• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Bacterial transport experiments in fractured crystalline bedrock

Posted on: Friday, 26 September 2003, 06:00 CDT

Abstract

The efficiency of contaminant biodegradation in ground water depends, in part, on the transport properties of the degrading bacteria. Few data exist concerning the transport of bacteria in saturated bedrock, particularly at the field scale. Bacteria and microsphere tracer experiments were conducted in a fractured crystalline bedrock under forced-gradient conditions over a distance of 36 m. Bacteria isolated from the local ground water were chosen on the basis of physicochemical and physiological differences (shape, cell-wall type, motility), and were differentially stained so that their transport behavior could be compared. No two bacterial strains transported in an identical manner, and microspheres produced distinctly different breakthrough curves than bacteria. Although there was insufficient control in this field experiment to completely separate the effects of bacteria shape, reaction to Gram staining, cell size, and motility on transport efficiency, it was observed that (1) the nonmotile, mutant strain exhibited better fractional recovery than the motile parent strain; (2) Gram- negative rod-shaped bacteria exhibited higher fractional recovery relative to the Gram-positive rod-shaped strain of similar size; and (3) coccoidal (spherical-shaped) bacteria transported better than all but one strain of the rod-shaped bacteria. The field experiment must be interpreted in the context of the specific bacterial strains and ground water environment in which they were conducted, but experimental results suggest that minor differences in the physical properties of bacteria can lead to major differences in transport behavior at the field scale.

Introduction

Bioaugmentation, which involves amendments of genetically engineered or specialized bacteria to contaminated subsurface environments for purposes of enhancing biodegradation, is becoming a potential alternative to conventional pump-and-treat technologies. Although considerable information has been gained from microcosm and in situ studies involving the tractability of bioaugmentation for remediation of contaminated granular aquifers (Li and Logan 1999; Harkness et al. 1999; Steffan et al. 1999; Munakata-Marr et al. 1997), little is known about the potential of bioaugmentation for enhancing biodegradation in contaminated fractured rock. There is a dearth of laboratory data concerning transport of bacteria through fractures and applicable field data are almost nonexistent. In the absence of experimental observations, bacteria are often treated like abiotic colloids in transport models (Kim and Corapcioglu 1996; Corapcioglu and Haridas 1984; Hornberger et al. 1992). In reality, bacteria vary considerably in terms of their size, morphology, motility, and surface chemistry, which lead to substantive differences in their propensities for attachment to solid surfaces within aquifers (Harvey and Garabedian 1991). A better understanding of the role of bacterial cell characteristics upon their transport behavior in fractured media is needed if more realistic bacterial transport models are to be developed and applied to bioaugmentation in contaminated bedrock. This information would also improve predictions of pathogenic bacteria through fractured media.

This paper documents a field-scale experiment comparing transport behavior of bacteria with varying physico-chemical characteristics and motility. Specifically, we examined the effects of cell size, morphology (general shape), Gram stain type (indicative of fundamental compositional differences in the cell envelope), and motility (independent locomotion by means of a flagellum or flagella). All of the aforementioned characteristics have the potential to affect transport through fractured rock. For example, one might hypothesize that a larger bacterium would be more resistant to filtration than a smaller bacterium due to its decreased rate of Brownian motion, but may be more susceptible to settling during transport. Flagellated bacteria may be more capable of moving into relatively immobile water, and thus be able to reach trapped contaminants. At the same time, the mean transport rate of flagellated bacteria may be smaller than that of a non-flagellated bacteria.

Physicochemical characteristics have been shown to affect transport in porous media at the field scale (Metge et al. 2000), but the effects in fractured bedrock are unknown. We considered it possible that hydrodynamic heterogeneity typical of structured media such as fractured bedrock may be more important to transport than the difference in bacteria-surface interactions. If this were the case, then bacteria species of similar size but varying physicochemical characteristics might transport in a similar manner. This is an important hypothesis, as it implies bacteria transport in structured geologic media can be described using relatively simple colloid-based transport models. The transport of all bacteria species might then be predicted based upon a "reference" bacterium. To test this hypothesis, we injected species of bacteria of similar size but varying physicochemical characteristics into a saturated crystalline bedrock, under an induced hydraulic gradient. Differences in bacterial transport behavior was assessed by comparing the arrival concentrations of bacteria at an extraction well. As will be described in this paper, bacteria species of similar size but varying physicochemical characteristics did not transport in a similar manner. There was insufficient control in this field experiment to understand bacteria-surface interactions or even to completely distinguish the effects of cell size, morphology, Gram type, and motility on transport rates. Certain interesting trends were identified with respect to the relative breakthrough of the bacteria, however, that may have important implications for enhancing bioremediation of structured geologic media.

Previous Work

We are aware of one field-scale tracer test regarding bacteria transport in fractured rock. Champ and Schroeter (1988) assessed the potential for bacterial transport in fractured crystalline rock by injecting Escherichia coli, polystyrene microspheres, and inorganic and radioactive solute tracers into the fractured bedrock underlying the Chalk River Nuclear Laboratories. They found that peak concentrations of microspheres and E. coli arrived at the extraction well faster than those of solute tracers, but were subject to considerable attenuation. Filter factors estimated from breakthrough curves in the fractured media were reported to be similar to those found in gravel aquifers, indicating a strong potential for bacterial transport in fractured crystalline rock. Story et al. (1995) performed laboratory bacterial tracer tests in samples of fractured sandstone and volcanic tuff. In sandstone cores, where water flowed evenly through the matrix, bacteria were transported in a dispersed manner throughout the sandstone, whereas bacteria were transported primarily along preferred flowpaths (fractures or macropores) in permeable tuff cores.

Field tracer tests have been conducted in fractured bedrock using participate tracers other than bacteria, such as microspheres and viruses. Vilks and Bachinski (1996), Reimus (1995), and Cumbie and McKay (1999) used polystyrene microspheres in laboratory studies of transport through fractured media. Champ and Schroeter (1988), Becker et al. (1999), and Vilks et al. (1997) used similar microspheres in field experiments conducted in crystalline bedrock. Tracer experiments that have compared microsphere and solute in fractured media have generally shown strong filtration of microspheres, and shorter mean breakthrough times of microspheres compared to solutes. Filtration of microspheres is apparently due to capture on fracture walls, diffusion into immobile pore spaces, and/ or gravitational settling (Cumbie and McKay 1999). Harvey et al. (1989) and Lawrence and Hendry (1996) point out, however, that because of the complexities involved with the transport, growth, and decay of bacteria, microspheres have limited applicability as an analog for bacteria in transport experiments. McKay et al. (1993) used bacteriophage (viruses that infect bacteria) along with solute tracers in fractured clay till, but viruses are generally ~10 limes smaller than bacteria, so are not expected to be good transport analogs for bacteria.

Harvey (1997) provides a review of field-scale microbe transport experiments in porous media. In some of these experiments, indigenous bacterial isolates were cultured in the laboratory, fluorescently labeled, and injected as ground water tracers (Harvey et al. 1989; Harvey and Garabedian 1991; Harvey et al. 1993). Use of indigenous aquifer bacteria avoids adverse impacts on ground water quality and increases the likelihood of survival in situ. As in fractured media, bacteria tend to be highly filtered during transport through sand and gravel aquifer systems.

Physicochemical and physiological factors such as size (Fontes et al. 1991), motility, surface chemistry (van Loosdrecht et al. 1987; Truesdail et al. 1998; Scholl and Harvey 1992), buoyant density (Harvey et al. 1997), shape (Weiss et al. 1995), and external cell structures (Otto et al. 1999) have been exam\ined in laboratory settings. Bacteria adhesion is controlled, in part, by electrostatic and hydrophobic interactions between cell walls and mineral surface or coatings. Bacterial cells tend to have complex three-dimensional surfaces that exhibit a moderately weak net negative charge over the pH ranges found in most ground water environments (van Loosdrecht et al. 1989). Cell-attachment behavior is highly influenced by the hydrophobicity of cell-surfaces (Fontes et al. 1991; van Loosdrecht et al. 1987; Rijnaarts et al. 1995b). Both electrostatic and hydrophobic interactions are highly affected by changes in water chemistry (Rijnaarts et al. 1995a; van Loosdrecht et al. 1989; Lawrence and Hendry 1996; Fontes et al. 1991; van Loosdrecht et al. 1987). Perhaps as a consequence, little direct correlation has been found between Gram staining and the degree of hydrophobicity of the cell envelope (Gannon et al. 1991). In addition, external cell structures such as fimbriae and extracellular polymers may play a dominant role in the reversibility of bacterial attachment to surfaces (Otto et al. 1999).

Iron oxyhydioxides that are commonly found on mineral surfaces in saturated subsurface environments have been shown to affect bacterial transport (Scholl and Harvey 1992; Knapp et al. 1998; Truesdail et al. 1998; Mills et al. 1994). In general, both mineral and bacteria surfaces have a weak negative charge under acidic to near-neutral conditions (Truesdail et al. 1998). However, iron oxyhydroxides that modify granular and fracture surfaces are positively charged under the aforementioned pH range and, consequently, facilitate bacterial attachment. Even at abundances present in natural porous media, laboratory experiments predict that iron oxyhydroxides can have a strong effect on bacterial transport (Scholl and Harvey 1992; Knapp et al. 1998). Natural abundances of iron oxyhydroxide coatings also have been found to affect viral transport in porous media field experiments (Pieper et al. 1997). We are unaware of experiments that investigate the relationship between bacterial transport and iron oxyhydroxide coatings on fracture surfaces. Although the effect of iron oxyhydroxide upon transport of bacteria in granular media has been investigated in a few studies (Mills et al. 1994; Scholl et al. 1990), more information is needed in order to predict the effects in aquifers, which vary considerably in pH, iron oxide abundance, and the degree by which iron oxyhydroxides are coated with organic matter.

In sum, research to date has demonstrated that the relationship between the bacterial properties and their transport behavior in saturated geologic media is complex. Under field conditions, bacteria suspended in moving ground water will encounter a variety of fluid velocities, mineral surfaces, and solution chemistry. A mechanistic interpretation of bacterial interactions for these experiments is not possible. Our objective in these experiments, therefore, was to (1) determine if different types of bacteria transported in a different manner in fractured bedrock, and (2) look for gross trends in cell properties with respect to transport behavior.

Experimental Methods

Bacterial Tracers

Indigenous bacteria were isolated from the aquifer at the Mirror Lake site in order to preclude microbial contamination of the Hubbard Brook ground water ecosystem. Bacterial isolates were prepared by plating serial dilutions of natural ground water onto 1/ 10 strength nutrient agar plates (Difco nutrient agar media with 1 g/ L glucose and ground water), which were then incubated aerobically at ~22[degrees]C for five to eight days. Initial screening of cultured ground water isolates was performed to assess overall morphology, Gram type, and motility. The four chosen isolates differed in terms of shape, Gram-stain reaction, motility, and size.

Table 1

Bacterial Strains Used as Tracers in Field Experiments

The bacterial strains used for the field tracer test are listed in Table 1, along with their genera, and physical and physiological characteristics. The strains isolated from the aquifer at the Mirror Lake (ML) site were designated ML1, ML2, ML2m, and ML3, where "m" indicates a strain produced through UV-induced mutation. Gram staining refers to a staining procedure designed to assess the presence (Gram negative) or absence (Gram positive) of a lipopolysaccaride outer membrane, exterior to the cell wall. The rationale for including both Gram types in the suite of test organisms is that the aforementioned differences in the nature of the cell envelopes may lead to measurable differences in affinities for mineral surfaces in ground water (Gannon et al. 1991). Two Gram- positive strains were obtained: a rod-shaped (strain ML1) and a coccoidal-shaped bacterium (strain ML3), subsequently identified by DNA-sequencing (Microbial Insights, Knoxville, Tennessee) as belonging to the genera Microbacterium and Staphylococcus, respectively. The motile Gram-negative rod (Pseudomonas sp. strain ML2) exhibited a spreading pattern with complete plate coverage within a few days. A nonmotile mutant strain (ML2m) of the pseudomonad was developed by subjecting a liquid culture to ultraviolet (254 nm) radiation for 10 minutes. The UV-exposed liquid culture was then reisolated on MacConkey plates. A colony that did not exhibit spreading was further examined using flagellar stain and subsequently subcultured and the degree of genetic similarity to the parent strain compared by DNA sequencing (Microbial Insights).

Prior to the field-scale test, the four isolates were cultured in low-nutrient liquid media and pelleted to concentrate bacteria (centrifuged at 3500 times g, 10[degrees]C, 15 minutes). The four isolates were stained with different DNA-specific fluorescent dyes that enabled each strain to be differentially enumerated under an epifluorescent microscope, based on wavelength of excitation and emission, and on morphological differences. Thus, several strains of bacteria could be injected simultaneously and enumerated separately in samples taken at the extraction well. The ML3 and ML2m strains were stained with 4,'6-diamidino-2-phenylindole (DAPI) at 1 mg/L (final concentration), whereas the ML2 strain was stained with Molecular Probes fluorochrome SYBR(R) Green I nucleic acid stain (product number S-7563 at a dilution of 1/1000). The ML1 was stained with ethidium bromide (100 mg/L, final concentration). All stained isolates were repelleted and resuspended in filtered (1 [mu]m pore size) Mirror Lake ground water. The suspensions of stained bacterial isolates were maintained at 3[degrees]C prior to the forced gradient injection test at the field site. These stains are not expected to alter the viability of the bacteria over the time spans of these experiments (Chen and Koopman 1997; Fuller et al. 2000).

Bacteria were enumerated in the field using an epifluorescent- microscopy direct-count method (Harvey et al. 1984). Samples taken at the withdrawal well were vacuum filtered (<10 psi vacuum) onto 0.2 [mu]m (pore size) black polycarbonate membrane filters, which were mounted on glass slides. Bacteria were enumerated under a Nikon Optiphot II epifluorescent microscope by choosing fields at random until at least 100 cells and seven fields were counted. Counts for each slide were repeated two or three times, and standard errors were estimated to be within 10% of the mean. Size ranges of the bacteria were accomplished using a computer imaging system, calibrated to stained microspheres of known diameter.

Field Scale Tracer Test

A tracer test was conducted at the U.S. Geological Survey Fractured Rock Research Site, located near Mirror Lake, in Grafton County, New Hampshire. The site lies within the Hubbard Brook Experimental Forest, operated by the U.S. Forest Service in the Southern White Mountains. Over the last 10 years, a number of hydraulic and tracer tests have been conducted within the FSE (Forest Service East) wellfield that have provided a conceptual model for flow and transport in the fractured bedrock at the site. Background and further details concerning characterization of the FSE wellfield are thoroughly documented in the published literature (Becker and Shapiro 2000; Day-Lewis et al. 2000; Shapiro and Hsieh 1998). Bedrock at the FSE wellfield is composed primarily of granitoids that have intruded the peletic schist country rock (Johnson and Dunstan 1998). A combination of tectonic and unloading stresses have resulted in the formation of a complex fracture network throughout the crystalline bedrock to a depth of at least 300 meters. Individual fractures tend to extend less than 10 meters in length, so permeability and transport is thought to be along interconnected fractures and fracture zones (Day-Lewis et al. 2000). The specific aperture and distribution of these fractures remains unknown, however. The tracer experiment was conducted between wells FSE 6 and FSE 9, which have been the subject of other tracer experiments (Becker et al. 1999; Becker and Shapiro 2000).

Iron oxyhydroxide coatings have been observed in outcrops and cores in and near the field site (Johnson and Dunstan 1998). The local ground water is in a slightly reduced state due to the proximity of a septic tank, so that iron precipitates out of water after a period of exposure to the atmosphere. It is possible, therefore, that the aeration of water during the injection process of this and previous experiments (Becker et al. 1999; Becker and Shapiro 2000) resulted in iron oxyhydroxide coatings on the mineral surfaces. Unfortunately, we have no way of directly confirming the presence of such coatings. The presence of mineral coatings is important to the interpretation of the experiment, as iron oxyhydroxide coatings have been shown to strongly affect bacterial transport in unstructured porous geologic media (Scholl and Harvey 1992; Knapp et al. 1998; Truesdail et al. 19\18; Mills et al. 1994).

The experiment was conducted in a radially convergent "slug- injection" configuration. In this configuration, FSE 6 was pumped at ~10.7 L/min for 41 hours to establish a pseudo-steady-state flow to the withdrawal well. A known volume of traced water was then injected in FSE 9 situated 36 m away, over a short period of time. The volume of injection fluid was controlled using a three pneumatic packer system in which an upper and lower packer hydraulically isolated the injection interval in the borehole, and a center packer sealed the conductive fracture (Shapiro and Hsieh 1996). With the center packer sealing the conductive fracture, fluid in the borehole was fully mixed with the tracer solution by circulating traced and untraced water through a mixing lank at the surface (total volume ~35 L). Mixing was accomplished using a pump located at the surface. Injection was initiated by deflating the center packer and letting a measured volume of traced water enter the formation under the force of gravity. The injection volume of 15 L took ~6 minutes to enter the formation. Cessation of injection occurred when the center packer was reinflated and fully sealed the conductive fracture. This method allowed for an accurate measurement of the mass of tracer and bacteria injected during the experiment. The injection volume was flushed with clean formation water after the completion of the experiment to remove remaining injectate.

Two injections were conducted to prevent flocculation of positively and negatively charged microspheres in the injection system. The first injection included deuterated water (D^sub 2^O where D is the hydrogen isotope ^sup 2^H), negatively charged microspheres (carboxylate-modified polystyrene, diameter 1.0 [mu]m with 0.066 [mu]m standard deviation, Interfacial Dynamics Corp., Portland, Oregon), and bacteria. The second injection, 24 hours after the first, was conducted in exactly the same manner but included only the positively charged microspheres (amidine-modified polystyrene, diameter 1.0 [mu]m with 0.066 [mu]m standard deviation, Interfacial Dynamics Corp., Portland, Oregon). These positively charged microspheres were never detected at the pumping well.

Deuterated water was used as a conservative solute tracer because at the injected concentrations it is not expected to impact the chemistry or density of the injectate mixture, as an anionic tracer might (Becker and Coplen 2001). Although deuterated water was injected in a pure form (D^sub 2^O), deuterated hydrogen is expected to achieve equilibrium with hydrogen in the formation water. As a result, the tracer measured at the extraction well is actually HDO (Becker and Coplen 2001). Flocculation of bacteria was not observed in test tubes prior to injection, but were introduced in sequence anyway to enhance dilution and thereby decrease the chance any flocculation might occur under the chemical environment of the formation water. The inoculate concentrations were diluted by a factor of ~30, once they were mixed in the borehole solution. Cell counts in the injectate fluid ranged from 2 x 10^sup 8^ to 3 x 10^sup 10^ cells/L. Samples from the withdrawal well, FSE 6, were collected over a 44-hour period and analyzed on site (except for deuterated water) in order to generate tracer breakthrough curves for each constituent.

Results and Discussion

Less than 4% of the bacteria and microsphere tracers were recovered, whereas >90% of the deuterated water tracer was recovered during the experiments (Table 2). None of the positively charged amidine-modified microspheres were observed at the extraction well. The poor relative recovery of the bacteria and microspheres indicates that particles are heavily filtered in this system. Similar results were observed in previous tracer tests between these two wells, wherein most of the solute tracer mass was recovered, compared with a fractional recovery of only 6% to 19% for the microspheres (Becker et al. 1999).

Breakthrough curves for all recovered tracers are shown in Figure 1. Data are plotted as the fraction of recovered mass versus time, because concentrations are too variable in range and from point-to- point to produce legible breakthrough curves. In addition, plotting data in this manner allows the relative arrival rate of some curves to be compared without sensitivity to the position of the breakthrough peak, which is a function of tracer recovery (Becker et al. 1999). The coccus (ML3) bacteria arrived slightly ahead of the deuterated water, indicating that there was some enhanced transport of this bacterium over the solute tracer. Enhancement of particulate tracers relative to solutes are expected due to hydrodynamic chromatography and selective hydrodynamic dispersion, and has been observed in previous colloid tracer experiments between these two wells (Becker et al. 1999). It is not possible to say whether the transport of the other particulate tracers were enhanced, however, because filtration lowered their concentration such that crossover was not observed in the fractional breakthrough curves.

Particulate tracers behaved in a dissimilar manner from one another; i.e., there were no two identical breakthrough curves. Microspheres appear to be a poor analog for bacteria in this system, generally arriving at the extraction well later than all but the Gram-positive nonmotile rods. It is important to note that polystyrene latex microspheres have essentially hydrophobic surfaces, but can be made stable (nonaggregating) by amending charged functional groups (such as carboxylate or amidine groups) to their surfaces. Thus, microspheres may have fundamentally different surface properties than bacteria that are typically neither highly hydrophobic or highly charged.

Bacteria motility appeared to have a negative impact on transport (Figure 1). The motile (flagellated) Gram-negative rod (ML2) were poorly recovered with respect to its nonmotile mutant strain (ML2m). During the first 80 minutes of breakthrough, the fractional breakthrough of strains ML2 and ML2m was ~15% to 20% of the solute tracer, indicating that the fastest moving bacteria were not highly filtered. However, subsequent breakthrough indicates that the motile strain was preferentially immobilized within the aquifer relative to the nonmotile strain. Motile bacteria should be subject to a greater effective diffusion rate than nonmotile bacteria of the same size and shape. According to standard deep-bed particle filtration theory, if two populations of colloidal particles have the same propensity for attachment (stickiness coefficient), the colloids exhibiting the highest diffusion rate should be subject to the greatest filtration, all other factors being equal (Rajogopalan and Tien 1976). Another explanation for the greater filtration of the motile strain over the nonmotile strain is that the motile strain may have migrated into relatively stagnant water, thus delaying their arrival and enhancing their probability for capture. It has been hypothesized that relatively stagnant water exists in adjacent to mobile water in fractured rock systems, and that the diffusive exchange between mobile and immobile water can create a delay in tracer arrival and cause extensive breakthrough tailing (Neretnieks et al. 1982; Birgersson et al. 1993; Maloszewski and Zuber 1993; Raven et al. 1988).

Table 2

Recovery of Tracers in Field Experiments

Gram-positive rod bacteria (ML1) appeared to be preferentially filtered over the nonmotile Gram-negative rod bacteria (ML2m) as shown in Figure 1. This does not necessarily suggest that Gram staining is an indicator of the attachment affinity of the bacteria, but rather the Microbacterium (Gram-positive) appears to have a greater affinity for attachment than Pseudomonas (Gram-negative) in this system. The Gram-staining procedure relies on a reaction between the stain and the cell envelope, so Gram staining may be an indicator of bacteria-surface interaction under certain conditions. This conclusion may not be drawn from this experiment, however, and laboratory studies by others have not resulted in a consistent correlation between Gram staining and transport behavior in soil (Gannon et al. 1991).

Figure 1. Cumulative breakthrough curve for all recovered tracers (positively charged microspheres were not detected at the extraction well).

Gram-positive coccus bacteria appeared to be preferentially transported over the Gram-positive rod bacteria (Figure 1). Fractional recovery of the coccoidal bacterium (strain ML3) was 17 times higher than that of the Gram-positive rod-shaped bacterium (strain ML1), for example (Table 2). It is possible that the spheroid shape of the coccoidal bacteria reduces the probability of attachment to mineral surface. A rod-shaped bacterium would have a greater portion of its cell surface in contact with the solid surface relative to a coccoid bacterium with same cell length. In addition, rod-shaped bacteria may be more apt to overcome electrostatic repulsion from the mineral surface than spherically shaped bacteria.

It is unlikely that cell shape affected straining of bacteria. Bouwer (1984) used glass columns packed with a uniform spherical medium and found that straining occurred when the diameter of suspended particles was >20% of the diameter of the porous medium. One might expect, therefore, that water-carrying bacteria would have to pass through apertures of 5 [mu]m or less for straining to occur. According to the "cubic" law the rate of flow through a fracture varies as a function of the cube of the aperture. The mean hydraulic connecting FSE6 and FSE 9 is thought to be ~400 [mu]m (Becker and Shapiro 2000,) so that these primary fractures would carry >500,000 times the flow of a 5 [mu]m fracture. As a consequence, an insignificant fraction of the flow would pass through a fracture small enough to strain bacteria.

Explanations of the influ\ence of cell shape on transport are conjecture, in any case, as it is impossible to separate the influence of cell shape from cell wall attributes and cell size. Cell size is potentially much more important than cell shape, for example, and the coccoid bacteria were approximately half the size of the rod bacteria (Table 1). Other researchers have found that particle size can be a critical parameter in filtration rates through fractured geologic media (Cumbie and McKay 1999). Previous tracer experiments (Becker et al. 1999) conducted between these same two wells (FSE 6 and FSE 9) resulted in much better transport of microspheres that were also carboxylate modified, of the same density, and from the same production process (Interfacial Dynamics, Portland, Oregon), but of a smaller diameter (0.2 [mu]m).

In Figure 2, the breakthrough from this experiment of deuterated water, coccoidal bacteria, and 1 [mu]m diameter carboxylate- modified microspheres are compared to the breakthrough of 0.2 [mu]m diameter carboxylate-modified microspheres used in a previous tracer experiment. Breakthrough is plotted as fraction of injected mass recovered versus volume pumped from the extraction well, to account for the difference in pumping rate between the two experiments (10.7 L/min in this experiment as compared to 9.8 L/min in the previous experiment). Note that if only size is considered, smaller particles are recovered most efficiently. Approximately 19% of the 0.2 [mu]m diameter microspheres were recovered versus <1% of the 1 [mu]m diameter microspheres, for example. It should be mentioned that the previous experiment was conducted as a "weak dipole" or approximately radially convergent test, wherein 5% of the pumped water was reinjected. This should not have had a large influence on the results, however, based on experience at Mirror Lake with different two-well tracer test configurations (Becker et al. 1999; Becker and Shapiro 2000).

Figure 2. Cumulative breakthrough curve for tracers of different size. Breakthrough of deuterated water, coccoid bacteria, and 1 [mu]m diameter microspheres were obtained from this experiment, whereas the breakthrough of 0.2 [mu]m diameter microspheres were obtained from a previous experiment conducted between the same two wells. Recovery is seen to be inversely related to particle size.

Conclusions

Interpretations of these experimental results must be made with reference to the specific strains of bacteria used, water chemistry, and surface characteristics of the transport media. It is encouraging, however, that some of the results concur with established conceptual models of transport. Motile bacteria were subject to greater filtration, supporting the hypothesis that motile strains are more likely to be filtered due to an increased probability of collision with surfaces. Likewise, it would be expected that motile bacteria would be better able to swim into stagnant water trapped in dead-end apertures and pore spaces in fractured bedrock, again resulting in a greater probability of surface collision and attachment. Taken in the context of an in situ biological remediation effort, motile bacteria would have the advantage of being able to access contaminants trapped in dead-end pores, but would have the disadvantage of migrating more slowly into the contaminated area.

The implication of these experiments with respect to cell physicochemical properties is less clear. For example, Gram- negative rod-shaped bacteria (ML2m) migrated more efficiently than Gram-positive bacteria. Although the difference in transport was striking, Gram staining is designed to assess the presence (Gram negative) or absence (Gram positive) of a lipopolysaccaride outer membrane, exterior to the cell wall, which may or may not affect bacteria surface interactions (Gannon et al. 1991). Consequently, we strongly caution the reader against generalizing the relative behavior of the two Gram-types observed in this experiment. The result does suggest, however, that Gram staining or other gross indicators may be beneficial for the isolation of bacteria that are more mobile in ground water. Careful laboratory research is needed in this area.

A coccus bacterial strain (ML3) appeared to be relatively mobile compared to the other bacteria, but it is not clear what characteristic of the coccus led to this behavior. Although the obvious difference between coccus (spherical) and bacillus (rod) species is their shape, it is unlikely that shape affected transport. Given the relative behavior of the other bacteria, cell wall characteristics and/or size was probably much more influential on transport. Similarly, the coccus strain responded positively to Gram staining, but the relationship between physicochemical properties of the cell envelope and filtration is indeterminate due to the influence of these other variables. The coccus isolate could only be cultured to a size about half that of the other strains and microspheres used in the experiments. Comparison of microspheres of different sizes in the same formation suggests that size is an extremely important factor in transport, at least in the 0.2 to 1 micron diameter range.

These experiments demonstrate that small difference in bacteria properties such as cell size, morphology, Gram type, and motility can lead to large differences in transport efficiency. It is unreasonable, therefore, to propose that the transport of one bacterial species will be a confident predictor of other bacteria species in this fractured crystalline rock. Similarly, the transport behavior of colloids such as microspheres are expected to be a poor predictor of bacteria species. If this single field test can be extrapolated to other structured and unstructured geologic media, the results suggest that mathematical models of transport will require multiple parameters describing bacterial cell physicochemical attributes if they are to reliably predict bacteria migration in ground water. It is extremely difficult, if not impossible, to impose the necessary experimental control to determine these multiple parameters in the field. We would argue that theoretical, laboratory, and field-transport experiments will need to be conducted in parallel to advance understanding of bacteria transport in ground water.

Acknowledgments

These investigations were conducted with the support of the Toxic Substances Hydrology and National Research Programs administered by the U.S. Geological Survey. We gratefully acknowledge the kind cooperation and support of the Hubbard Brook Experimental Forest, which is operated and maintained by the Northeastern Forest Experiment Station, USDA Forest Service, Radnor, Pennsylvania. We thank Jim Schuetz for his aid with field operations. The manuscript was greatly improved as a result of the thoughtful comments provided by Isabelle Cozzarelli and Martha Scholl for the U.S. Geological Survey, and Walton Kelly, Phil Oberlander, and an anonymous reviewer for Ground Water.

Vol. 41, No. 5-GROUND WATER-September-October 2003 (pages 682- 689)

References

Becker, M.W., and T.B. Coplen. 2001. Deuterated water as a conservative artificial ground-water tracer. Hydrogeology Journal 9, no. 5: 512-516.

Becker, M.W., P.W. Reimus, and P. Vilks. 1999. Transport and attenuation of carboxylate-modified latex microspheres in fractured rock laboratory and field tracer tests. Ground Water 37, no. 3: 387- 395.

Becker, M.W., and A.M. Shapiro. 2000. Tracer transport in fractured crystalline rock: Evidence of nondiffusive breakthrough tailing. Water Resources Research 36, no. 7: 1677-1686.

Birgersson, L., L. Moreno, I. Neretnieks, H. Widen, and T. Agren. 1993. A tracer migration experiment in a small fracture zone in granite. Water Resources Research 29, no. 12: 3867-3878.

Bouwer, H. 1984. Elements of soil science and groundwater hydrology. In Groundwater Pollution Microbiology, ed. C.P. Gerba. New York: Wiley.

Champ, D.R., and J. Schroeter. 1988. Bacterial transport in fractured rock: A field-scale tracer test at the Chalk River Nuclear Laboratories. Water Science Technology 20, no. 11/12: 81-87.

Chen, J., and B. Koopman. 1997. Effect of fluorochromes on bacterial surface properties and interaction with granular media. Applied and Environmental Microbiology 63, no. 10: 3941-3945.

Corapcioglu, M.Y., and A. Haridas. 1984. Transport and fate of microorganisms in porous media: A theoretical investigation. Journal of Hydrology 72, no. 1-2: 149-169.

Cumbie, D.H., and L.D. McKay. 1999. Influence of diameter on particle transport in a fractured shale saprolite. Journal of Contaminant Hydrology 37, no. 1-2: 139-157.

Day-Lewis, F.D., P.A. Hsieh, and S.M. Gorelick. 2000. Identifying fracture-zone geometry using simulated annealing and hydraulic- connection data. Water Resources Research 36, no. 7: 1707-1721.

Fontes, D.E., A.L. Mills, G.M. Hornberger, and J.S. Herman. 1991. Physical and chemical factors influencing transport of microorganisms through porous media. Applied and Environmental Microbiology 57, no. 9: 2473-2481.

Fuller, M.E., S.H. Streger, R.K. Rothmel, B.J. Mailloux, J.A. Hall, T.C. Onstott, J.K. Fredrickson, D.L. Balkwill, and M.F. DeFlaun. 2000. Development of a vital fluorescent staining method for monitoring bacterial transport in subsurface environments. Applied and Environmental Microbiology 66, no. 10: 4486-4496.

Gannon, J.T., V.B. Manilial, and M. Alexander. 1991. Relationship between cell surface properties and transport of bacteria through soil. Applied and Environmental Microbiology 57, 190-193.

Harkness, M.R., A.A. Bracco, M.J. Brennan, K.A. Deweerd, and J.L. Spivack. 1999. Use of bioaugmentation to stimulate complete reductive dechlorination of trichloroethene in Dover soil columns. Environmental Science and Technology 33, 1100-1109.

Harvey, R.W. 1997. Microorganisms as tracers in groundwater injection and recovery experiments: A review. FEMS Microbiology Reviews 20, 461-472.

Ha\rvey, R.W., and S.P. Garabedian. 1991. Use of colloid filtration theory in modeling movement of bacteria through a contaminated sandy aquifer. Environmental Science and Technology 25, no. 1: 178-185.

Harvey, R.W., L.H. George, R.L. Smith, and D.R. LeBlanc. 1989. Transport of microspheres and indigenous bacteria through a sandy aquifer: Results of natural- and forced-gradient tracer experiments. Environmental Science and Technology 23, no. 1: 51-56.

Harvey, R.W., N.E. Kinner, D. McDonald, D.W. Metge, and A.L. Bunn. 1993. Role of physical heterogeneity in the interpretation of small-scale laboratory and field observations of bacteria, microbial- sized microsphere, and bromide transport through aquifer sediments. Water Resources Research 29, no. 8: 2713-2721.

Harvey, R.W., D.W. Metge, N. Kinner, and N. Mayberry. 1997. Physiological considerations in applying laboratory-determined buoyant densities to predictions of bacterial and protozoan transport in groundwater: Results of in-situ and laboratory tests. Environmental Science and Technology 31, 289-295.

Harvey, R.W., R.L. Smith, and L. George. 1984. Effect of organic contamination upon microbial distributions and heterotrophic uptake in a Cape Cod, Massachusetts, aquifer. Applied Environmental Microbiology 48, 1197-1202.

Hornberger, G.M., A.L. Mills, and J.S. Herman. 1992. Bacterial transport in porous media: Evaluation of a model using laboratory observations. Water Resources Research 28, 915-938.

Johnson, C.D., and A.H. Dunstan. 1998. Lithology and fracture characterization from drilling investigations in the Mirror Lake area, Grafton County, New Hampshire. U.S. Geological Survey Toxic Substance Hydrology Program Water-Resources Investigations Report 98- 4183 98-4183: 41.

Kim, S., and M.Y. Corapcioglu. 1996. A kinetic approach to modeling mobile bacteria-facilitated groundwater contaminant transport. Water Resources Research 32, no. 2: 321-331.

Knapp, E.P., J.S. Herman, G.M. Hornberger, and A.L. Mills. 1998. The effect of distribution of iron-oxyhydroxide grain coatings on the transport of bacterial cells in porous media. Environmental Geology 33, no. 4: 243-248.

Lawrence, J.R., and M.J. Hendry. 1996. Transport of bacteria through geologic media. Canadian Journal of Microbiology 42, 410- 422.

Li, Q., and B.E. Logan. 1999. Enhancing bacterial transport for bioaugmentation of aquifers using low ionic strength solutions and surfactants. Water Research 33, 1090-1100.

Maloszewski, P., and A. Zuber. 1993. Tracer experiments in fractured rocks: Matrix diffusion and the validity of methods. Water Resources Research 29, no. 8: 2723-2735.

McKay, L.D., J.A. Cherry, R.C. Bales, M.T. Yahya , and C.P. Gerba. 1993. A field example of bacteriophage as tracers of fracture flow. Environmental Science and Technology 27, no. 6: 1075-1079.

Metge, D.W., R.W. Harvey, A.M. Shapiro, M. Becker, S. Collins, S.P. Buckwalter, R.L. Sheets, and G. Rowe. 2000. What surrogates can tell us about the transport behavior of pathogens in fractured-rock and granular aquifers. Eos Transactions AGU 81, no. 48, Fall Meeting Supplement, Abstract B51C-03.

Mills, A.L., J.S. Herman, G.M. Hornberger, and T.H. DeJesus. 1994. Effect of solution ionic strength and iron coatings on mineral grains on the sorption of bacterial cells to quartz sand. Applied and Environmental Microbiology 60, no. 9: 3300-3306.

Munakata-Marr, J., V.G. Matheson, L.J. Forney, J.M. Tiedje, and P.L. McCarty. 1997. Long-term biodegradation of trichloro-ethylene influenced by bioaugmentation and dissolved oxygen in aquifer microcosms. Environmental Science and Technology 31, 786-791.

Neretnieks, I., T. Eriksen, and P. Taehtinen. 1982. Tracer movement in a single fissure in granitic rock: Some experimental results and their interpretation. Water Resources Research 18, no. 4: 849-858.

Otto, K., H. Elwing, and M. Hermansson. 1999. The role of type 1 fimbriae in adhesion of Escherichia coli to hydrophilic and hydrophobic surfaces. Colloids and Surfaces B: Biointerfaces 15, no. 1: 99-111.

Pieper, A.P., J.N. Ryan, R.W. Harvey, G.L. Amy, T.H. Illangasekare, and D.W. Metge. 1997. Transport and recovery of bacteriophage PRD1 in a sand and gravel aquifer: Effect of sewage- derived organic matter. Environmental Science and Technology 31, no. 4: 1163-1170.

Rajogopalan, R., and C. Tien. 1976. Trajectory analysis of deep- bed filtration with the sphere-in-cell porous media model. AIChE Journal 22, no. 3: 523-533.

Raven, K.G., K.S. Novakowski, and P.A. Lapcevic. 1988. Interpretation of field tracer tests of a single fracture using a transient solute storage model. Water Resources Research 24, no. 12: 2019-2032.

Reimus, P.W. 1995. The use of synthetic colloids in tracer transport experiments in saturated rock fractures. Report no. LA- 13004-T, Los Alamos, New Mexico: Los Alamos National Laboratory.

Rijnaarts, H.H.M., W. Norde, E.J. Bouwer, J. Lyklema, and A.J.B. Zehnder. 1995a. Reversibility and mechanism of bacterial adhesion. Colloids and Surfaces B: Biointerfaces 4, no. 1: 5-22.

Rijnaarts, H.H.M., W. Norde, J. Lyklema, and A.J.B. Zehnder. 1995b. The isoelectric point of bacteria as an indicator for the presence of cell surface polymers that inhibit adhesion. Colloids and Surfaces B: Biointerfaces 4, no. 4: 191-197.

Scholl, M.A., and R.W. Harvey. 1992. Laboratory investigations on the role of sediment surface and groundwater chemistry in transport of bacteria through a contaminated sandy aquifer. Environmental Science and Technology 26, no. 7: 1410-1417.

Shapiro, A.M., and P.A. Hsieh. 1996. A new method of performing controlled injection of traced fluid in fractured crystalline rock. In U.S. Geological Survey Toxic Substances Hydrology Program Technical Meeting, ed. D.W. Morganwalp and D.A. Aronson, 131-136. Colorado Springs, Colorado: USGS.

Shapiro, A.M., and P.A. Hsieh. 1998. How good are estimates of transmissivity from slug tests in fractured rock? Ground Water 36, no. 1: 37-48.

Steffan, R.J., K.L. Sperry, M.T. Walsh, S. Vainberg, and C.W. Condee. 1999. Field-scale evaluation of in situ bioaugmentation for remediation of chlorinated solvents in groundwater. Environmental Science and Technology 33, 2271-2781.

Story, S.P., P.S. Amy, C.W. Bishop, and F.S. Colwell. 1995. Bacterial transport in volcanic tuff cores under saturated flow conditions. Geomicrobiology Journal 13, 249-264.

Truesdail, S.E., J. Lukasik, S.R. Farrah, D.O. Shah, and R.B. Dickinson. 1998. Analysis of bacterial deposition on metal (hydr) oxide-coated sand filter media. Journal of Colloid and Interface Science 203, 369-378.

van Loosdrecht, M.C., M.J. Lyklema, W. Norde, G. Schraa, and A.J.B. Zehnder. 1987. The role of bacterial cell wall hydrophobicity in adhesion. Applied and Environmental Microbiology 53, no. 8: 1893- 1897.

van Loosdrecht, M.C., M.J. Lyklema, W. Norde, G. Schraa, and A.J.B. Zehnder. 1989. Bacterial adhesion: A physicochemical approach. Microb. Ecol. 17, 1-15.

Vilks, P., and D.B. Bachinski. 1996. Colloid and suspended particle migration experiments in a granite fracture. Journal of Contaminant Hydrology 21, no. 1-4: 269-279.

Vilks, P., L.H. Frost, and D.B. Bachinski. 1997. Field-scale colloid migration experiments in a granite fracture. Journal of Contaminant Hydrology 26, no. 1-4: 203-214.

Weiss, T.H., A.L. Mills, G.M. Hornberger, and J.S. Herman. 1995. Effect of bacterial cell shape on transport of bacteria in porous media. Environmental Science and Technology 29, 1737-1740.

by Matthew W. Becker1, David W. Metge2, Samantha A. Collins1, Allen M. Shapiro3, and Ronald W. Harvey2

1Department of Geology, University at Buffalo, Buffalo, NY 14260, mwbecker@geology.buffalo.edu

2U.S. Geological Survey, 3215 Marine St., Boulder, CO 80303

3U.S. Geological Survey, 431 National Center, Reston, VA 20192

Received June 2002, accepted January 2003.

Copyright Ground Water Publishing Company Sep/Oct 2003

More News in this Category