USGS Groundwater Information
Geochemical and Microbial Evidence of Fuel Biodegradation in a Contaminated Karst Aquifer in Southern Kentucky, June 1999
By Tom D. Byl, Gregg E. Hileman, Shannon D. Williams, and James J. Farmer
Complex hydrogeologic conditions coupled with poorly understood biodegradation processes in karst aquifers have led many to believe that the potential for natural attenuation of petroleum fuel hydrocarbons is limited. This research addressed the capacity for biodegradation processes in karst. Ground-water samples were collected for bacteria and geochemical analysis from several contaminated monitoring wells in an unconsolidated regolith and karst aquifer that had varying concentrations of dissolved fuel. Bacteria concentrations were greatest in ground-water samples containing the greatest fuel contamination. Additionally, bacteria isolated from fuel-contaminated ground-water samples readily grew in Petri dishes with dissolved gasoline fuel as the only source of food. The wells with screens intersecting less contaminated sections of the aquifer had greater dissolved oxygen concentrations (6.3 milligrams per liter) than those intersecting more contaminated sections (dissolved oxygen less than 0.1 milligrams per liter). Also, where the oxygen concentrations were diminished, geochemical evidence indicated that anaerobic processes were active. This evidence includes elevated levels of ammonia and ferrous iron in the fuel-contaminated ground-water samples. Based on these results, biodegradation of fuel constituents in the karst aquifer is indicated, and therefore, natural attenuation should not be disregarded because of preconceptions about low microbial activity in karst aquifers.
Approximately 40 percent of the United States east of the Mississippi River is underlain by various types of karst aquifers (Quinlan, 1989), and more than two-thirds of the State of Tennessee is underlain by carbonate rocks that are classified as karst (Wolfe and others, 1997). Potential sources of ground-water contamination are numerous in karst regions; however, the fate of ground-water contaminants such as petroleum hydrocarbons in karst aquifers is poorly understood because of the complex hydrology (Field, 1993) and the lack of biodegradation studies. Furthermore, ground-water models that simulate flow or predict the fate and transport of contaminants in unconsolidated, porous-media aquifers have limited application to karst aquifers. Most natural attenuation and bioremediation guidelines specify that these models are not applicable in fractured rock or karst aquifers (U.S. Environmental Protection Agency, 1997).
The U.S. Geological Survey (USGS) conducted a preliminary study to evaluate whether microbes in a karst aquifer have a capacity to biodegrade the fuel-related compounds benzene, toluene, ethyl-benzene, and xylenes (BTEX). This report presents field and laboratory data collected to determine the potential for biodegradation of jet-fuel contamination in a karst aquifer in Southern Kentucky. The study site was selected because of the known presence of BTEX in the ground water and the availability of hydrologic and historic water-quality data. The objective of the study was to characterize the microbial population and geochemical conditions present in ground water at the study site to determine if BTEX biodegradation in the karst aquifer was significant enough to warrant further investigation.
METHODS AND MATERIALS
Ground-water samples were collected from seven monitoring wells located within a 5-hectare study area with a known jet-fuel spill, north of Clarksville, Tennessee, in southern Kentucky (fig.1). The wells were selected based on depths of screens and recent reports of fuel contaminant concentrations. The goal was to examine the geochemistry and bacteria present under a broad range of hydrogeologic (fig. 1) and contaminant conditions. The monitoring wells selected were screened in regolith, epikarst, and bedrock.
Water-quality samples were collected by the USGS during June 1999. Ground-water samples were collected from the wells using a submersible pump. Wells were micropurged as according to Puls and Paul (1995). The pump was lowered to the middle of the screen before starting. Water was pumped at a rate of approximately 1 liter per minute. An electronic tape measure was used to concurrently monitor the water level in the well casing. Water recharging the well was considered to be equal to the water being pumped when the water level became steady. At the point when water recharge equaled pump discharge and the specific conductance, turbidity, and fluorescence values in the pumped water no longer changed over a 1-minute period, water samples were collected. The first ground-water samples collected were for volatile organic compounds, followed by bacteria, and geochemical analyses respectively. Bacteria growth tests were begun within 6 hours of sample collection. Dissolved oxygen, alkalinity, carbon dioxide, pH, specific conductance, and temperature were measured immediately upon sample collection (Wood, 1981). The other ground-water samples were preserved according to Wood (1981) and the Hach Company (1992), and were analyzed within the holding period. The water sample for BTEX analysis was collected in three 40-milliliter (mL) vials and analyzed by gas chromatography at the USGS, National Water Quality Laboratory in Arvada, Colo. Analyses for nutrients and geochemical constituents were conducted using spectrophotometric methods described by Hach Company (1992), and, by Byl and Williams (2000).
Bacteria were enumerated and identified in ground-water samples from the monitoring wells to supplement the interpretation of the geochemical information. Bacteria were enumerated from four of the seven monitoring wells using two methods (data from the remaining three wells were not used because the samples were not processed within 24 hours of collection). The facultative and aerobic heterotrophic bacteria were enumerated using dilution buffer and tryptic soy agar plate counts as described by Eaton and others (1995). Aerobic and facultative bacteria capable of using gasoline as their sole food source also were enumerated using sterile media pads soaked with dilution buffer containing dissolved gasoline. Sterile filters with 0.45-micrometer pore sizes were used to inoculate the growth media. A 0.01- or 1-mL aliquot of sample 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 growth media using sterile forceps, and the plates were placed in an incubator at 35 °C for 5 days. Bacteria colony-forming units were counted after 24, 48, and 120 hours. Results are reported as colony-forming units per milliliter of sample water using data from the 48-hour count (table 1).
Bacteria types present in the samples were identified by the use of the ribonucleic (RNA)-oligonucleotide hybridization method (Amann and others, 1995; Byl and others, 1997). The RNA-oligonucleotide hybridization method is a technique that exploits unique nucleotide sequences in the ribosomal RNA (rRNA) to identify bacteria (Amann and others, 1995). The method can be used to identify groups of bacteria such as sulfate-reducing bacteria or specific genera such as Nitrosomonas sp. The level of identification 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).
RESULTS AND DISCUSSION
Results of the analyses demonstrated geochemical conditions that are indicative of fuel biodegradation in contaminated wells. A strong inverse pattern was found between dissolved oxygen (DO) levels and BTEX concentrations in the water samples, implying a biological oxygen demand resulting from BTEX consumption (tables 1 and 2). In addition, a great quantity of bacteria was present in each of the samples, with the largest populations associated with the most contaminated wells. Subsurface bacteria known to influence BTEX degradation processes were identified using RNA-hybridization probes (Farmer and others, 1998, Chapelle, 1993).
The bacteria identified were Pseudomonas sp., Pseudomonas aureginosa, and two groups of bacteria known as ammonia-oxidizing and iron-oxidizing bacteria. Pseudomonas bacteria, one of the most common soil bacteria, contain a variety of enzymes that degrade BTEX, such as toluene di-oxygenase and catechol di-oxygenase (Chapelle, 1993). Pseudomonas aureginosa is a specific species in the Pseudomonas family that has been shown to be important in BTEX biodegradation (Houghton and Shanley, 1994). The number of Pseudomonas aureginosa generally increased with increasing BTEX concentrations in the samples, indicating they were flourishing in the contaminated waters.
Ammonia-oxidizing bacteria also were identified in samples from four of the wells. These bacteria types have an enzyme called ammonia mono-oxygenase that can degrade BTEX in the presence of DO (Bedard and Knowles, 1989). Bacteria associated with iron geochemistry also have been shown to degrade BTEX (Chapelle, 1993). Only one highly contaminated well (CAAF-15) tested positive for iron-oxidizing bacteria. Elevated concentrations of dissolved iron in wells CAAF-2 and MCI-1 indicated that iron-reducing bacteria were active in other parts of the aquifer. In general, a consortium of aerobic and anaerobic bacteria appeared to be present in water samples from the contaminated wells implying that environmental conditions change in the aquifer or that heterogeneous niches exist within the karst aquifer with regard to redox conditions.
Perhaps the most direct evidence that bacteria in the karst aquifer at this site are capable of degrading fuel came when bacteria were grown directly on gasoline as the sole food source. Filters containing indigenous bacteria from the aquifers were placed on sterile media pads soaked with dilution buffer (Eaton and others, 1995) with gasoline dissolved in it. The control tests contained sterile media pads soaked with dilution buffer only, and no gasoline. Results of these tests are shown by the number of colony-forming units (CFU) per milliliter that grew on the pads (table 1). Bacteria in water samples from well MCI-1, a karst bedrock aquifer well, grew rapidly on the gasoline-amended media, while nothing grew on the filters placed on the control pads containing dilution buffer only, indicating the bacteria used the gasoline as a food source to grow and replicate. Bacteria from three other well samples, CAAF-8B, CAAF-10, and MCI-4, also grew on gasoline-amended media, but not on the control pads soaked with only dilution buffer. Samples from wells CAAF-2, CAAF-6, and CAAF-15 were not tested for growth on gasoline because of time constraints.
Results from these tests indicate several preliminary lines of evidence exist that bacteria are actively degrading fuels in the karst aquifer at the site. The geochemical evidence includes such indicators as oxygen consumption, ammonia production, iron dissolution, and sulfur reduction as bacteria use these constituents in their metabolic processes (table 2). The bacteria evidence includes identification of bacteria known to consume BTEX, high bacteria counts, and growth of indigenous bacteria using fuel as their only food source. These data demonstrate that bacteria thrive in water from both the deep and shallow wells at the site. Based on these results, biological degradation of BTEX is believed to occur in both unconsolidated regolith and bedrock parts of the fuel-contaminated aquifer.
Implications of the findings
The concentrations of bacteria present in the bedrock aquifer samples were high enough to be comparable with concentrations reported for contaminated sand aquifers (Ghiorse and Wilson, 1988). Thus, some of the fuel biodegradation utilization capacity derived for sand aquifers may apply to ground water in regolith, epikarst, and karst aquifers. A stoichiometric balance of terminal electron acceptors utilized in the consumption of BTEX by bacteria is used to derive a utilization capacity (Newell and others, 1996). Comparing geochemical constituents in relatively clean wells to concentrations in more contaminated wells derives the estimated concentration of terminal electron acceptors in an aquifer. Using the methods and stoichiometric equations found in the U.S. Environmental Protection Agency model BIOSCREEN: Natural Attenuation Decision Support System (Newell and others, 1996), the biodegradation capacity would be nearly 4 and 3 milligrams BTEX per liter of water based on the concentrations of DO, nitrate (NO3), ferrous iron (Fe2+) and sulfate (SO42-) found in samples from the regolith and bedrock aquifers, respectively (table 3). These results imply that BTEX dissolved in the regolith and karst aquifers are likely to be consumed by bacteria given sufficient time (the highest BTEX concentrations identified in the karst aquifer have been less than 1 mg/L). Note that the biodegradation capacity does not provide a rate of biodegradation, so factors such as ground-water flow velocity and contaminant retention time in the aquifer are important unknown variables. A study that integrates a more detailed assessment of the hydrogeology, along with further geochemistry and biology would be required to determine if BTEX biodegradation occurs at a rate sufficient to protect target receptors.
Amann, R.I., Ludwig, W., and Schleifer, K.H., 1995, Phylogenetic identification and in situ detection of individual microbial cells without cultivation: Microbiological Reviews, v. 59, no. 1, p. 143-169.
Bedard, C., and Knowles, R., 1989, Physiology, biochemistry, and specific inhibitors of CH4, NH4+, and CO oxidation by methanotrophs and nitrifiers: Microbiology Review, v. 53, no. 1, p. 68-84.
Britton, L.J., and Greeson, P.E., eds., 1989, Methods for collection and analysis of aquatic biological and microbiological samples: Techniques of Water-Resources Investigations of the U.S. Geological Survey, Book 5, chapter A4, 363 p.
Byl, T.D., and Williams, S.D., 2000, Biodegradation of Chorinated ethenes at a karst site in Middle Tennessee. U.S. Geological Survey Water-Resources Investigations Report 99-4285, 58 pages. Also available at http://pubs.water.usgs.gov/wri994285/
Byl, T.D., Farmer, J.J., Williams, S.D., Wolfe, W.J., and Bailey, F.C., 1997, Identification of bacteria in contaminated ground water to determine the feasibility of intrinsic remediation [abs.], in Bridging the Global Environment-Technology, Communication, and Education, 18th, San Francisco, 1997: Society of Environmental Toxicology and Chemistry, p. 312.
Chapelle, F.H., 1993, Ground-water microbiology and geochemistry: New York, N.Y., John Wiley and Sons, Inc., 424 p.
Eaton, A.D., Clesceri, L.S., Greenberg, A.E., and Branson, M.A.H., eds., 1995, Standard methods for the examination of water and wastewater (19th ed.): Washington, D.C., American Public Health Association, 1268 p.
Farmer, J.J., Byl, T.D., Williams, S.D., and Bailey, F.C., 1998, Identification of bacteria in contaminated ground water using the RNA-hybridization technique [abs.], in The Natural Connection-Environmental Integrity and Human Health, 19th, Charlotte, N.C.: Society of Environmental Toxicology & Chemistry, p. 3.
Field, M.S., 1993, Karst hydrology and chemical contamination: Journal of Environmental Systems, v. 22, no.1, p. 1-26.
Ghiorse, W.C., and Wilson, J.T., 1988, Microbial ecology of the terrestrial subsurface: Advances in Applied Microbiology, v. 33, p. 107-172.
Hach Company, 1992, Hach water analysis handbook: Loveland, Colo., Hach Company, 831 p.
Houghton, J.E., and Shanley, M.S., 1994, Catabolic potential of Pseudomonads: A regulatory perspective, in Chaundry, G.R., ed., Biological degradation and bioremediation of toxic chemicals: Portland, Ore., Dioscorides Press, p. 11-32.
Newell, C.J., McLeod, R.K., Gonzales, J.R., and Wilson, J.T., 1996, Bioscreen: Natural attenuation decision support system user's manual, Version 1.3 (p. 4-11), EPA/600/R-96/087: Office of Research and Development, Washington, D.C. 65 pages. Also available at www.epa.gov/ada/download/models/bioscrn.pdf (1.3MB PDF)
Puls, R.W., and Paul, C.J., 1995, Low-flow purging and sampling of ground water monitoring wells with dedicated systems: Ground Water Monitoring & Remediation v. 15, no.1, p. 116-123.
Quinlan, J.F., 1989, Ground-water monitoring in karst terranes: recommended protocols and implicit assumptions: Las Vegas, Nev., U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory, EPA/600/X-89/050, 100 p.
U.S. Environmental Protection Agency, Region 4, 1997, Draft EPA Region 4 Suggested practices for evaluation of a site for natural attenuation (biological degradation) of chlorinated solvents, Version 3.0: Atlanta, Ga., U.S. Environmental Protection Agency, Region 4, 41 p.
Wolfe, W.J., Haugh, C.J., Webbers, Ank, and Diehl, T.H., 1997, Preliminary conceptual models of the occurrence, fate, and transport of chlorinated solvents in karst aquifers of Tennessee: U.S. Geological Survey Water-Resources Investigations Report 97-4097, 80 p.
Wood, W.W., 1981, Guidelines for the collection and field analysis of ground-water samples for selected unstable constituents (2d ed.): Techniques of Water-Resources Investigations of the U.S. Geological Survey, Book 1, chap. D2, 24 p.
In Eve L. Kuniansky, editor, 2001, U.S. Geological Survey Karst Interest Group Proceedings, U.S. Geological Survey Karst Interest Group Proceedings, St. Petersburg, Florida, February 13-16, 2001: USGS Water-Resources Investigations Report 01-4011
The use of firm, trade, and brand names in this report is for identification purposes only and does not consitute endorsement by the U.S. Government.