of the aqueous layer)
to the absorbance of the octane-free suspension.

Hydrophobic-interaction chromatography, which measures the
amount of bacteria retained by a hydrophobic gel (Dillon et al.,

7


-------
1986; Mozes and Rouxhet, 1987), was also used to determine cell
hydrophobicity. Octyl-sepharose gel (Pharmacia, Uppsala, Sweden)
was washed with 4 M NaCl prior to being packed in duplicate wool-
plugged Pasteur pipettes to a final bed volume of 2 ml. Bacteria
grown in Trypticase-soy broth for 24 h at 3 0°C were washed twice
with 4 M NaCl and resuspended in 4 M NaCl to a density of
approximately 108 cells per ml. The bacterial suspension (1 ml)
was then applied to the gel bed and eluted with 5 ml of 4 M NaCl.
Another column was eluted with deionized water to assess the extent
of filtration of the cells. The percentage of hydrophobicity is
expressed as 100 times the ratio of absorbance of the original cell
suspension at 400 nm minus that of the eluate (nonretained
bacteria) to the absorbency of the original suspension. The data
indicate the percentage of cells retained by the hydrophobic gel.

Net surface electrostatic charge was determined by
electrostatic-interaction chromatography (ESIC) and by measuring
zeta potential. ESIC is a measure of the bacterial affinity for
cation- or anion-exchange resins (Pedersen, 1981; Stenstrom, 1989).
Pasteur pipettes plugged with glass wool were filled with either
CM-sepharose CL-6B anion exchanger or DEAE-sepharose CL-6B cation
exchanger (Pharmacia) to a final bed volume of 2 ml. Duplicate
columns of the two resins were washed with 0.9% NaCl prior to
column packing. Bacteria grown in Trypticase-soy broth for 24 h at
30°C were washed twice with 0.9% NaCl and resuspended in 0.9% to a
density of 108 cells/ml. The bacterial suspension (1.0 ml) was
added to the columns, and the cells were eluted with 5 ml of 0.9%
NaCl. The percentage of cells retained is expressed in the same
way as for the BATH procedure.

The electrophoretic mobility of the bacteria was measured with
a Lazer Zee meter model 501 (Mozes et al., 1987; van der Mei et
al., 1987). This instrument measures the zeta potential by
determining the rate of bacterial movement in a known electric
field. The potential difference between the electrodes was 150 V.
The cells were suspended in deionized water (pH 7) at a
concentration of 10® cells per ml before determination of
electrophoretic mobility at 20°C. The measurements of zeta
potential were corrected to 3°C by multiplying the measured zeta
potential by (1-0.02(20-3).

TRANSPORT THROUGH AQUIFER SOLIDS

The bacterial strain used in the experiments was isolated by
Gannon et al. (1991) from samples of Kendaia loam obtained in
Aurora, NY. This strain (Pseudomonas sp. strain KL2) is
amphitrichously flagellated and is characterized by an average cell
diameter of 0.4 /im and an average cell length of 1.8 /im. From the
19 strains tested by Gannon et al. (1991), this strain was selected
because it showed higher than average mobility in homogeneous
columns of disturbed Kendaia loam soil. The bacteria were grown at
30°C for 24 h (200 mL of medium in a 250-mL flask) in Trypticase-

8


-------
soy broth on a rotary shaker (100 rpm). After being collected by
centrifugation and washed twice with deionized water with a
resistivity of 18.3 Mohms-cm, the bacteria were resuspended in
either deionized water or a 0.01 M NaCl solution to achieve
concentrations of 1.0 X 10s and 1.0 X 109 cells per mL. The pH of
these bacterial suspensions was 7.0. Ten minutes prior to the
breakthrough experiments, the extent of cell aggregation was
assessed by examining wet mounts under a light microscope and by
comparing differences in turbidity of the bacterial suspension at
0 and 5 min using a Spectronic 20 spectrophotometer at a wavelength
of 550 nm, following procedures described by Calleja (1984) . No
bacterial aggregation was observed in either deionized water or
NaCl solution.

The sandy aquifer material used originated from Eastport, NY.
The Suffolk County Department of Health provided auger drillers to
obtain aquifer material from a depth of 11 m. Special precautions
were taken not to mix the aquifer material with surface soil.
After being air-dried and passed through a 2-mm sieve, this aquifer
material was found to contain 0.07% clay, 1.88% silt, 4.55% fine
sand, 3.1% medium sand and 90.4% coarse sand.

Breakthrough experiments were conducted using 0.3-m long
plexiglass columns (Soil Measurement Systems, Tucson, AZ) with an
internal diameter of 0.05 m. Bottles containing 0.01 M NaCl,
deionized water, or the bacterial suspension were connected to the
column via a peristaltic pump used to maintain a constant flow
velocity. The sand, which was constantly stirred to avoid particle
segregation while it was poured into the column, was packed
uniformly to a bulk density of 1.750±0.012 Mg M"3 by gently tapping
the wall of the columns. One pore volume (PV) was equal to 204
cm3. All the tubing, glassware and valves used in the experiments
were autoclaved to avoid contamination. After being sterilized by
irradiation with 60Co (2.5 Mrad) , the columns were set up vertically
and saturated with sterile deionized water. Upon saturation, a
constant upward flow velocity was established using a peristaltic
pump. Two velocities were selected: 1.0 X 10"4 m s"1 (3.5 PV h"1)
and 2.0 X 10"4 m s"1 (6.9 PV h"1) .

Two types of boundary conditions could in principle be adopted
in column breakthrough experiments: (a) a step increase in
concentration or (b) a pulse-type boundary condition. The latter
was chosen, largely to limit the risk of clogging at the inlet end
of the column. To obtain a pulse-type boundary condition, the
bacterial suspensions were substituted for a period of 1 h for the
sterile deionized water that previously percolated through the
column. For the experiments in which bacteria were suspended in a
NaCl solution, the bacterial pulse was followed by a sterile 0.01
M NaCl solution. Otherwise, the flow of deionized water was
resumed. In an experiment designed to determine the effect of a
sudden reduction in the ionic strength of the solution, a pulse of
bacteria suspended in the NaCl solution was followed by 0.01 M NaCl

9


-------
solution for 1 h and then by deionized water for 2 h. In each
experiment, the effluent from the column was collected using a
fraction collector. The bacterial concentration in each fraction
was determined by the drop plate method (Reed and Reed, 1948) on
Trypticase-soy agar using triplicate samples for each dilution.
All the breakthrough experiments were carried out in duplicate at
3±1.5°C to limit bacterial growth and death.

Numerical integration of the untransformed data, using the
trapezoidal method, yielded the percentage of bacteria recovered in
the effluent after a given period of time.

The breakthrough of a tracer (chloride) was determined
separately under the same conditions by applying a 1-h pulse of
0.01 M NaCl followed by deionized water. The chloride
concentration in successive fractions of the effluent was measured
with a chloridometer (Haake Buchler Instruments, Saddlebrook, NJ) .
In all cases, perfectly symmetric breakthrough curves showed that
chloride behaved as a tracer and that preferential pathways were
absent from the column.

STATISTICAL ANALYSIS OF DATA

The data were subjected to regression analysis using the Data
Desk software package (Odestra Corp., Northbrook, IL) .
Significance was determined by the F-ratio, which is the ratio of
the mean square for regression to the mean square residual.

RESULTS

TRANSPORT THROUGH SOIL

Bacteria were obtained from Kendaia loam that were able to
grow using benzene, chlorobenzene, or toluene as sole source of C
and energy. They were identified by standard bacteriological
techniques as strains of Pseudomonas. Achromobacter. Bacillus, and
Enterobacter. The isolates able to degrade benzene were designated
Pseudomonas Benl and Achromobacter Ben2, those degrading
chlorobenzene were designated Pseudomonas CB1, Bacillus CB2, and
Bacillus CB3, and the bacteria growing on toluene were designated
Enterobacter Toll, Enterobacter Tol2, Enterobacter Tol3, and
Pseudomonas Tol4. The bacteria obtained from diesel fuel-
contaminated soil were designated Achromobacter DF1, Pseudomonas
DF2, and Flavobacterium DF3, those from the sewage treatment plant
were designated Enterobacter strains IS1 and IS2, and those from
Kendaia loam as Enterobacter KL1 and Pseudomonas KL2 and KL3.

Pseudomonas KL2 was used to study the direction of movement of
bacteria through the soil. The distribution of bacteria in the
column showed that the direction of bacterial flow was chiefly
downwards in both soil columns tested (TABLE 1) . The presaturation
of the soil columns resulted in vertical water flow,- and bacteria

10


-------
TABLE 1. Distribution of Pseudomonas KL2 in columns of soil after
passage of four pore volumes of water

Sample	Sample	Cell no. X 10^ sen g

Location	depth (cm)	Column 1	Column 2

Center

0.0-1.7

980

2,200



1.7-3.4

1,400

730



3.4-5.0

720

690

Left, 0.4-2.0 cmf

0-5.0

79

21

Right, 0.4-2.0 cmt

0-5.0

14

0.75

Left, 2.4-4.0 cmf

0-5.0

0.75

0.01

Right, 2.4-4.0 cmt

0-5.0

0.90

<0.01

tLateral distance and left or right of center core.

11


-------
thus moved with the water. Because little horizontal movement was
evident and essentially no bacteria reached the walls, movement of
cells along the column wall did not appear to contribute to
measurements of the vertical mobility of bacteria.

In the studies of bacterial transport, the agreement between
the numbers of colony-forming units on replicate plates from
triplicate columns of soil was good (TABLE 2), except for Bacillus
CU4519 and Bacillus CB3, for which fewer than 105 of the 10s cells
added to the top of the soil were transported and the agreement
among replicate counts was poor. In tests of the numbers of added
bacteria that were recovered, replicate counts of the number of
colony-forming units agreed with a standard deviation of ±2 0%.

Marked differences in the fraction of the added bacteria that
moved through the soil were evident among the various species
(TABLE 2). For Enterobacter IS2, 15% of the cells were
transported, whereas only 0.01% of the added cells of Bacillus CB3
were found in the effluent as viable bacteria. No consistent
pattern in mobility was evident among the strains of a genus; thus,
more than 5% of the cells of two strains of Enterobacter and three
strains of Pseudomonas were transported as compared to <1% of three
other strains of the same two genera. The benzene degraders
(strains Benl and Ben2) moved to a greater extent than the isolates
able to use chlorobenzene (strains CB1, CB2, and CB3) or toluene
(Toll, Tol2, Tol3, and Tol4). Of the eight isolates selected for
their presumed greater mobility, more than 2% of the cells of six
strains (Enterobacter IS2, Enterobacter IS1, Pseudomonas KL2,
Pseudomonas DF2, Achromobacter DF1, and Enterobacter KL1) were
transported through the soil, whereas the percentages were lower
for Pseudomonas KL3 and Flavobacterium DF3. For the bacterium
originally obtained from a groundwater aquifer (Arthrobacter Lula
D), 7.7% of the cells moved through the soil.

The percentage of the added cells that were recovered in the
soil and the effluent varied from essentially 100% for
Achromobacter Ben2 and Bacillus CB2 to 4.3% for Bacillus CB3 (TABLE
2). Although some strains of a genus showed high recoveries, far
lower recoveries were found among other strains of the same genus.
The fact that the recoveries of only 5 of the 19 strains were below
25% indicates that high percentages of the cells of many strains
did not lose viability and were not irreversibly sorbed to soil
particles. A comparison of recovery percentages and the transport
percentages indicates that many of the viable cells were retained
in the soil column. The data show that the recoveries exceeded 50%
for 7 of the 10 isolates for which >2% of the cells were
transported, but such high recoveries were found for only 1 of the
9 isolates for which <1% of the cells were transported. Regression
analysis indicated that transport was correlated with recovery (F
ratio = 5.08, significant at P = 0.05).

12


-------
TABLE 2. The numbers and percentages of bacteria transported and
recovered

Bacterium

Bacteria
transportedf
(%)

Bacteria
recoveredf
(%)

Enterobacter IS2

15±2

61±6

Enterobacter IS1

13±3

60±13

Pseudomonas KL2

8 . 2±1.2

4 6±6

Arthrobacter Lula D

7.7±1.9

39±4

Pseudomonas DF2

6.9±0.4

53±12

Achromobacter Ben2

6.8±0.8

107±13

Pseudomonas Benl

5.9±0.4

71±4

Bacillus CB2

4.1±1.2

104±20

Achromobacter DF1

3.9±0.8

39±4

Enterobacter KL1

2.2±0.4

86±7

Pseudomonas KL3

0.9±0.3

48±3

Enterobacter Tol2

0.9±0.1

14±1

Enterobacter Tol3

0.9±0.1

64±1

Pseudomonas Tol4

0 . 3±0.1

6.5±1.5

Enterobacter Toll

0.2±0.1

34±1

Pseudomonas CB1

0.2±0.1

25±5

Bacillus CU4519

0.1±0.1

9 . 4±1.3

Flavobacterium DF3

0.1±0.01

5 . 4±1.5

Bacillus DB3

0.01±0.00

4 . 3±0.4

fMean ± standard deviation

13


-------
The values for adsorption coefficient (Kd) of the bacteria
ranged from 0.0 (i.e., a lower-bound Kd = no bacteria adsorbed) for
Enterobacter IS1 to infinity (i.e., a higher bound Kd = all
bacterial were adsorbed) for Pseudomonas Tol4 (TABLE 3) . A
comparison of the mean values of the lower- and higher-bound Kd
values with the transport percentages shows that 8 of 10 bacteria
for which >2% of the cells were transported had Kd values <10.0,
whereas 8 of 9 bacteria for which <1% of the cells were transported
had Kd values >10.0. Thus, a high percentage of cells of strains
with low Kd values moved relatively freely through the soil,
whereas a low percentage of cells of strains with high Kd values
were transported at a significant rate under identical conditions.

The relationship between the percentage of the cells
transported and the lower- but not the upper-bound Kd values was
statistically significant. The F values for the regressions were
7.40 and 3.26, respectively.

Regressions indicated that recoveries were related to the Kd
values (P = 0.05). When the lower-bound Kd values exceeded 20 or
the upper-bound Kd values exceeded 100 (Enterobacter Tol2,
Pseudomonas Tol4 and CB1, and Bacillus CU4519) , 25% or fewer of the
cells were recovered. This probably reflects the inability to
detect a significant fraction of the sorbed cells by the counting
procedure.

CELL PROPERTIES

The hydrophobicities of the bacteria varied markedly.
Measurements of hydrophobicity based on bacterial adherence to
octane (the BATH method) revealed that less than 10% of the cells
of Enterobacter sp. IS2, Pseudomonas sp. KL 2 and Bacillus strains
CU4 519 and CB3 adhered to the hydrocarbon, whereas more than 70% of
the cells of Bacillus sp. CB2 and Pseudomonas sp. Tol 4 were
retained (TABLE 4) . No relationship is evident between the results
of the BATH assay and the percentage of cells transported or the
classification of the isolates.

For most of the test bacteria, more cells were retained by the
hydropic gel in the determinations by hydrophobic-interaction
chromatography (HIC) than by octane. Except for Achromobacter sp.
Ben2, more than 25% of the cells adhered to the gel (TABLE 4). No
relation is evident between the results of the HIC assay and the

14


-------
TABLE 3. Adsorption coefficients of bacteria

Bacterium

Adsorution
Lower
bound

coefficient (Kd)
Higher
bound

Enterobacter IS2

4.5

5.5

Enterobacter IS1

0.0

0.0

Pseudomonas KL2

5.5

8 . 5

Arthrobacter Lula D

5.5

8.0

Pseudomonas DF2

0.85

0.9

Achromobacter Ben2

13

25

Pseudomonas Benl

6.0

8.5

Bacillus CB2

18

36

Achromobacter DFl

7.5

11

Enterobacter KL1

3.5

4.5

Pseudomonas KL3

9.0

24

Enterobacter Tol2

29

145

Enterobacter Tol3

6.5

8.5

Pseudomonas Tol4

45

00

Enterobacter Toll

9.5

16

Pseudomonas CB1

26

150

Bacillus CU4519

30

410

Flavobacterium DF3

13

17

Bacillus CB3

12

24

15


-------
TABLE 4. Relationship between characteristics of bacteria
and their transport through soil



Cells

















trans-



Cells

retained

m

Zeta

Cell





ported

BATH

HIC

Anion

Cation

potential

length

Cap

Bacteria

(%)

assay

assay

exchange

exchange

(-mV)

(f/m)

sul<

Enterobacter IS2

15

8

29

98

14

24

0.7

+

Enterobacter IS1

13

21

29

98

29

24

0.7

+

Pseudomonas KL2

8.2

6

28

97

9

21

1.4

+

Arthrobacter Lula D

7.7

13

91

98

7

31

0.8

-

Pseudomonas DF2

6.9

17

29

99

5

19

0.8

+

Achromobacter Ben2

6.8

31

7

98

23

18

1.4

-

Pseudomonas Benl

5.9

25

45

100

5

26

2.1

-

Bacillus CB2

4.1

85

58

98

21

8

0.9

+

Achromobacter DR1

3.9

61

66

98

10

13

1.9

+

Enterobacter KL1

2.2

17

28

99

25

21

1.8

-

Pseudomonas KL3

0.9

21

32

98

7

32

1.8

-

Enterobacter Tol2

0.9

19

39

99

47

21

2.6

-

Enterobacter Tol3

0.9

11

41

98

15

19

2.2

-

Pseudomonas Tol4

0.3

75

50

96

10

21

1.7

+

Enterobacter Tolll

0.2

69

50

99

9

11

1.0

-

Pseudomonas CB1

0.2

44

64

97

18

21

2.5

-

Bacillus CU4519

0.1

5

77

100

20

36

1.6

-

Flavobacterium DF3

0.1

29

59

98

39

13

1.3

-

Bacillus CB3

0. 01

5

61

98

7

15

2.1

+


-------
percentage of cells transported or the genera in which the bacteria
are classified. Moreover, as indicated by the F value, regression
analysis showed that the correlation between the results of the
BATH and HIC assays was poor. Elution of the columns with
deionized water rather than 4 M NaCl led to recovery of 100% of the
added cells of all strains tested, indicating that the cells were
not removed in the assay by filtration.

Measurements were made of the net surface charge by
determinations involving ESIC and zeta potentials. In the
chromatographic procedure, anion- and cation-exchange resins were
used. Most cells of the 19 strains were retained by the anion-
exchange resins, and 97 to 100% of the cells were retained by this
method (TABLE 4). On the other hand, appreciable differences were
noted in the percentage of cells retained by the cation-exchange
resins. More than 30% of the cells of Enterobacter sp. Tol2 and
Flavobacterium sp. DF3 but less than 10% of the cells of
Pseudomonas strains KL2, DF2, Benl, and KL3, Arthrobacter sp. Lula
D, Enterobacter Tol4, and Bacillus sp. CB3 were retained. The zeta
potentials of the bacteria also differed, ranging from values of -8
mV for Bacillus sp. CB2 to -36 mV for Bacillus sp. CU4519. A
relation was not evident between the surface charge of the 19
bacteria determined by these means and their mobilities through
soil or generic placement.

The cell sizes and the presence of capsules were determined
for each of the 19 organisms. The width of the cells did not
differ appreciably, ranging from 0.5 to 0.7 jJm. In contrast, the
lengths varied from 0.7 to 2.6 fJm (TABLE 4). Transport of the
bacteria was statistically related to length of cells (F = 11.8,
significant at P = 0.05) as indicated by regression analysis.

Eight of the strains produced capsules and ten formed none.
However, nonencapsulated and capsulated cells were evident among
the strains showing both good (>1% transport) and poor mobilities
(<1% transport).

Only 10 of the 19 bacteria were tested for the presence of
flagella. Eight strains were flagellated. Enterobacter strains
IS2 and IS1, Pseudomonas strains DF2 and Benl, and Bacillus sp.
CU4 519 had peritrichous flagella, Pseudomonas sp. KL2 and Bacillus
strains CB2, CU4519, and CB3 bore lophotrichous flagella
Pseudomonas sp. KL2 and Bacillus sp. CB2 were amphitrichous, and
Pseudomonas sp. DF2 and Bacillus strains CU4519 and CB3 were
monotrichous. Flagella were thus present on cells of strains that
exhibited poor (<1% transport) and good (>1% transport) mobilities.

TRANSPORT THROUGH AQUIFER SOLIDS

In the experiments involving low flow rate (1.0 X 10~Sii s"1) and
low bacterial concentration (1.0 X 108 cells mL"1), a distinct
effect of the ionic strength of the carrying liquid phase on the

17


-------
extent of bacterial breakthrough was evident (Fig. 2). When the
bacteria were suspended in 0.01 M NaCl solution, their
concentration (C) in the effluent never reached more than 2.2% of
the input concentration, (Co). In contrast, the relative
concentration C/Co approached unity after approximately 70 min when
the bacteria were suspended in deionized water. After 2 h of flow
through the columns, 100% of the chloride tracer and 60% of the
bacteria suspended in deionized water were transported through the
column vs. only 1.5% of the cells suspended in 0.01 M NaCl
solution.

Compared with the breakthrough of chloride, the breakthrough
of bacteria suspended in deionized water was significantly retarded
(Fig. 3). The time for first measurable appearance of cells in the
effluent was not greatly different in both cases; i.e.,
approximately 12 min for chloride and 15 min for bacteria.
However, whereas the relative concentration of chloride in the
effluent increased rapidly thereafter, reaching a value of 1.0 only
30 min after application of the pulse, the ratio C/Co for the
bacteria increased more gradually. The tail ends of the
breakthrough curves for chloride and bacteria suspended in
deionized water appeared to coincide, a sharp drop occurring in
both cases at about 75 min. A similar drop in C/Co at
approximately 7 5 min was also apparent for bacteria suspended in
0.01 M NaCl solution. However, it is necessary to bear in mind the
experimental error involved in. the evaluation of the bacterial
concentration as illustrated by the scatter. The bacterial
concentration in the effluent after 100 min became negligibly small
both in the systems with deionized water and 0.01 M NaCl solution.

In the experiments involving a higher flow rate (2.0 X 10"4m
s"1) , the breakthrough was similar to that described above with a
few differences. When bacteria were suspended in 0.01 M NaCl
solution, more cells (3.9% vs. 1.5%) were transported through the
column, and their relative concentration reached 7.3% (vs. 2.2%
observed earlier) (Fig. 4). Nevertheless, reduction of the ionic
strength of the liquid phase resulted in a significantly increased
rate of transport (77%) and in a peak relative concentration
reaching 1.0. The higher flow rate produced a shortened time to
the breakthrough of chloride; after only 5 min, chloride appeared
in the effluent, and its relative concentration then increased
sharply to reach 1.0 in less than 10 min. Essentially all (99%) of
the chloride passed through the column. The bacteria suspended in
deionized water followed the same pattern (Fig. 5) with, however,
less apparent delay than in the first experiment. The time for
bacterial breakthrough was between 5 and 10 min, and the relative
cell concentration reached 1.0 approximately 30 min after
application of the pulse. Thereafter, the behavior of the bacteria
suspended in deionized water coincided with that of chloride. In
particular, the tail ends of their breakthrough curves were
congruent, precisely the same drop in C/Co being observed in both
case between 65 and 70 min. A similar decrease of C/C0 was

18


-------
0.080

0.060

0.040

0.020

Hours

Figure 2. Breakthrough of Pseudomonas sp. KL2 with 0.01 M NaCl
(flow rate 1.0 X 10"4 in s"1, equivalent to 3.5
pore volume h"1).

19


-------
1.0
0.8
0.6
0.4
0.2

Deionized

M.

°o° cPO O 000 o cp°^o°1^

o w o

• •

• •



3 •

• • %.

• • •

0.5

1.0
Hours



o chloride
• bacteria

ftabt I t»t •

1.5

Figure 3. Breakthrough of Pseudomonas sp. KL2 with deionized water
(flow rate 1.0 X 10"4 m s"1, equivalent to 3.5 pore
volumes h"1) .

20


-------
1

CJ
CJ

B

•H

4->

cd
U
4->

C


•H

Pi

0.080

0.060

0.040

0.020

0.01 M NaCl

• •

_L

* • • f

0.5

1.0

Hours

1.5

Figure 4. Breakthrough of Pseudomonas sp. KL2 with 0.01 M NaCl

solution and flow rate of 2.0 X 10"4 m s"1, equivalent to
6.9 pore volumes h"1.

21


-------
1

u
u

8

•H

cd

4-i
&
03
CJ

S

u

>
•H

4J

CO

'—I

Si

Pd

1.0

0.8

0.6

0.4

0.2

peionized H^O

O o° o " o_

^ % CP

o chloride
• bacteria

0.5

1.0
Hours

1.5

Figure 5. Breakthrough of Pseudomonas sp. KL2 with deionized water
at a flow rate of 2.0 X 10"4 m s"1, equivalent to 6.9 pore
volumes h"1.

22


-------
observed within the same time period when the bacteria were
suspended in 0.01 M NaCl solution.

When a cell density of 1.0 X 109 cells mL"1 was used, the
breakthrough curves for bacteria in 0.01 M NaCl solution and
deionized water were almost identical except that the relative
concentration of bacteria at the peak approached 1.0 at the low
ionic strength, whereas C/CQ was 0.7 at the higher ionic strength
(Fig. 6 and 7) . A precipitous drop in relative concentration
occurred in both cases between 70 and 8 0 min. This time range
corresponded closely to the tail ends of the chloride and bacterial
breakthrough curves that were previously noted under similar flow
conditions. The percentages of bacteria transported through the
column were 44% with 0.1 M NaCl and 57% with deionized water.

The effect of reducing the ionic strength (replacing 0.01 M
NaCl solution with a pulse of deionized water) of the inflowing
solution on the breakthrough of bacteria is shown in Fig. 8. A
control was conducted in which 0.01 M NaCl was used for the entire
test period. When only NaCl solution was used, a single peak
(0.022) of bacterial breakthrough followed by a tail was observed.
The continuous application of NaCl solution did not displace
bacteria already retained in the column. However, when the NaCl
solution was replaced with deionized water, a second peak of
bacterial breakthrough was observed. The relative bacterial
concentration at the second peak reached 0.12, which was 4 times
higher than the first peak. The decrease in ionic strength lowered
the retention of bacteria in the column.

DISCUSSION

Macropore flow may be a major mechanism of bacterial transport
in soils. Therefore, the use of undisturbed soil columns might
have provided data on bacterial transport that would have
particular relevance to circumstances prevalent in the field (Smith
et al., 1985). However, columns of homogeneous soil were used to
avoid uncontrolled, preferential movement of bacteria through
macropores and thus to permit a definition of the factors that
control the movement of bacteria through the soil matrix itself.
For the development of the procedure for determinations of
mobility, a loamy soil was selected to avoid the extremes of
limited bacterial sorption and nearly free movement in sandy soils
on the one hand and the restricted movement resulting from
mechanical filtration and increased sorption in fine structured
soils on the other hand. Had these more extreme conditions been
imposed, the ability of the procedure to detect small differences
in bacterial mobilities might not have been evaluated.

Saturated soil with a constant head of water was used to mimic
bacterial movement under conditions of saturated flow. This
permitted the occurrence of mass transport of the cells in the
sufficiently large pores of the homogeneous soil. Bitton et al.

23


-------
CJ>
CJ>

c
o

•H

4-1

tfl

5-1

4-1

G
0)
U
C
o

CJ

0)
>
•H
4-1

to

r-H

O)
ctj

Hours

Figure 6. Breakthrough of Pseudomonas sp. KL2 with 0.01 M NaCl at

1.0 X 10V cells mL"1 (flow rate 1.0 X 10'* m s"1 equivalent

to 3.5 pore volumes h"1).

24


-------
Figure 7. Breakthrough of Pseudomonas sp. KL2 with deionized water

at 1.0 X 10v cells mL"1 (flow rate
equivalent to 3.5 pore volumes h"1) .

1.0 X 10"

m s-1

25


-------
0.12

0.09

0.06

0.03

B ~

ei . *B

nfel	1—«-

2	3

Hours

M

A

Figure 8. Breakthrough of Pseudomonas sp. KL2 at 1.0 X 109 cells
mL~1 using 0.01 M NaCl for 1 h followed by deionized
water (flow rate 1.0 X 10"4 m s"1 equivalent to 3.5 pore
volumes h"1) .

26


-------
(1974) showed that movement of bacteria through soil columns
stopped when the water content was at or below field capacity, and
Madsen and Alexander (1982) demonstrated that movement of bacteria
through soil was not detectable in the absence of a transporting
agent such as water.

The procedure described here has several advantages for
testing bacterial mobility in the soil matrix. Spurious data on
mobility resulting from bacteria moving at the interface between
the soil and the column wall are avoided through the use of a
relatively wide column and the coating of the column walls with
petrolatum to bind soil particles to the walls. Channels through
which bacteria could move preferentially were eliminated by
grinding and sieving the soil prior to preparing the column and, to
ensure reproducible measurements of transport, by uniform packing
to a fixed bulk density. Bacterial death from predation or
parasitism was avoided because such predators and parasites were
killed by irradiating the soil. Increases in cell numbers arising
from growth and decreases associated with starvation were prevented
by performing the tests of transport at 2 to 5°C. Furthermore, the
marked differences in mobility among the isolates suggest that the
proposed procedure does indeed distinguish among bacteria with
different capacities for movement.

For inoculation of the soil surface with bacteria that can
degrade organic pollutants at some underground site, some of the
added cells must move through the soil to the below ground zone of
contamination. Evidence exists, however, that introduced organisms
may fail because they are not transported to sites containing the
chemical (Goldstein et al., 1985). In this context, it is worth
noting that many of the carefully controlled experiments in which
inoculation resulted in biodegradation required transport to soil
depths of only 10 cm (Barles et al., 1979; Edgehill and Finn, 1983;
McClure, 1972). Measurements such as those described in the
present study will enable extrapolation of the potential
penetrability of the bacteria to considerably greater depths.

The capacity to degrade a chemical in culture is a necessary
but not sufficient requisite for successful biodegradation in the
field because, in addition to other traits, the inoculum strain
must possess the traits that enable it to move through soil,
subsoil, or aquifer materials to reach the area of chemical
contamination. The ability of bacteria to be transported to
subsurface contaminated zones may be evaluated from bacterial
properties such as cell size and their susceptibility to
adsorption. The correlation between these properties and mobility
can be determined by the procedure here proposed. Such a
correlation may enable the classification of bacteria according to
their potential mobility. The susceptibility of a bacterium to
predation or parasitism in the matrix through which it must pass
must also be determined before its ability to reach the zone of
contamination can be predicted.

27


-------
The data show a significant inverse correlation between the
lower-bound Kd values and the fraction of cells transported through
the column. The values for recovered cells transported varied from
nearly zero to 25% at Kdl <2, whereas the values never exceeded 8%
at Kdl >2. Adsorption, when sufficiently strong, can effectively
retard the bacteria. If adsorption is weak, the transport of cells
through the soil matrix may be controlled by mechanical filtration.
The correlation between Kdl and Kdh was good enough to make their
use in many cases interchangeable.

The dimensions of the bacterial cells affected their mobility.
As the cell length increased, the highest value for recovered cells
transported among bacteria of a given length decreased. Thus, the
longer the cell, the less its likelihood of passing through the
soil matrix. Cell length was not significantly correlated with the
recovery, Kdh, or Kdl.

Cell hydrophobicity, net surface electric charge, and the
presence of capsular polysaccharides were evaluated because they
are properties of bacterial cells that appear to be involved in
adsorption of bacteria to solid surfaces (Fattom and Shilo, 1984;
Rosenberg and Kjelleberg, 1986). The sizes of the cells
(Corapcioglu and Haridas, 1984; Pekdeger and Matthess, 1983) were
tested because larger cells may be more readily removed by
filtration than smaller cells. The presence of flagella might also
impede movement, so their occurrence on the test strains was
investigated. Varying degrees of hydrophobicity were observed
among the bacteria, but the results of the BATH and HIC assays did
not agree. The nonuniformity of the bacterial surface may cause a
bacterium to be hydrophilic in one assay and hydrophobic in another
(Stenstrom, 1989) . Moreover, hydrophobicity is not a definitive
characteristic but varies with the hydrocarbon used for the assay
(Marshall and Cruickshank, 1973; Stotzky, 1985).

Measurements of zeta potential revealed that all of the test
bacteria had net negative surface charges. However, bacteria with
positive charges on nonuniform cell surfaces may adhere to
negatively charged particles of soil (Marshall, 197 6). No pattern
was observed between the extent of transport and bacterial surface
charges.

Adherence of bacteria to solid surfaces may result from their
having extracellular polysaccharides (Costerton et al. , 1978 ;
Sutherland, 1983). Of the strains for which more than 1.0% of the
cells passed through the soil, six had capsules, indicating that
capsules do not necessarily hinder transport. Measurements were
made of cell size because large bacteria presumably are less likely
to pass through soil pores than small cells, and a statistical
relationship between size and transport of bacteria through soil
was evident. Although cell appendages could impede mobility, a
relationship was not observed between flagellation and transport.

28


-------
Motility was not considered important in the present study because
transport was tested at low temperatures.

The present findings suggest that it should be possible to
obtain bacteria that have both the capacity to biodegrade unwanted
organic compounds and the ability to move through earth materials
to sites containing these chemicals, as indicated by the
observation that high percentages of two benzene degraders and one
chlorobenzene-utilizing bacterium moved through soil in appreciable
numbers.

The movement of bacteria through homogeneous sand increased
when the inflowing solution had a low ionic strength. When
deionized water was used, the bacteria moved readily through the
sand column after an initial period of retention. The high
mobility of bacteria may result from the relatively inert surfaces
of the aquifer materials at low ionic strength, thus limiting
adsorption, and large pores and pore necks which allowed bacteria
to pass through with minimal filtration. The breakthrough of
bacteria and chloride differed in that bacterial breakthrough was
slower, transport of bacteria was less than chloride, and the
relative concentration of bacteria increased gradually and tailed
slightly. These differences indicated that the cells were briefly
retained in the porous medium.

The retention of bacteria is generally attributed to
filtration in the contact zones of adjacent pores, sedimentation in
the pores and/or adsorption (Corapcioglu and Haridas, 1984) . Gerba
et al. (197 5) noted that removal of bacteria appears to occur
largely at or near the soil surface. Filtration at constrictions
in drainage channels and in nontransmissing pores can limit
bacterial movement through porous media (Vinten and Nye, 1985) .
Bacterial adsorption may result from van der Waals forces,
electrostatic interactions, hydrophobicity and polymer bridging
(Marshall et al. , 1971). If the aquifer material has low
adsorption capacity, the potential adsorption sites will be filled
by the bacteria that arrive first. The bacteria that follow can
either travel freely or replace those cells that are adsorbed.
Thus, bacterial breakthrough would be retarded only until the
adsorption sites were filled. In addition, the adsorption of
bacteria may reduce the diameter of pore necks and enhance
filtration to some extent. The salt solution dramatically reduced
bacterial breakthrough. The total bacterial breakthrough was low
because the cells were removed from solution by interactions with
the sand matrix. Cations such as Ca, Mg, and Na increase bacterial
adsorption by neutralizing the negative surface charge of the
bacteria (Yates and Yates, 1988). The enhanced retardation of
bacterial transport in the presence of NaCl may result from a
further compression of the electrical double-layer arising from the
increase in ionic strength. The diffuse double-layer may be more
compressed and the interparticle distances reduced, thus creating
more opportunities for adsorption (Marshall, 1975). When deionized

29


-------
water was used, nearly all the bacterial cells moved through the
system after the initial stage of retention, suggesting that the
pores and pore necks were not restricting bacterial passage. It is
unlikely that NaCl increased the number of terminal pores or
changed the pore size distribution of the aquifer material because
of the low clay content of the aquifer material.

The replacement of the NaCl solution with deionized water
resulted in a second peak of bacterial breakthrough. The retention
caused by the adsorption mechanisms appeared to be reversible when
ionic strength was changed. Filtration would not likely be
reversible unless the porous medium was disturbed.

Differences in the transport of bacteria noted in experiments
with 0.01 M NaCl and deionized water were more marked at 1.0 X 108
than at 1.0 X 109 cells per mL. This may be linked with available
filtration and adsorption sites. At low cell concentrations, the
retention sites presumably were not filled so that most of the
bacteria were adsorbed, whereas most of the retention sites
presumably were filled at a higher cell concentration, so that the
additional bacteria traveled freely or replaced bacteria that were
retained. However, tests with deionized water failed to reveal
appreciable differences between the two cell concentrations,
possibly a result of the reduction in ionic strength that decreased
the adsorption capacity of the sand. An increase in flow velocity
somewhat reduced bacterial retention, a decrease that may be
related to the reduction in bacterial residence or "reaction" time
at higher flow velocity.

The experimental results suggest that adsorption significantly
contributed to the retention of bacteria and that bacterial
movement through aquifer sand was enhanced by reducing the ionic
strength of the inflowing solution. Cell density and flow velocity
also influenced bacterial movement. For bioremediation of
contaminated sandy aquifers, manipulation of ionic strength may
thus be a means to facilitate movement of bacteria to the site of
organic contamination.

30


-------
REFERENCES

Barles, R. W. , C. G. Daughton, and D. P. H. Hsieh. 1979.
Accelerated parathion degradation in soil inoculated with
acclimated bacteria under field conditions. Arch. Environ.
Contain. Toxicol. 8:647-660.

Baveye, P., and A. Valocchi. 1989. An evaluation of mathematical
models of the transport of biologically reacting solutes in
saturated soils and aquifers. Water Resour. Res. 25:1413-
1421.

Bell, R. G., and J. B. Bole. 1978. Elimination of fecal coliform
bacteria from soil irrigated with municipal sewage lagoon
effluent. J. Environ. Qual. 7:193-196.

Bitton, G., J. M. Davidson, and S. R. Farrah. 1979. On the value
of soil columns for assessing the transport pattern of viruses
through soils: a critical outlook. Water Air Soil Pollut.
12:449-457.

Bitton, G. and C. P. Gerba. 1984. Groundwater pollution
microbiology: The emerging issue. p. 1-7. In G. Bitton and
C. P. Gerba (eds) Goundwater pollution microbiology. J. Wiley,
New York.

Bitton, G., N. Lahav, and Y. Henis. 1974. Movement and retention
of Klebsiella aeroaenes in soil columns. Plant Soil 40:373-
380.

Borden, R. C. , and P. B. Bedient. 1986. Transport of dissolved
hydrocarbons influenced by oxygen-limited biodegradation. 1.
Theoretical development. Water Resour. Res. 22:197 3-1982.

Brown, K. W., H. W. Wolf, K. C. Donnelly, and J. F. Slowey. 1979.
The movement of fecal coliforms and coliphages below septic
lines. J. Environ. Qual. 8:121-125.

Calleja, G. C. 1984. Identifying the problems: Theoretical and
experimental approaches, p. 21-91. In Microbial aggregation.
CRC Press, Boca Raton, Florida.

Corapcioglu, M. Y., and A. Haridas. 1984. Transport and fate of
microorganisms in porous media: a theoretical investigation.
J. Hydrol. 72:149-169.

Corapcioglu, M. Y., and A. Haridas. 1985. Microbial transport in
soils and groundwater. A numerical model. Adv. Water Resour.
8:188-200.

Costerton, J. W., G. G. Geesey, and K.-J. Cheng. 1978. How
bacteria stick. Sci. Am. 238:86-95.

31


-------
Dazzo, F., P- Smith, and D. Hubbell. 1973. Vertical dispersal of
fecal coliforms in Scranton fine sand. Proc. Soil Crop Sci.
Soc. Fla. 32:99-102.

Dillon, J. k. , J. A. Fuerst, A. C. Hayward, and G. H. G. Davis.
1986. a comparison of five methods for assaying bacterial
hydrophobicity. j. Microbiol. Methods 6:13-19.

Duboise, S. M. , B. E. Moore, and B. P. Sagik. 1976. Poliovirus
survival and movement in a sandy forest soil. Appl. Environ.
Microbiol. 31:536-543.

Edgehill, R. u., and R. K. Finn. 1983. Microbial treatment of
soil to remove pentachlorophenol. Appl. Environ. Microbiol.
45:1122-1125.

Fattom, A., and M. Shilo. 1984. Hydrophobicity as an adhesion
mechanism of benthic cyanobacteria. Appl. Environ. Microbiol.
47:135-143.

Gannon, J. T. , V- B. Manilal, and M. Alexander. 1991.
Relationship between cell properties and the transport of
bacteria through soil. Appl. Environ. Microbiol. 57:190-193.

Gerba, C. P-, and J. C. Lance. 1978. Poliovirus removal from
primary and secondary sewage effluent by soil filtration.
Appl. Environ. Microbiol. 36: 247-251.

Gerba, C. P., C. Wallis, and J. L. Melnick. 1975. Fate of
wastewater bacteria and viruses in soil. J. Irrig. Drain.
Div., Proc. Am. Soc. Civ. Eng. 101(IR3):157-174.

Germann, P. F., M. S. Smith, and G. W. Thomas. 1987. Kinematic
wave approximation to the transport of Escherichia coli in the
vadose zone. Water Resour. Res. 23:1281-1287.

Goldshmid, J., D. Zohar, Y. Argaman, and Y. Kott. 1973. Effect of
dissolved salts on the filtration of coliform bacteria in sand
dunes, p. 147-157. In S. H. Jenkins (ed.) Advances in water
pollution research. Pergamon Press, New York.

Goldstein, R. M., L. M. Mallory, and M. Alexander. 1985. Reasons
for possible failure of inoculation to enhance biodegradation.
Appl. Environ. Microbiol. 50:977-983.

Hagedorn, C., E. L. McCoy, and T. M. Rahe. 1981. The potential
for groundwater contamination from septic effluents. J.
Environ. Qual. 10:1-8.

Harvey, R. W., L. H. George, R. L. Smith, and D. R. LeBlanc. 1989.
Transport of microspheres and indigenous bacteria through a

32


-------
sandy aquifer: Results of natural- and forced-gradient tracer
experiments. Environ. Sci. Technol. 23:51-56.

Hattori, T. and R. Hattori.	197 6. The physical environment in

soil microbiology: An	attempt to extend principles of

microbiology to soil	microorganisms. CRC Crit. Rev.
Microbiol. 4:423-461.

Jackson, M. L. 1974. Soil chemical analysis - Advanced course,
p. 110-114. Published by author, Dept. of Soil Sci., Univ. of
Wisconsin, Madison, Wis.

Krone, R. B., G. T. Orlob, and C. Hodgkinson. 1958. Movement of
coliform bacteria through porous media. Sew. Indus. Wastes
30:1-13.

Madsen, E. L., and M. Alexander. 1982. Transport of Rhizobium and
Pseudomonas through soil. Soil Sci. Soc. Am. J. 46:557-560.

Marshall, K. C. 1975. Clay mineralogy in relation to survival of
soil bacteria. Annu. Rev. Phytopathol. 13:357-373.

Marshall, K. C. 197 6. Interfaces in microbial ecology. Harvard
University Press, Cambridge, MA.

Marshall, K. C. 1980. Adsorption of microorganisms to soils and
sediments. p. 317-329. In G. Bitton and K. C. Marshall
(eds.) Adsorption of microorganisms to surfaces. Wiley, New
York.

Marshall, K. C., and R. H. Cruickshank. 1973. Cell surface
hydrophobicity and the orientation of certain bacteria at
interfaces. Arch. Mikrobiol. 91:29-40.

McCarty, P. L., B. E. Rittmann, and E. J. Bouwer. 1984.
Microbiological processes affecting chemical transformations
in groundwater. p. 89-115. In G. Bitton, and C. P. Gerba
(ed.) Groundwater pollution microbiology. Wiley, New York.

McClure, G. W. 1972. Degradation of phenylcarbamates in soil by
mixed suspension of IPC-adapted microorganisms. J. Environ.
Qual. 1:177-180.

Molz, F. J., M. A. Widdowson, and L. D. Benefield. 1986.
Simulation of microbial growth dynamics coupled to nutrient
and oxygen transport in porous media. Water Resour. Res.
22:1207-1216.

Mozes, N., F. Marchal, M. P. Hermesse, J. L. Van Haecht, L.
Reuliaux, A. J. Leonard, and P- G. Rouxhet.	1987.

Immobilization of microorganisms by adhesion: Interplay of

33


-------
electrostatic and non-electrostatic interactions. Biotech.
Bioeng. 30:439-450.

Mozes, N. , and P. G. Rouxhet. 1987. Methods for measuring
hydrophobicity of microorganisms. J. Microbiol. Methods 6:99-
112 .

Pedersen, K. 1981. Electrostatic interaction chromatography, a
method for assaying the relative surface charges of bacteria.
FEMS Microbiol. Lett. 12:365-367.

Peele, T. C. 193 6. Adsorption of bacteria by soils. Cornell
Univ- Agric. Sta. Memoir 197, p. 3-18.

Pekdeger, A., and G. Matthess. 1983. Factors of bacteria and
virus transport in groundwater. Environ. Geol. 5:49-52.

Rahe, T. M. , C. Hagerdon, E. L. Mc Coy, and G. F. Kling. 1978.
Transport of antibiotic-resistant Escherichia coli through
western Oregon hillslope soils under conditions of saturated
flow. J. Environ. Qual. 7:487-494.

Reed, R. W. , and G. B. Reed. 1948. Drop plate method of counting
viable bacteria. Can. J. Res. (Sec. E), 26:317-326.

Rittmann, B. E. , P. L. McCarty, and P. V. Roberts. 1980. Trace-
organics biodegradation in aquifer recharge. Ground Water
18:236-243 .

Rosenberg, M. , D. Gutnick, and E. Rosenberg. 1980. Adherence of
bacteria to hydrocarbons: A simple method for measuring cell-
surface hydrophobicity. FEMS Microbiol. Lett. 9:29-33.

Rosenberg, M., and S. Kjelleberg. 1986. Hydrophobic interactions:
Role in bacterial adhesion. Adv. Microb. Ecol. 9:353-393.

Schaub, S. A., and C. A. Sorber. 1977. Virus and bacteria removal
from wastewater by rapid infiltration through soil. Appl.
Environ. Microbiol. 33:609-619.

Smith, M. S., G. W. Thomas, R. E. White, and D. Ritonga. 1985.
Transport of Escherichia coli through intact and disturbed
soil columns. J. Environ. Qual. 14:87-91.

Sosbey, M. D., C. H. Dean, M. E. Knuckles, and R. A. Wagner. 1980.
Interactions and survival of enteric viruses in soil
materials. Appl. Environ. Microbiol. 40:92-101.

Stenstrom, T. A. 1989. Bacterial hydrophobicity, an overall
parameter for the measurement of adhesion potential to soil
particles. Appl. Environ. Microbiol. 55:142-147.

34


-------
Stotzky, G. 1985. Mechanisms of adhesion to clays with reference
to soil systems, p. 195-253. In D. C. Savage and M. Fletcher
(ed.), Bacterial adhesion: mechanisms and physiological
significance. Plenum Publishing Corp., New York.

Sutherland, I. W. 1983. Microbial exopolysaccharides - Their role
in microbial adhesion in aqueous systems. CRC Crit. Rev.
Microbiol. 10:173-201.

Tan, Y. 1989. Transport of bacteria in porous media. Ph.D.
thesis. Australian National Univ., Canberra, Australia.

Taylor, S. W. , and P. R. Jaffe. 1990. Substrate and biomass
transport in a porous medium. Water Resour. Res. 26:2181-
2194 .

van der Mei, H. C., A. H. Weerkamp, and H. J. Busscher. 1987. A
comparison of various methods to determine hydrophobic
properties of streptococcal cell surfaces. J. Microbiol.
Methods 6:277-287.

Vinten, A. J. A., and P- H. Nye. 1985. Transport and deposition
of dilute colloidal suspensions in soil. J. Soil Sci. 36:531-
541.

Viraraghavan, T. 1978. Travel of microorganisms from a septic
tile. Water Air Soil Pollut. 9:355-362.

Wilson, J. T. , J. F. McNabb, D. L. Balkwill, and W. C. Ghiorse.
1983. Enumeration and characterization of bacteria indigenous
to a shallow water-table aquifer. Ground Water 21:134-142.

Wollum, A. G. , II., and D. K. Cassel. 1978. Transport of
microorganisms in sand columns. Soil Sci. Soc. Am. J. 42:72-
76.

35


-------