Potential for intrinsic bioremediation of a DNT-contaminated aquifer

Potential for intrinsic bioremediation of a DNT-contaminated aquifer

Bradley, P M

Introduction

Nitroaromatic-explosives compounds like 2,4-dinitrotoluene and 2,6-dinotrotoluene (2,4-DNT and 2,6-DNT, respectively) are common contaminants of surface soils and shallow aquifers at many military sites throughout the United States (Higson, 1992). Nitroaromatic compounds are a significant environmental concern due to their toxicity to fish (Osmon and Klausmeier, 1972; Smock et al., 1976), algal species (Smock et al., 1976; Won et al., 1976), microorganisms (Won et al., 1976; Klausmeier et al., 1973), and other organisms (Won et al., 1976). A number of alternatives, including incineration, anaerobic bioslurry, and soil composting, are available for the restoration of explosives-contaminated soil [for an overview of previous research, see Spain (ed.), 1995]; however, application of these techniques to clean up nitroaromatics-contaminated aquifers is problematic. The increasing closure of military bases throughout the United States has made identification of cost-effective and environmentally acceptable methods for remediation of nitroaromatics-contaminated ground water an environmental imperative.

Public concern over environmental contamination combined with an increasing emphasis on the price of restoration has led regulatory agencies to consider intrinsic bioremediation as

remedial alternative for the cleanup of contaminated sites. Because this approach avoids many of the implementation difficulties and financial costs inherent in physical, chemical, and bioengineered methods of ground-water cleanup, intrinsic bioremediation is particularly attractive as an alternative for aquifer restoration. Although historically intrinsic bioremediation has been confined primarily to the cleanup of petroleum-contaminated aquifers (McAllister and Chiang, 1994; Salanitro, 1993), the potential for intrinsic bioremediation of ground-water contamination exists at any site characterized by an indigenous community of aquifer microorganisms capable of contaminant biodegradation.

The shallow aquifer at the inactive Weldon Spring Ordnance Works, Weldon Spring, Missouri is contaminated with 2,4-DNT and 2,6-DNT as a result of explosives manufacturing and processing activities carried out during World War II (Bradley et al., 1994). Analyses of ground-water samples collected at the site reveal that dissolved concentrations of 2,4-DNT and 2,6-DNT decrease with distance along the general path of ground-water flow (Schumacher et al., 1993). This decrease is consistent with a previous report of DNT degradation by microorganisms indigenous to the aquifer at Weldon Spring (Bradley et al., 1994) and indicates that in situ microbial activity may restrict ground-water transport of DNT at the site.

From an intrinsic bioremediation perspective, however, microbial degradation is desirable only if the compounds that accumulate during degradation are less toxic than the original contaminants or, under ideal conditions, innocuous endproducts such as CO sub 2 . Although previous research demonstrated that microorganisms indigenous to the aquifer at Weldon Spring readily transformed 2,4-DNT and 2,6-DNT to transient, monoamino-intermediates, the final products of DNT degradation by the aquifer microbial community were not identified (Bradley et al., 1994). The purpose of the studies reported here was to evaluate the potential for intrinsic bioremediation of the DNT-contaminated aquifer by examining the ability of the indigenous aquifer microorganisms to completely degrade radio-labeled 2,4-DNT and 2,6-DNT to sup 14 CO sub 2 .

Site Background

During its operation between 1941 to 1945, the Weldon Spring Ordnance Works produced more than 320 million kg of 2,4,6-trinitrotoluene (TNT) and smaller quantities of DNT. Even though explosives manufacturing and handling activities ceased in 1945, recent surveys revealed widespread contamination of the surface soils and underlying aquifer by nitroaromatic compounds. Of 6,200 surface soil samples collected at the site, 15.9% contained TNT concentrations greater than 30 mg/kg dry soil (Fink, 1992). Soil 2,4-DNT concentrations as high as 3 mg/kg have been reported within contaminated areas at Weldon Spring (Schumacher et al., 1992), but, in general, TNT represents greater than 99.9% of the soil contamination at the site. In contrast, ground-water samples collected from the shallow aquifer indicate that TNT contamination of the aquifer is uncommon, but ground-water DNT contamination is widespread. The general lack of TNT contamination in the ground water is attributable to preferential microbial degradation of TNT as it slowly leaches through the overlying soil column (Schumacher unpublished data; Bradley et al., 1994; Schumacher et al., 1992).

The stratigraphy of the site has been described in detail elsewhere (Schumacher et al., 1993). Much of the site is underlain by overburden deposits of variable thickness consisting of a surface layer of relatively organic rich (0.35-1.5% dry weight), modified loess, a fractured clay till, and a residuum. Consolidated carbonated bedrock is present immediately beneath the overburden deposits and is characterized by an extensive network of fractures. Shallow ground-water flow at Weldon Spring takes place primarily within the highly permeable residuum and the fractured carbonate bedrock. The top of the water table generally is found within the residuum.

The Weldon Spring Ordnance Works is situated along an east-west trending ridge that serves as an approximate groundwater flow divide between the Mississippi River basin to the north and the Missouri River basin to the south (Figure 1). (All Figures and Tables omitted) Results of a dye tracer test indicate that ground water is not transferred between the two basins (Schumacher et al., 1993). The study site is located north of the dividing ridge, and ground water generally flows beneath the study site northeast toward a nearby spring, Burgermeister Spring. Because MW17 is located along the ridge that delineates the ground-water flow divide at this site, some uncertainty exists about the direction of groundwater flow through this well (Table 1). Although the fractured nature of the bedrock portion of the shallow aquifer makes determination of the rate of ground-water flow difficult, the time required for ground water to travel from the divide to Burgermeister Spring (about 4 km downgradient) is estimated to be between 2 to 10 years (Chapelle, unpublished data).

Recently, ground-water samples were collected from monitoring wells located along the shallow ground-water flowpath and analyzed for dissolved 2,4-DNT and 2,6-DNT (Schumacher et al., 1993). The results of these analyses are summarized in Table 1 and Figure 2. The ground-water survey (Schumacher et al., 1993) indicated that the dissolved concentrations of 2,4-DNT and 2,6-DNT decrease along the approximate ground-water flow path from 44 mu g/l and 61 mu g/l, respectively, in samples collected at MW17 near the recharge area to less than the detection limit of 0.1 mu g/l in samples collected approximately

km downgradient near Burgermeister Spring (Figure 2).

Chemicals

2,4-DNT, and 2,6-DNT were obtained from SRI International (Menlo Park, CA). 2-amino-nitrotoluene, 2-amino-6-nitrotoluene, and 4-amino-2-nitrotoluene were obtained from Aldrich Chemical Company (Milwaukee, WI). The purity of all standards was 99% or greater. Uniformly ring labeled 2,4-DNT (7.0 mCi/mmol and 2,6-DNT (2.2 mCi/mmol) were obtained from Sigma Chemical Company (St. Louis, MO). The purity of the radiolabeled substrates was determined by thin layer chromatography and high performance liquid chromatography (HPLC) and found to be greater than 98% and 97.4% for 2,4-DNT and 2,6-DNT, respectively.

Methods

The ability of aquifer microorganisms indigenous to Weldon Spring to mineralize 2,4-DNT and 2,6-DNT was investigated using aquifer material from the residuum. The aquifer material contained less than 0.01% by weight of organic carbon and carbonate carbon. The concentration of DNT sorbed onto the collected aquifer material was less than 0.01 mu moles/kg dry material. The ground water at the site was aerobic (dissolved oxygen concentrations >= 1.0 mg/l). Additional microcosms were created using contaminated surface soil collected at an abandoned munitions wash house and surface soil collected at an adjacent site with no detectable contamination and no history of nitroaromatic exposure. The TNT concentration of the contaminated surface-soil samples was approximately 0.5 mu moles/kg dry soil, and DNT concentrations were less than the analytical detection limit of 0.01 mu moles/kg dry soil.

Experimental microcosms were prepared aerobically by placing 3 g saturated aquifer material or soil in 30 ml serum vials and sealing with Teflon-coated butyl stopper/base trap assemblies as described previously (Bradley et al., 1994). The microcosms were amended with 0.5 ml of an aqueous solution containing approximately 4 X 10 sup 5 dpm of uniformly ring-labeled 2,4-DNT or 2,6-DNT. Abiological controls were prepared by adding HgCl sub 2 (5 mM) and autoclaving the microcosms (121degC for 1 h).

Microcosms were incubated statically in the dark at room temperature. For 28 days, triplicate experimental microcosms and a single killed control were periodically sacrificed to quantify the sup 14 CO sub 2 produced. Microcosms were acidified with 1000 mu l of 13 N H sub 3 PO sub 4 . Evolved sup 14 CO sub 2 was collected by placing 300 mu l of 3 N KOH in suspended base traps and shaking the acidified microcosms for 48 h. The sup 14 CO sub 2 recovered in the base solution was quantified by liquid scintillation counting. The recovery efficiency of sup 14 CO sub 2 in the sample material was determined using H sup 14 CO sub 3 . Reported values were corrected for recovery efficiency, the activity recovered at time zero, and the activity detected in sterilized control vials. The activity associated with radiolabeled volatile intermediates other than CO sub 2 was estimated using toluene as a trapping agent and found to be less than 1% of the activity recovered in the base traps.

Additional aquifer microcosms were prepared as described previously (Bradley et al., 1994) using unlabeled substrates and monitored by HPLC analysis in order to quantify changes in the concentrations of 2,4-DNT, 2,6-DNT, and their respective monoamino-nitrotoluene degradation intermediates. For all studies, statistically significant differences between treatment means were identified by analysis of variance and the Student-Newman-Kuels Multiple Comparison test (Jandel Scientific, 1994).

Results and Discussion

The results of the ground-water survey at Weldon Spring indicated that dissolved concentrations of 2,4-DNT and 2,6-DNT are reduced to below the 0.1 mu g/l detection limit before the ground water reaches the point of contact at Burgermeister Spring(Table 1 and Figure 2). The fact that DNT contamination is not detected in downgradient wells, even though the groundwater transport time is estimated to be less than 10 years and the site has been contaminated for more than 50 years suggests that some factor, such as microbial degradation, is serving to contain the transport of DNT contamination at the site. A number of factors, such as dispersion, dilution, adsorption, and biodegradation, can contribute to the natural attenuation of dissolved contaminants along a ground-water flow path, but only biodegradation results in a reduction of the contaminant mass. Several studies have reported transformation of DNT compounds by a variety of microorganisms (Bradley et al., 1994; Hallas and Alexander, 1983; Liu et al., 1984; McCormick et al., 1978; Parrish, 1977; Spanggord et al., 1991; Valli et al., 1992). Recently, significant degradation of DNT by aquifer micoorganisms was demonstrated at Weldon Spring (Bradley et al., 1994).

These observations have raised the possibility that intrinsic bioremediation may be a realistic option for restoring DNT-contaminated ground water at Weldon Spring following the removal of the surface contamination. However, the final products of DNT degradation by the aquifer microorganisms at the site have not been identified (Bradley et al., 1994). From an intrinsic bioremediation perspective, microbial degradation is desirable only if the compounds that accumulate during degradation are less toxic than the original contaminants or, under ideal conditions, innocuous endproducts such as CO sub 2 . The results of previous investigations of microbial DNT degradation indicate that the ability to mineralize DNT is not widespread (Bradley et al., 1994; Hallas and Alexander, 1983; Liu et al., 1984; McCormick et al., 1978; Parrish, 1977; Spanggord et al., 1991; Valli et al., 1992). Among the many organisms found capable of DNT degradation (Bradley et al., 1994; Hallas and Alexander, 1983; Liu et al., 1984; McCormick et al., 1978; Parrish, 1977; Spanggord et al., 1991; Valli et al., 1992), only the white rot fungus, Phanaerochaete chrysoporium, has demonstrated significant mineralization of DNT to CO sub 2 (Valli et al., 1992).

The microbial community indigenous to the shallow aquifer at Weldon Spring readily degraded 2,4-DNT (Figure 3). The concentration of 2,4-DNT in the aquifer microcosms was reduced by about 80% within 30 days. The reduction in 2,4-DNT concentration observed in the experimental microcosms was attributed to biological activity, because no significant decrease in concentration occurred in sterilized control treatments. The decrease in the concentration of 2,4-DNT was accompanied by accumulation of monoamino-nitrotoluene compounds. 2,4-DNT was transformed, primarily, to 4-amino-2-nitrotoluene (22.3% final concentration) and, to a lesser extent, 2-amino-4-nitrotoluene (6.3% final concentration), as described previously (Bradley et al., 1994; Liu et al., 1984; McCormick et al., 1978).

The microorganisms indigenous to the shallow aquifer at Weldon Spring mineralized a significant fraction of added [U-ring- sup 14 C]2,4-DNT to sup 14 CO sub 2 within 28 days (Figure 4, Table 2). Approximately 28% of the 2,4-DNT radiolabel was recovered as

approximately 20% of the added 2,4-DNT remained undegraded` while approximately 22% and 6% were transformed to` 4-amino-2-nitrotoluene and 2-amino-4-nitrotoluene,` respectively.` Several unidentified, transient-intermediates which were` detected` only after the onset of monoamino-nitrotoluene accumulation` represent the remaining 24% of the added 2,4-DNT.` ` A similar pattern of transformation and subsequent` mineralization` was observed in aquifer microcosms containing 2,6-DNT;` however, degradation of 2,6-DNT proceeded more slowly` than degradation of 2,4-DNT (Figures 3 and 4). At the end of` the` incubation, approximately 67% of the 2,6-DNT remained` undegraded` and approximately 14% was transformed to` 2-amino-6-nitrotoluene.` In the radiolabeled microcosms only 8% of the` [ sup 14 C]2,6-DNT radiolabel was recovered` as sup 14 CO sub 2 (Figure 4,` Table 2). The remaining 11% of the degraded 2,6-DNT was` transformed to several unidentified, transient intermediates` which were detected only after the onset of` monoamino-nitrotoluene` accumulation. The difference in the efficiency of` substrate transformation and mineralization observed between` 2,4-DNT and 2,6-DNT microcosms may be due to in situ` acclimation` to 2,4-DNT. At the Weldon Spring site, 2,4-DNT is a` significant component of the surface contamination which is` the probable source for the ground-water contamination` (Schumacher et al., 1993; Schumacher et al., 1992). By` comparison,` 2,6-DNT is found only in trace quantities at the site` (Schumacher et al., 1993; Schumacher et al., 1992).` ` Significant mineralization of 2,4-DNT and 2,6-DNT also` was observed in the microcosms containing surface soil` collected` at the Weldon Spring site (Table 2). Microcosms containing` uncontaminated surface soil mineralized 2,4-DNT and 2,6-DNT` with recoveries comparable to those of aquifer microcosms. The` microbial communities indigenous to the contaminated surface` soil at Weldon Spring did not mineralize 2,4-DNT or 2,6-DNT` readily (Table 2). Soil DNT concentrations as high as 3 mg/kg` have been reported within contaminated areas at Weldon Spring` (Schumacher et al., 1992); however, TNT represents greater than` 99.9% of the soil contamination at the site. The contaminated` soil used in this study contained detectable concentrations of` TNT (about 0.5 mu moles/kg dry soil). Although TNT` concentrations` were not monitored in this study, the potential exists that` the contaminated-soil microbial community preferentially` utilized` TNT over DNT. Transformation and subsequent` mineralization` of TNT by the microorganisms indigenous to the` contaminated` soil at Weldon Spring have been reported previously` (Bradley et al., 1994).` ` The results of the experiments presented here provide the` first direct evidence for mineralization of 2,4-DNT and 2,6-DNT` by aquifer microorganisms. As such, the results have important` implications for the remediation of DNT-contaminated ground` water at this and other nitroaromatic-contaminated sites.` Intrinsic` bioremediation, using the ability of indigenous` microorganisms` to mineralize contaminants as a remediation technology, is` a fairly recent approach to site cleanup (McAllister and` Chiang,` 1994; Salanitro, 1993). To date, however, this approach has` been` restricted primarily to petroleum hydrocarbons that are known` to be easily biodegraded (McAllister and Chiang, 1994;` Salanitro,` 1993). Our results, which show rapid mineralization of` radio-labeled` 2,4-DNT and 2,6-DNT to sup 14 CO sub 2 ,` demonstrate that the` indigenous microorganisms provide a significant degradative` capacity for DNT-contaminated ground water at this site. This,` in turn, suggests that intrinsic bioremediation may be a viable` remediation alternative for nitroaromatic-contaminated ground` water, and indicates that this technology may be applicable to` ` wider range of chemical contaminants than has been previously` considered.` ` Acknowledgments` ` This research was supported by the U.S. Army Corps of` Engineers, Kansas City District.` ` References` ` Bradley, P. M., F. H. Chapelle, J. E. Landmeyer, and J. G.` Schumacher 1994. Microbial transformation of` nitroaromatics` in surface soils and aquifer materials. Appl. Environ.` Microbiol. v. 60, pp. 2170-2175.` ` Fink, S. A. 1992. Uptake of nitroaromatic compounds by Weldon` Spring soils. Thesis, Univ. of Missouri, Rolla. 71 pp.` ` Hallas, L. E. and M. Alexander. 1983. Microbial transformation` of` nitroaromatic compounds in sewage effluent. Appl. Environ.` Microbiol. v. 45, pp. 1234-1241.` ` Higson, F. K. 1992. Microbial degradation of nitroaromatic` compounds.` In: S. L. Neidleman and A. I. Laskin (eds.), Advances` in Applied Microbiology. Academic Press, Inc., New York. pp.` 1-19.` ` Klausmeier, R. E., J. L. Osmon, and D. R. Walls. 1973. The` effect of` trinitrotoluene on microorganisms. Dev. Ind. Microbiol. v.` 15,` pp. 309-317.` ` Liu, D., K. Thomson, and A. C. Anderson. 1984. Identification` of` nitroso compounds from biotransformation of` 2,4-dinitrotoluene.` Appl. Environ. Microbiol. v. 47, pp. 1295-1298.` ` McAllister. P. M. and C. Y. Chiang. 1994. A practical approach` to` evaluating natural attenuation of contaminants in ground` water. Ground Water Monitor. Rev. Spring, pp. 161-173.` ` McCormick, N. G., J. H. Cornell, and A. M. Kaplan. 1978.` Identification` of biotransformation products from 2,4-dinitrotoluene.` Appl. Environ. Microbiol. v. 35, pp. 945-948.` ` Osmon, J. L. and R. E. Klausmeier. 1972. The microbial` degradation` of explosives. Dev. Ind. Microbiol. v. 14, pp. 247-252.` ` Parrish, F. W. 1977. Fungal transformation of` 2,4-dinitrotoluene and` 2,4,6-trinitrotoluene. Appl. Environ. Microbiol. v. 34, pp.` 232-233.` ` Salanitro, J. P. 1993. The role of bioattenuation in the` management of` aromatic hydrocarbon plumes in aquifers. Ground Water` Monitor. Rev. Fall.` ` Schumacher, J. G., C. E. Lindley, and F. S. Anderson. 1992.` Migration` of nitroaromatic compounds in unsaturated soil at the` abandoned Weldon Spring Ordnance Works, St. Charles` County, Missouri. Proceedings from the 16th annual Army` Research and Development Symposium. CETHA-TS-CR-92062.` pp. 173-192.` ` Schumacher, J. G., S. J. Sutley, and J. D. Cathcart. 1993.` Geochemical` data for the Weldon Spring training area and vicinity` property,` St. Charles County. Missouri-1990-92. U.S. Geological` Survey Open-File Report 93-153. 86 pp.` ` SigmaStat User’s Manual. 1992. Jandel Scientific, San Rafael,` CA.` ` Smock, L. A., D. L. Stoneburner, and J. R. Clark. 1976. The` toxic` effects of trinitrotoluene (TNT) and its primary degradation` products on two species of algae and the fathead minnow.` Water Res. v. 10, pp. 537-543.` ` Spain, J. E. (ed.) 1995. Biodegradation of Nitroaromatic` Compounds.` Plenum Press, New York. 197 pp.` ` Spanggord, R. J., J. C. Spain, S. F. Nishino, and K. E.` Mortelmans.` 1991. Biodegradation of 2,4-dinitrotoluene by a Psuedomonas` sp. Appl. Environ. Microbiol. v. 57, pp. 3200-3205.` ` Valli, K., B. J. Brock, D. K. Joshi, and M. H. Gold. 1992.` Degradation` of 2,4-dinitrotoluene by the lignin-degrading fungus` Phanaerochaete` chrysosporium. Appl. Environ. Microbiol. v. 58, pp.` 221-228.` ` Won. W. D., L. H. DiSalvo, and J. Ng. 1976. Toxicity and` mutagenicity` of 2,4,6-trinitrotoluene and its microbial metabolites. Appl.` Environ. Microbiol. v. 31, pp. 576-580.` ` by P. M. Bradley(a), F. H. Chapelle(a), J. E. Landmeyer(a),` and J. G. Schumacher(b)` ` a U.S. Geological Survey-Water Resources Division, Stephenson` Center-Suite 129, 720 Gracern Rd., Columbia, South Carolina` 29210-7651.` (Bradley is corresponding author, Tel: 803-750-6100, Fax:` 803-750-6181,` E-mail: pbradleydsccmb.er.usgs.gov.)` ` b U.S. Geological Survey-Water Resources Division, 1400` Independence` Road, MS 200, Rolla, Missouri 65401.` ` Received June 1995, revised January 1996, accepted February` 1996.`

Copyright Ground Water Publishing Company Jan/Feb 1997

Provided by ProQuest Information and Learning Company. All rights Reserved