Bacterial Deposition in Porous Media - American Chemical Society

Cell-solid interactions, and as a consequence R0, are determined by the following factors: (i) the type of cell- coating and the substratum hydrophobi...
0 downloads 0 Views 806KB Size
Environ. Sci. Technol. 1996, 30, 2877-2883

Bacterial Deposition in Porous Media: Effects of Cell-Coating, Substratum Hydrophobicity, and Electrolyte Concentration H U U B H . M . R I J N A A R T S , * ,† WILLEM NORDE,‡ EDWARD J. BOUWER,§ JOHANNES LYKLEMA,‡ AND A L E X A N D E R J . B . Z E H N D E R †,| Department of Microbiology, Wageningen Agricultural University, Hesselink van Suchtelenweg 4, 6703 CT Wageningen, The Netherlands, Department of Physical and Colloid Chemistry, Wageningen Agricultural University, Dreijenplein 6, 6703 HB Wageningen, The Netherlands, and Department of Geography and Environmental Engineering, The Johns Hopkins University, Baltimore, Maryland 21218

Deposition of seven bacterial strains on spherical glass and Teflon collectors was studied in vertical downflow columns at an ionic strength (I) of 0.1 M. The various bacteria had either one of the following types of major cell-surface constituents: nonpolysaccharide (NP), amphiphilic (AMPH), or anionic polysaccharide (AP) macromolecules. Deposition was analyzed in terms of the clean bed collision efficiency R0 (the probability of a cell to attach upon reaching a substratum free of cells) and a blocking factor B (the ratio of the area blocked by an attached cell to the geometric area of a cell). The value of R0 decreased from 1.0 to about 0.01 in the following order of cell-surface constituents/collector combinations: NP/Teflon and AMPH/Teflon > NP/glass > AMPH/ glass, AP/Teflon, and AP/glass. The value for B, at a Peclet number of 1 × 105, increased from about 3 to 18 in the order NP/low cell charge < NP or AMPH/ high cell charge < AP/high cell charge. This indicates that cell-cell repulsion enhances blocking. Blocking is higher on Teflon than on glass. Most likely cell-surface macromolecules adsorb in the surroundings of the attached cells and enhance blocking on Teflon. The deposition of four bacterial strains was investigated at 0.0001 M e I e 0.1 M. For I values smaller than a critical level, R0 decreased with decreasing I. The critical I is determined by the range over which cell-surface macromolecules can penetrate the repulsive Gibbs energy barrier between cell and solid. The value for B increases about 1 order of magnitude upon changing I from 0.1 to 0.001 M. Maximal control of microbial mobility in porous media can be reached in systems for which B and R0 are high at high I (0.1 M): the high B value minimizes

S0013-936X(96)00598-6 CCC: $12.00

 1996 American Chemical Society

the occurrence of pore-clogging whereas the dependencies of R0 and B on I allow manipulation of deposition by varying the ionic strength.

Introduction Control of microbial transport and retention in porous media is required for effective and safe bioremediation of contaminated soil and groundwater (1-4). A significant step toward a better understanding of bacterial mobility in porous media was made in the preceding paper in which the basic mechanisms contributing to microbial deposition in coarse grain media were analyzed (4). The present paper elaborates on that work and addresses the physicochemical factors affecting bacterial transport in coarse grain media. The initial bacterial deposition rate in coarse grain media is appropriately described by the following equation (4):

-

dc dΘ )G ) kcR dt dt

(1)

where c is the cell concentration in the bulk liquid, G ) 3(1/ - 1)/(πasab2),  is the porosity, as is the grain radius (m), ab is the radius of the bacterium (m), and Θ is the fraction of surface covered with cells, and k (s-1) is the mass transfer coefficient that is a function of the collector mass transfer efficiency, as, and the superficial velocity of the fluid phase (4). The adhesion or collision efficiency R represents the probability of a particle to attach upon reaching the collector surface and is a function of Θ (4):

R ) R0(1 - BΘ)

(2)

Here is R0, the clean bed collision efficiency (Θ ) 0), and B, the blocking factor that is defined as the ratio of the area blocked by an attached cell to the geometric area of that cell (4, 5). The R0 is primarily determined by cell-solid interactions whereas B is strongly affected by cell-cell interactions. In addition, both R0 and B depend on the hydrodynamic condition of the system (4). Cell-solid interactions, and as a consequence R0, are determined by the following factors: (i) the type of cellcoating and the substratum hydrophobicity (6-18), (ii) the amount of negative charge on the solid and the bacterial surface and the ionic strength (I) of the medium (12, 13, 15, 18-28), and (iii) the range over which cell-surface macromolecules can penetrate the Gibbs energy barrier that exists between bacterium and solid phase (24). When these interactions do not inhibit deposition, R0 ) 1. Blocking (B) may be significantly enhanced by repulsive cell-cell interactions that depend on the type of cell-coating and the ionic strength of the medium (4, 5, 27, 29, 30). The present study focuses on the physicochemical factors affecting bacterial transport through R0 and B. The dependencies of R0 and B on three physicochemical system * Corresponding author present address: TNO Institute of Environmental Sciences, Energy Research and Process Innovation, P.O. Box 6011, 2600 JA Delft, The Netherlands. † Department of Microbiology. ‡ Department of Physical and Colloid Chemistry. § Department of Geography and Environmental Engineering. | Present address: EAWAG/ETH, CH 8600, Du ¨bendorf Switzerland.

VOL. 30, NO. 10, 1996 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2877

TABLE 1

Designations and Types of Macromolecular Cell-Coating for Bacteria Studied strain

other designationa

type of cell-coatingb

C1 C3 C4 C5 C6

Coryneform Bacteria Arthrobacter sp. strain DSM 6687 Rhodococcus sp. strain C125 Rhodococcus erythropolis A177 Corynebacterium sp. strain DSM 6688 Corynebacterium sp. strain DSM 44016

AP AMPH-II AMPH-I NP NP

P3 P4

Pseudomonads Pseudomonas sp. strain B13 Pseudomonas putida mt2

AP NP

a The sources of organisms other than DSM or ATCC strains are described by Rijnaarts et al. (14). b The types of macromolecular cellcoating were deduced in a previous study (16). The following types can be discriminated: AP, anionic polysaccharide; AMPH-I, amphiphilic macromolecules of which the hydrophilic parts dominates interactions with all solid surfaces; AMPH-II, amphiphilic macromolecules of which the hydrophilic parts dominates interactions with hydrophilic surfaces and the hydrophobic parts controls the interactions with hydrophobic surfaces; NP, nonpolysaccharide cell-coating.

properties were determined: (i) the type and structure of the cell-coating, (ii) the hydrophobicity of the substratum, and (iii) the ionic strength of the aqueous phase. Column experiments were performed with defined materials, i.e., with seven bacterial strains possessing a range of cellsurface constituents (14), Teflon and glass spherical collectors, and aqueous media with various ionic strengths. The implications of our findings for microbial transport in natural porous media, like soils and aquifers, are discussed.

Materials and Methods Aqueous Media, Collectors, and Columns. Phosphatebuffered saline solutions (PBS) (pH ) 7.2) with I varying between 0.0001 and 0.1 M were previously described (14). Spherical collectors of PFA-Teflon and glass and procedures for cleaning the collectors and packing the columns are described elsewhere (14). The collector radii were 190 ( 50 µm for Teflon and 225 ( 25 for glass. The columns had an internal diameter of 1.00 cm, the porous bed length L was 9 ( 0.3 cm, and the porosity  was 0.34 ( 0.02 (4). Bacteria. The microorganisms used are listed in Table 1, and their following characteristics have been described in other papers: (i) the origins of the strains and the procedures for cultivation and preparation (14), (ii) the effective cell radii, diffusion constants, and cell surface hydrophobicities (14), (iii) the negative ζ-potential at pH 7 at various I (24), (iv) the isoelectric points and type of macromolecular cell-coating (16), and (v) the range over which the cell-surface macromolecules can penetrate electrostatic barriers (24). Column Experiments. Experiments were conducted in duplicate with vertical down-flow columns. Bacterial suspensions in PBS with an OD280 value of 0.6 were prepared by diluting samples of concentrated stock suspensions. The resulting suspensions varied in cell concentration c0 between 3 × 107 and 5 × 108 cells cm-3, depending on the strain used. These were applied to the columns by a peristaltic pump. The flow rate was kept constant for each column but varied between 16 and 21 cm3 h-1 for different columns. The effluent cell concentration c, as measured by OD280, was monitored with time. The level of c0 (OD280) was found to be constant throughout each experiment. The columns were operated for 2 h.

2878

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 30, NO. 10, 1996

Two types of experiments were performed with both glass and Teflon collectors: (i) experiments at I ) 0.1 M with all bacterial strains listed in Table 1 and (ii) experiments with strains C3, C6, P3, and P4 at ionic strengths varying between 0.0001 and 0.1 M. Determination of r0 and B. The values of R0 and B are determined from the breakthrough response of bacterial cells according to method 2 described in the preceding paper (4). The following parameters required for calculations were obtained from previous research: the Hamaker constant Abs(w) for the solid-water-bacterium system (2 × 10-22 J for Teflon and 6.2 × 10-21 J for glass (15)) and the radii and diffusion coefficients of the bacteria (14).

Results and Discussion Effects of Type of Cell-Coating and Substratum Hydrophobicity on Bacterial Transport. These effects were studied at I ) 0.1 M. The breakthrough curves of strains C1, C5, and C6 are shown in Figure 1A. The corresponding R0 values (Figure 1B) were higher for Teflon than for glass. The adhesion efficiency R decreased with Θ except for strain C1 on glass where R increased with Θ. Values of R0 and B were deduced from the R-Θ plots by linear regression analysis for all bacterium/solid combinations tested at I ) 0.1 M (Table 2). No reliable estimates of B could be obtained for the low deposition cases (R0 < 0.1), since R is very sensitive to c/c0 at c/c0 levels close to unity. Effects on Clean Bed Collision Efficiency R0. The R0 value corresponds to a Gibbs energy barrier or activation energy for adhesion ∆Gq0 (in kT) (k (J K-1) is the Boltzmann constant, T (K) is the absolute temperature] (15), i.e., ∆Gq0 ) -ln(R0). This parameter is completely determined by steric interactions between the cell-surface macromolecules and the substratum at I ) 0.1 M (15, 24). The steric hindrance, as indicated by ∆Gq0, is clearly related to the composition of the cell-coating, see Figure 2. It decreases (∆Gq0 increases) upon changing from coatings of the anionic polysaccharide (AP) type via AMPH to NP. Quantitatively, these trends also depend on the substratum hydrophobicity (Figure 2) for all but one bacterium/collector combination. Low or absent steric hindrance is observed for nonpolysaccharide cell-coatings combined with a hydrophobic substratum (Figure 2). Strain C1 on Teflon is an exception. In contrast to other anionic polysaccharide cell-coatings like that of P3, the anionic cell surface polyelectrolytes of C1 adsorb readily on Teflon and make initial adhesion on this substratum 100% efficient. Effects on Blocking Factor B. Blocking is determined by hydrodynamic factors and by cell-cell interactions (4, 5). The Peclet number NPe, which is the ratio of convection to diffusion, defines the hydrodynamic condition during deposition (4):

NPe ) 2Uas/Db

(3)

where Db (m2 s-1) is the diffusion coefficient of the bacterium. Blocking increases with increasing NPe (4, 5). The Peclet number varied between 6 × 104 and 1.8 × 105 among the different column tests due to small differences in collector radius, bacterial radius, and interstitial fluid velocity U. The blocking data were made comparable by normalizing them to NPe ) 105 according to

B* )

a+b B a10-5NPe + b

(4)

FIGURE 1. Breakthrough curves (A) and r-Θ plots (B) for the strains C1 (circles), C5 (squares), and C6 (triangles) for columns with glass (filled symbols) and Teflon (open symbols) collectors and I ) 0.1 M. TABLE 2

Experimental Values of Clean Bed Collision Efficiency r0 and Blocking Factor B at Ionic Strength of 0.1 Ma cellcoatingb strain AP AP AMPH-I AMPH-II NP NP NP

P3 C1 C4 C3 C6 C5 P4

glass

r0 0.02 ( 0.01 0.05 ( 0.02 0.012 ( 0.005 0.021 ( 0.010 0.28 ( 0.01 0.24 ( 0.01 0.12 ( 0.01

Teflon

B NDc NEGd ND ND 8.2 ( 0.7 8.1 ( 1.4 2.9 ( 2.9

r0

B

0.013 ( 0.006 1.00 ( 0.04 0.09 ( 0.02 0.83 ( 0.01 0.77 ( 0.02 0.98 ( 0.03 0.83 ( 0.02

ND 17.4 ( 1.0 ND 10.7 ( 1.0 12.5 ( 0.3 9.5 ( 0.1 2.5 ( 0.9

a The results are listed in the order of increasing R and decreasing 0 B. b Abbreviations for the types of cell-coatings as in Table 1. c ND, could not be determined because of data scatter and c/c0 values close to unity. d NEG, a negative B value was obtained as a result of an effect opposite to blocking, i.e., R increases with increasing Θ.

where B* is the normalized blocking factor. The constants a and b can be derived experimentally when the B-NPe dependency is linear over the NPe range studied as was demonstrated for strain C3 on Teflon (4) (a ) 2.9 ( 0.2 and b ) 7.3 ( 0.2). These values of a and b were used for normalizing all blocking results obtained at I ) 0.1 M (Table 3). The B* values for strains C5 and C6 show that blocking on glass is less pronounced than on Teflon, suggesting that cell-solid interactions influence B* in addition to cell-cell interactions. A likely explanation is that cell-surface macromolecules adsorb during adhesion (15). Their affinity for the solid phase and, as a consequence their tendency to stretch along the solid-liquid interface (Figure 3A), increases with increasing substratum hydrophobicity (3133). As a result, the adsorbed polymers cause greater blocked areas on Teflon than on glass.

FIGURE 2. Activation energy (∆Gq0) for initial deposition at I ) 0.1 M, derived from r0 values (15), as a function of type of cell-coating, for columns with glass (open symbols) or Teflon (filled symbols) collectors. Strain designations refer to Table 1.

The magnitude of B* increased in the order of increasing cell-surface charge and decreasing cell hydrophobicity (Table 3), namely, low charged nonpolysaccharide coating (strain P4) < highly charged nonpolysaccharide or amphiphilic coating (strains C3, C5, and C6) < highly charged cells covered with anionic polysaccharide (C1). This is consistent with increases in steric and electrosteric (cellcell) repulsion as cell-surface charge increases and cell hydrophobicity decreases (16, 24). Blocking is exceptional for strain C1 on glass and Teflon (Figure 1A,B; Tables 2 and 3). On Teflon, this strain displays the strongest blocking among all bacterium/collector combinations studied. The adsorbed anionic polysaccharide tails appear to form large blocked areas on Teflon

VOL. 30, NO. 10, 1996 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2879

TABLE 3

Normalized Blocking Factor B* on Glass and Teflon at Ionic Strength of 0.1 M, for Some of Organisms Testeda B* d strain C1 C3 C6 C5 P4

cell-coatingb AP AMPH-II NP NP NP

ζ (mV)c 38 50 42 28 11

glass NEGe NDf 8.1 ( 0.7 7.0 ( 1.3 3.0 ( 3.0

Teflon 17.8 ( 1.0 10.4 ( 1.0 14.0 ( 0.3 9.5 ( 0.1 2.7 ( 1.0

a The type of cell-coating and the ζ-potential of the bacteria at an ionic strength of 0.1 M are also listed. b Abbreviations for types of cellcoatings as in Table 1. c The values of ζ were collected from ref 24. d Blocking was normalized to a Peclet number of 1 × 105. e NEG, a negative B value was obtained as a result of an effect opposite to blocking, i.e., R increases with increasing Θ. f ND, could not be determined because of data scatter and c/c0 values close to unity.

FIGURE 4. Comparison of collision efficiencies in static batch systems, collected from ref 15, with that of column systems (r0) at an ionic strength of 0.1 M. The solid line indicates r0,column ) rbatch. The strain designations refer to Table 1.

FIGURE 3. Different mechanisms of transfer of bacterial surface macromolecules to solid substrata. (A) Adsorption in the surroundings of an attached cell. (B) Transfer by a collision of the cell with the substratum. (C) Excretion and adsorption without cell-solid contact.

(Figure 3A). On glass, the initial deposition is very low, but the deposition becomes more favorable during the course of the experiment, which is most likely due to the formation of a conditioning macromolecular layer according to mechanisms shown in Figure 3A,B. Polymer excretion and adsorption without cell-solid contact (Figure 3C) is not likely to play a role for the systems studied here (14). Comparison of R0 in Porous Media and R in Static Batch Systems. The trend in Figure 2 confirms previous observations on deposition of bacteria on flat sheet collectors in batch systems (15). Comparison of the batch and columnderived values of R (Figure 4) shows good agreement (same order of magnitude) for all but one bacterial/collector combinations. The column results should be considered more reliable than the batch results because columns provide better defined hydrodynamic conditions and allow the determination of B. However, the reasonable correlation between Rbatch and R0 indicates that blocking was not a major factor influencing the batch results. This was the consequence of diffusion controlling particle transport in the batch system, which is a random nondirectional process that excludes blocking according to a “hydrodynamic shading” mechanism (4). A great discrepancy between batch and column results was only observed for strain C1 on Teflon (Figure 4). Possibly the adsorption of the surface polymers of this organism caused a substantial blocking under static batch conditions. Effect of Ionic Strength I on Bacterial Transport. Typical examples of the results of deposition experiments at various ionic strengths are shown in Figure 5A,B, where

2880

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 30, NO. 10, 1996

breakthrough curves and R-Θ plots of strain C3 and Teflon are shown. Initial deposition (R0) decreased and the slope of the R-Θ plots varied with I. The variation between the duplicates gives a good impression of the reproducibility, which is generally satisfactory. An exception is the low deposition case at I ) 0.001 M where the uncertainty in B exceeds 50% of the average value. The dependencies of R0 and B on I are shown in Figure 6A-D for all the cases tested. Values of B on glass could only be estimated for strain C6 at I g 0.001 M and for strain P4 at I ) 0.1 M (Figure 6D). Effect of I on R0. The R0 values of strains C3 and C6 on Teflon decreased in a similar way with decreasing I (Figure 6A), whereas for strain P4 and P3 the values were hardly influenced by I. Previous investigations with static batch systems demonstrated that cell-surface macromolecules can penetrate the repulsive double-layer barriers between cells and solids (15, 24). Most bacterial surface polymers can reach the substratum at high I. For example, at an I of 0.1 M, R0 is at its maximum and is controlled by steric interactions (Figure 7A). The separation between the bacterial and solid surface created by the electrostatic barrier becomes wider upon decreasing the ionic strength (Figure 7C). Below a certain critical ionic strength level Is, the number of the cell-surface macromolecules that are not long enough to reach the substratum increases with decreasing I. As a consequence, double-layer repulsion reduces R0 with decreasing I at I < Is (Figure 7C). The initial deposition (R0) of strain P3 is low and hardly influenced by the I (Figure 6A,B). Probably the anionic polysaccharides of this strain penetrate into the electrostatic barrier region between cell and solid and thus inhibit deposition even at an ionic strength as low as 0.001 M (24). The deposition of the pseudomonad P4 on Teflon is high and only inhibited by double-layer repulsion at I < 0.001 M ((24) Figure 6A). This indicates that rather long structures are also present on the cell-surface of this strain which, in this case, bridges the cell-solid separation (15, 16, 24).

FIGURE 5. Breakthrough curves (A) and r-Θ plots (B) for strain C3 and columns with Teflon collectors at various ionic strengths I. Filled and open symbols indicate duplicate results.

FIGURE 6. Clean bed collision efficiency r0 (A and B) and blocking factor B (C and D) as a function of the ionic strength I for strains C3, C6, P3, and P4 in columns with glass and Teflon collectors.

For the strains C3 and C6, the Is levels are about 0.01 M in both column (Figure 6A,C) and batch (24) systems. This suggests that the extension of the cell surface macromolecules for these organisms is somewhat smaller than for the strains P3 and P4. A difference between the Is value in columns (0.001 M; Figure 6B) and in batch (0.01 M (24)) was found for strain P4 on glass. Except for this small discrepancy, the extension and chemical nature of bacterial surface macromolecules appear to exert a similar influence on the initial deposition in dynamic (columns) and static (batch) systems at ionic strengths between 0.1 and 0.0001 M. Effect of I on B. The blocking factor B of strains P4, C3, and C4 displayed an I dependency opposite to that of R0 (Figure 6B): B increased with decreasing I. The low deposition of strain P3 prevented an estimation of B.

FIGURE 7. Illustration summarizing the ionic strength effects on bacterial deposition. The escape probability is indicated by the arrows directed away from the substratum: a small arrow indicates a low escape probability (high collision efficiency r) whereas a thick and long arrow indicates a high escape probability (low r). High ionic strength: (A) Cell-surface macromolecules can penetrate the electrostatic barrier (eb) and can easily reach the substratum (s), which results in a maximum r. (B) Blocking is at a minimum as a result of minimal cell-cell repulsion. Nevertheless, the blocked area (ba) reduces r. Low ionic strength: (C) Most cell-surface macromolecules cannot reach the substratum (s), which results in a contribution of double-layer repulsion; r0 decreases (escape probability increases) with decreasing I. Some cells can still attach. (D) Electrostatic cell-cell repulsion is increased, which causes an enhanced blocking and an additional reduction in r.

Blocking increases with decreasing ionic strength (Figure 6B,D), i.e., with increasing electrostatic repulsion between adhered and depositing cells (Figure 7B,D). The electrostatic effect is further demonstrated by the increase in B

VOL. 30, NO. 10, 1996 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2881

FIGURE 8. Clean bed collision efficiency r0 as a function of blocking factor B. This figure summarizes all deposition data. The bacterium/ substratum/ionic strength combinations in sectors I-IV display a different deposition and transport behavior as explained in the text. The number after the strain designation indicates the ionic strength. When no number is given, I ) 0.1 M.

with decreasing I being more pronounced for the highly charged strains C3 and C6 than for the weakly charged P4 cells (Table 3; Figure 6A). These observations confirm predictions and findings of others (4, 5, 27, 30). Effect of I on Bacterial Transport through R0 and B. In general, the ionic strength influenced bacterial deposition by affecting both the initial adhesion efficiency R0 and the blocking factor B: R0 decreases (Figures 7A, 7C, 6A, and 6C) and B increases with decreasing I (Figures 7B, 7D, 6B, and 6D). Both effects cause a reduction of R upon decreasing I (eq 2). Relevance for Microbial Mobility in Porous Media. The parameters R0 and B characterize the first two steps of bacterial accumulation at solid-liquid interfaces: (i) the initial deposition (R0) and (ii) the level of surface saturation that can be reached (B). Since B indicates the extent of cell-cell interactions it may also provide information on bacterial deposition at surface coverages close to or beyond saturation. Weakly blocking bacterial species will form multiple cell layers at higher coverages and therefore tend to clog a porous medium after prolonged application of the cell suspension (4, 30). In contrast, the deposition of strongly blocking organisms stops after saturation of the surface. This prevents clogging and allows newly introduced cells to be transported down-gradient throughout the porous medium (4, 34). An evaluation of the mobility of the different bacterial species in the model porous media can be illustrated using an R0-B plot (Figure 8) that is divided into four regions. Region I comprises bacterium/collector combinations with a high R0 and a low B value, which cause fast deposition to a high coverage level. A continued feeding of a cell suspension with region I characteristics would cause multiple cell layers to form and eventually lead to poreclogging. This was demonstrated to occur for strain P4 on Teflon (4). Region II contains bacterium/collector combinations with a fast initial deposition, but strong blocking will result in low final coverages, resulting in an absence of clogging and an unhindered microbial transport through the porous medium after surface saturation. This was observed with strain C3 and a Bacillus species in columns packed with Teflon collectors (4) and sand grains (34), respectively. Region III includes the bacteria with a slow deposition (low R0 value) that can reach high coverages

2882

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 30, NO. 10, 1996

because B is low; they may even cause pore-clogging after a prolonged input of a cell suspension. Region IV contains the cases for which deposition is highly unfavorable due to low R0 and high B. Although the boundaries between the different sectors in Figure 8 cannot be fully established with the data, it appears that most bacterium/collector combinations studied fall into regions II and IV, which are considered to be nonclogging systems. Bacterial transport in such systems can be quantitatively described by an advective-dispersion equation extended with the deposition rate term (4) given by eqs 1 and 2. Variation of the ionic strength can be used to influence microbial mobility (i.e., cause a shift from sector II to sector IV). The bacterium/collector combinations in regions I and III may be considered as clogging systems. Microbial transport can only be quantitatively described with eqs 1 and 2 for low coverage conditions before the formation of multiple cell layers. Reduction in I may shift the system to nonclogging where microbial mobility is increased. A shift to sector II or sector IV is therefore caused, as shown for strain P4 on Teflon (Figure 8). Plots similar to Figure 8 may be helpful to assess the mobility of pathogens, beneficial microorganisms, and genetically engineered bacteria in natural and engineered systems. For instance, such a plot constructed with data obtained with aquifer and sediment grains in combination with a bacterial dispersal model that includes eqs 1 and 2 will greatly facilitate the selection of an appropriate organism and suitable ionic strength for seeding a contaminated aquifer with a bacterium specialized in degrading a specific pollutant. The results presented here were obtained under nongrowth conditions. In practice, growth will often occur after or simultaneously with deposition (35, 36). Since the blocking factor is indicative for cell-cell interactions, B may be of relevance for ascertaining the ease at which deposited cells can grow into a stabile biofilm. A first confirmation of this hypothesis is that Pseudomonas strains similar to the weakly blocking Ps. putida (P4) are known to colonize surfaces rapidly and to form stabile biofilms (37).

Conclusion Bacterial deposition in porous media can be adequately described with the clean bed adhesion efficiency R0 and blocking factor B under low coverage conditions. The dependencies of these parameters on factors like the type and length of cell-surface macromolecules, substratum hydrophobicity, and ionic strength provide ample possibilities to manipulate microbial mobility in soil, aquifers, and biofilm reactors. Furthermore, B appears to be a suitable characteristic to discriminate between microorganisms that are capable of pore-clogging at higher attached cell concentrations and those that can be transported freely after surface saturation. The relationship between B and the occurrence of pore-clogging and formation of biofilms emerge as primary subjects for further research. The relationships between the deposition parameters R0 and B and cell-coating, substratum hydrophobicity, and ionic strength hypothesized from the present study should be further explored using a broader range of different bacteria and various natural and man-made materials. This would greatly facilitate and improve the assessment of the mobility and retention of bacteria in natural porous media and engineered biofilm reactors.

Acknowledgments This research was funded by The Netherlands Integrated Soil Research Programme, Grant C6-1\8939.

Literature Cited (1) Updegraff, D. M. In Modeling the environmental fate of microorgansims; Hurst, C. J., Ed.; American Society of Microbiology: Washington, DC, 1991; pp 1-20. (2) Bouwer, E. J.; Zehnder, A. J. B. Trends Biotechnol. 1993, 11, 360367. (3) Van der Meer, J. R.; Bosma, T. N. P.; De Bruin, W. P.; Harms, H.; Holliger, C.; Rijnaarts, H. H. M.; Tros, M. E.; Schraa, G.; Zehnder, A. J. B. Biodegradation 1992, 3, 265-284. (4) Rijnaarts, H. H. M.; Bouwer, E. J.; Norde, W.; Lyklema, J.; Zehnder, A. J. B. Environ. Sci. Technol. 1996, 30, 2869-2876. (5) Dabros, T.; Van de Ven, T. G. M. Colloids Surf. A: Physicochem. Eng. Aspects 1993, 75, 95-104. (6) Absolom, D. R.; Lamberti, F. V.; Policova, Z.; Zingg, W.; Van Oss, C. J.; Neumann, A. W. Appl. Environ. Microbiol. 1983, 46, 90-97. (7) Bendinger, B.; Rijnaarts, H. H. M.; Altendorf, K.; Zehnder, A. J. B. Appl. Environ. Microbiol. 1993, 59, 3973-3977. (8) Busscher, H. J.; Weerkamp, A. H.; van der Mei, H. C.; Pelt, A. J. W.; de Jong, H. P.; Arends, J. Appl. Environ. Microbiol. 1984, 48, 980-983. (9) Busscher, H. J.; Uyen, M. H. M. J. C.; Pelt, A. J. W.; Busscher, H. J.; Weerkamp, A. H. FEMS Microbiol. Rev. 1987, 46, 165-173. (10) Dickson, J. R.; Koohmaraie, K. Appl. Environ. Microbiol. 1989, 55, 832-836. (11) Fletcher, M.; Loeb, G. L. Appl. Environ. Microbiol. 1979, 37, 6772. (12) Gilbert, P.; Evans, D. J.; Evans, E.; Duguid, I. G. J. Appl. Bacteriol. 1991, 71, 72-77. (13) Harvey, R. W. In Modeling the environmental fate of microorgansims; Hurts, C. J., Ed.; American Society for Microbiology: Washington, DC, 1991; pp 89-114. (14) Rijnaarts, H. H. M.; Norde, W.; Bouwer, E. J.; Lyklema, J.; Zehnder, A. J. B. Appl. Environ. Microbiol. 1993, 59, 3255-3265. (15) Rijnaarts, H. H. M.; Norde, W.; Bouwer, E. J.; Lyklema, J.; Zehnder, A. J. B. Colloids Surf. B: Biointerfaces 1995, 4, 5-22. (16) Rijnaarts, H. H. M.; Norde, W.; Lyklema, J.; Zehnder, A. J. B. Colloids Surf. B: Biointerfaces 1995, 4, 191-197. (17) Van Loosdrecht, M. C. M.; Lyklema, J.; Norde, W.; Schraa, G.; Zehnder, A. J. B. Appl. Environ. Microbiol. 1987, 53, 1893-1897. (18) Van Loosdrecht, M. C. M.; Norde, W.; Lyklema, J.; Zehnder, A. J. B. Aquat. Sci. 1990, 52, 103-114.

(19) Elimelech, M.; O’Melia, C. R. Environ. Sci. Technol. 1990, 24, 1528-1536. (20) Gannon, J.; Tan, Y.; Baveye, P.; Alexander, M. Appl. Environ. Microbiol. 1991, 57, 2497-2501. (21) Gordon, A. S.; Millero, F. Appl. Environ. Microbiol. 1984, 47, 495-499. (22) Martin, R. E.; Hanna, L. M.; Bouwer, E. J. Environ. Sci. Technol. 1991, 25, 2075-2082. (23) Meinders, J. M.; Busscher, H. J. Colloid Polym. Sci. In press. (24) Rijnaarts, H. H. M. Interactions between Bacteria and Solid Surfaces in Relation to Bacterial Deposition in Porous Media. Ph.D. Thesis, Wageningen Agricultural University, Wageningen, The Netherlands, 1994. (25) Rutter, P. R.; Vincent, B. In Microbial adhesion and aggregation; Marshall, K. C., Ed.; Springer-Verlag; Berlin, 1984 (26) Rutter, P. R.; Vincent, B. In CRC Series in mathematical models for microbiology; Bazin, J. M., Ed.; CRC Press, Inc.: Boca Raton, FL, 1988; pp 87-107. (27) Sjollema, J.; Weerkamp, A. H.; Busscher, H. J. Biofouling 1990, 1, 101-112. (28) Van Loosdrecht, M. C. M.; Lyklema, J.; Norde, W.; Schraa, G.; Zehnder, A. J. B. Appl. Environ. Microbiol. 1987, 53, 1898-1901. (29) Adamczyk, Z.; Siwek, B.; Zembala, M. Colloids Surf. 1992, 62, 119-130. (30) Song, L.; Elimelech, M. Colloids Surf. A: Physicochem. Aspects 1993, 73, 49-63. (31) Bonekamp, B. C. Colloids Surf. 1989, 41, 267-286. (32) Fleer, G. J.; Scheutjens, J. M. H. M. In Coagulation and flocculation; Dobias, B., Ed.; Marcel Dekker Inc.: New York, 1993; pp 209-263. (33) Van der Schee, H. A.; Lyklema, J. J. Phys. Chem. 1984, 88, 66616667. (34) Lindqvist, R.; Enfield, C. G. Microb. Ecol. 1992, 24, 25-41. (35) Cunningham, A. B.; Bouwer, E. J.; Characklis, W. G. In Biofilms; Characklis, W. G., Marshall, K. C., Eds.; John Wiley & Sons Inc.: New York, 1990; pp 697-732. (36) Sharma, P. K.; McInerney, M. J.; Knapp, R. M. Appl. Environ. Microbiol. 1993, 59, 3686-3694. (37) Caldwell, D. E.; Lawrence, J. R. Microb. Ecol. 1986, 12, 299-312.

Received for review July 9, 1996. Accepted July 9, 1996.X ES9605984 X

Abstract published in Advance ACS Abstracts, August 1, 1996.

VOL. 30, NO. 10, 1996 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2883