Biodegradation of Chlorinated Ethenes at a Karst Site in Middle Tennessee

Water-Quality Data Collection

Much of the water-quality data was already available from TDEC-DSF files (site 59-502), including chlorinated-ethene, ethene, and ethane data. Chlorinated-ethene data (TCE and byproducts of TCE degradation) were available for water samples collected quarterly between 1985 and 1998 from shallow and deep monitoring wells. Ethene and ethane (byproducts of VC degradation) data were available for water samples collected from selected shallow and deep wells beginning in 1997. These data included analytical printouts, information on analytical methods, quality control, and quality assurance data. The data were tabulated and examined, and changes in the analytical laboratory methods used were noted.

Additional data-collection activities conducted by the USGS included periodic water-quality sampling of selected shallow and deep monitoring wells, collection of water samples from selected deep wells for bacteria identification and enumeration, and continuous monitoring of water quality and water levels for selected deep wells. Water-quality samples were collected by the USGS during seven sampling events: November 1996; February, August, and November 1997; and February, April, and May 1998. Field measurements of temperature, pH, specific conductance, alkalinity, and dissolved oxygen (DO) were conducted using methods described by Wood (1981) and Hach Company (1992). Nutrient, major anion and cation, and total organic carbon analyses were conducted at the USGS regional laboratory in Ocala, Fla., using methods described by Fishman and Friedman (1989). Nutrient and major anion and cation analyses for samples collected during 1998 were conducted using spectrophotometric methods described by Hach Company (1992).

Continuous monitoring of water quality and water levels was conducted from March through May 1998. Data were collected at 15-minute intervals from four wells (1D, 2D, 3D, and 12D). The water-quality monitors located in the screened intervals of wells near water-producing zones measured conductivity, temperature, pH, ORP, and DO. The various sensors in the monitors were calibrated prior to placement in the wells and included rapid-pulse DO sensors that do not require stirring of the water in a well. Water-quality and water-level monitors were checked and calibrated every 2 to 3 weeks. Precipitation data (15-minute intervals) were obtained from a USGS gaging station approximately 1.5-km southeast of the study site. Precipitation data (daily totals) collected by employees at the site were obtained to verify that the data collected from the USGS gaging station were consistent with data from the site.

Geochemical Indicators

Geochemical data were used as indicators of terminal electron acceptor processes occurring in ground water at the site (Bouwer, 1994; Chapelle and others, 1995; Wiedemeier and others, 1998). The geochemical data used included DO, NO₃^-, NH₃, Mn²⁺, Fe²⁺, SO₄^2-, and S^2- concentrations (table 4). Calcium carbonate saturation indexes were calculated using methods described by Eaton and others (1995). Field measurements such as pH, alkalinity, specific conductance, temperature, and ORP were used to provide additional information concerning biodegradation processes and changes in water chemistry associated with the site hydrology.

Due to the complex ground-water flow paths in karst, geochemical isopleth maps may be difficult to construct in karst aquifers and the identification of discrete oxidation-reduction zones along a contaminant-plume transect is impractical (Chapelle and others, 1995). Thus, the geochemical data used to identify oxidation-reduction trends must be presented some other way. Radial diagrams can be used to identify spatial and temporal trends for geochemical indicators of biodegradation (Carey, 1998). Radial diagrams were used to illustrate oxidation-reduction conditions by arranging the axes of the diagrams in the sequential order of preferred electron acceptors (DO, NO₃^-, Mn, Fe, and SO₄^2-). Axis ranges were based on concentrations considered indicative of specific geochemical conditions (Wiedemeier and others, 1998). The axes for Mn²⁺ and Fe²⁺ were oriented with concentrations increasing toward the origin since these compounds are byproducts of Mn⁴⁺ and Fe³⁺ reduction (fig. 10).

Chlorinated-Ethene and Degradation Product Data

The chlorinated ethene released at this site was TCE. Degradation byproducts of TCE include DCE, VC, ethene, and ethane. Decreases in parent compounds and detection of chlorinated-ethene degradation byproducts may provide evidence that biodegradation has occurred (table 5). During reductive dechlorination, all three isomers of DCE can be produced; however, cDCE is the most commonly produced isomer and 1,1-DCE is the least commonly produced. Chlorinated-ethene, ethene, and ethane data for water samples from shallow and deep wells were converted to micromolar concentrations and examined for spatial and temporal trends that could indicate the occurrence of reductive dechlorination. Aerobic degradation of chlorinated ethenes often results in intermediate byproducts that spontaneously degrade into a variety of compounds. Intermediate byproducts of aerobic degradation, such as TCE-epoxide and chloroacetic acids, were not examined during this study.

Microcosm Experiments

In addition to the field evidence, laboratory microcosms were used to determine if microorganisms at the site had the potential to degrade contaminants. A standard microcosm method does not exist; however, a variety of methods have been reported. In many of the studies, solid and liquid aquifer materials were placed in the microcosms (Hopkins and others, 1993; Chapelle and others, 1996; Wilson and others, 1996). Other studies have used only ground water from contaminated sites in the microcosms (Nelson and others, 1988; Krumme and others, 1993; Moran and Hickey, 1997).

Batch-type microcosms containing only water were used during this study because they may be more representative of conditions in karst aquifers. Since preliminary field data indicated the occurrence of aerobic and anaerobic conditions at the site, microcosms were set up under aerobic conditions to monitor for the occurrence of cometabolic degradation pathways. After depletion of DO, the microcosms were monitored for anaerobic degradation (reductive dechlorination). Water samples collected from wells 1D, 2D, 3D, and 12D on February 17, 1998 and May 21, 1998 were used to construct microcosms during two experiments. Four replicates were established for each treatment, and control procedures included sterilization of selected microcosms (table 6). During experiment 1, bacteria counts indicated that the sterilization procedure used for controls (experiment 1, treatment 9) were not successful. The control microcosms were cross-contaminated with bacteria from the other microcosms during the addition of TCE. The second microcosm experiment was set up to establish new sterile controls and define the loss of TCE due to abiotic factors.

Microcosms were set up in 40-mL glass vials equipped with Teflon-lined septum caps then stored inverted in the dark until the selected sampling dates. On the sampling dates, samples from the microcosms were analyzed for chlorinated ethenes using gas chromatography (GC) with hexane extractions and an electron capture detector. Additional chlorinated ethene analyses were performed using GC with headspace analysis and a photoionization detector.

Bacteria Identification and Enumeration

The geochemical information described earlier in this report helped to determine if conditions were suitable for biodegradation of chlorinated ethenes and to tentatively identify microbial processes in the karst aquifer. Other lines of evidence such as laboratory microcosm data and bacteria identification and enumeration strengthen the interpretation of the geochemical information. Bacteria in water samples collected on August 20 and 21, 1998, from selected deep wells were identified by the use of the RNA-oligonucleotide hybridization method. Facultative and aerobic heterotrophic bacteria were enumerated from microcosm samples that contained water from various wells using tryptic soy agar plate counts. This information was used to provide further evidence that sufficient bacteria capable of solvent biodegradation were present in the karst aquifer.

The RNA-oligonucleotide hybridization method is a technique that takes advantage of unique nucleotide sequences in the ribosomal RNA (rRNA) to identify the bacteria (Amann and others, 1995). This method can be used to identify groups of bacteria such as sulfate-reducing bacteria or specific genera such as Nitrosomonas sp. The identification level used in this study was general bacteria groups such as iron-oxidizers, methanotrophs, ammonia-oxidizers, and sulfate-reducers (Byl and others, 1997; Farmer and others, 1998).

Bacteria samples were collected with sterile disposable bailers. Each water sample was poured into a sterile 1-liter bottle, placed on ice, and brought back to the laboratory. Sterile polycarbonate filters with 0.22-micrometer pores were used to concentrate the bacteria in the water samples. Bacteria were washed off the polycarbonate filters using a few drops of sterile phosphate buffer (PB) solution. The bacteria were preserved using a 4-percent paraformaldehyde-PB solution. The samples were stored in solution at 4 °C until they were subjected to the hybridization tests (Braun-Howland and others, 1992).

The sequences for the RNA-oligonucleotide probes of the various bacteria groups were selected from previously published sequences (Tsien and others, 1990; Kane and others, 1993; Amann and others, 1995). Bacteria have long been recognized to play an important role in oxidizing Fe²⁺ to Fe³⁺. The reverse process, bacteria mediated reduction of Fe³⁺ to Fe²⁺, has been studied as well (Chapelle, 1993). Field microbial investigations are indicating that bacteria-catalyzed Fe³⁺ reduction and Fe²⁺ oxidation are occurring within close proximity of each other (Lovley and Phillips, 1988; Lovley and others, 1990), which implies a synergistic relation between the organisms. This potential synergistic relation is one reason why these bacteria have been difficult to isolate and culture in the laboratory. To avoid this problem, Siering and Ghiorse (1997) have developed a rRNA-targeted hybridization probe to identify iron- and manganese-oxidizing bacteria in environmental samples. The RNA-hybridization probe was used to identify these potential synergistic bacteria in ground-water samples from this site.

The oligonucleotide probes were tagged with rhodamine, fluorescein, or acridine orange dye. The probes were stored at -20 °C until used. Bacteria were prepared for hybridization by taking a 100-µL sample from the preserved cells and washing the preservative out by using a series of washes and centrifugation steps (Brawn-Howland and others, 1992) and replacing the preservative with a hybridization buffer. The bacteria cells were incubated for 30 minutes at 90 °C to relax and open the RNA coils for hybridization with the oligonucleotide probes. The fluorescent-tagged probes were added to the hybridization mixture and incubated at 37 °C for 2 hours on a gentle shaker. Probes hybridized to complimentary sequences in the bacteria. The cells were spun down in a centrifuge and resuspended in clean PB solution. The un-hybridized probes were washed away in a series of PB-solution centrifugation washes. After two washes, the cells were resuspended and filtered through a 0.2-micrometer, clear polycarbonate filter. The filter with the cells was placed directly on a slide and examined using an epifluorescent microscope (X 1,000 magnification) equipped with appropriate excitation and emission filters for the targeted dye.

The tryptic soy agar and sterile dilution buffer used to count facultative anaerobic and aerobic heterotrophic bacteria were prepared as described by Eaton and others (1995). Autoclaved glass filtration devices were used to hold 0.45-µm pore-size sterile filters. A 0.01- or 1-mL aliquot of shaken well water was transferred to 20 mL of sterile dilution buffer and drawn onto the filter as described in the membrane-filtration method (Britton and Greeson, 1989). The filters were placed on the agar plates by using sterile forceps and placed in an incubator at 35 °C. Bacteria-colony forming units were counted after 48 hours. Results were reported as colony forming units per 100 mL of water.