Uptake and Transformation of Soil [14

Uptake and Transformation of Soil [14...
0 downloads 0 Views 388KB Size
Environ. Sci. Technol. 2010, 44, 6325–6330

Uptake and Transformation of Soil [14C]-Trinitrotoluene by Cool-Season Grasses J E N N I F E R M . D U R I N G E R , * ,† A. MORRIE CRAIG,‡ DAVID J. SMITH,§ AND RUFUS L. CHANEY| Department of Environmental and Molecular Toxicology, Oregon State University, 139 Oak Creek Building, Corvallis, Oregon 97331, College of Veterinary Medicine, Oregon State University, Corvallis, Oregon 97331, Animal MetabolismAgricultural Chemicals Research Unit, USDA ARS, Fargo, North Dakota 58105-5674, and USDA Environmental Management and By-Product Utilization Laboratory, Building 007, 10300 Baltimore Avenue, BARC-West, Beltsville, Maryland 20705

Received December 3, 2009. Revised manuscript received July 7, 2010. Accepted July 14, 2010.

This study investigated the fate and uptake of [14C]-TNT from soil into orchardgrass (Dactylis glomerata), perennial ryegrass (Lolium perenne), and tall fescue (Festuca arundinacea) over a one year period in a greenhouse-controlled environment. Pots (n ) 4 for each grass, containing 10 mg cold TNT/kg soil + 1.2 mg [14C]-TNT/kg soil and controls with no TNT) were exposed to light and temperature conditions typical of June at 45°N for 369 days. Three plant harvests were made (63, 181, and 369 days), and soil and plant materials were monitored for [14C]-TNT and metabolite concentrations. The 11.2 mg/kg TNT dose was not phytotoxic to the plant species tested. Continual uptake of TNT into grass blades was observed over the oneyear period, with a total accumulation of 1.3%, 0.9%, and 0.8% of the initial soil [14C]-TNT dose for orchard grass, perennial ryegrass, and tall fescue, respectively. All [14C]-TNT residue in plant material was incorporated as bound residue. At final harvest, radioactivity was concentrated most highly in the root > crown > blade for all species. Soil TNT was gradually reducedtoaminodinitro-toluenesandthenfurthertoanunidentified metabolite(s). Overall, orchardgrass appeared to be the most efficient species at taking up TNT.

Introduction Nitroaromatic compounds are widespread contaminants on over 16 000 Department of Defense (DoD) facilities (1), including manufacturing/handling sites and munitions ranges. An estimated 700 000 cubic yards of soil and 10 billion gallons of groundwater require treatment at tremendous cost to the DoD (2). 2,4,6-Trinitrotoluene (TNT) and hexahydro1,3,5-trinitro-1,3,5-triazine (RDX) are the primary contami* Corresponding author phone: 541-737-9414; fax: 541-737-8160; e-mail: [email protected]. † Department of Environmental and Molecular Toxicology, Oregon State University. ‡ College of Veterinary Medicine, Oregon State University. § Animal Metabolism-Agricultural Chemicals Research Unit, USDA ARS. | USDA Environmental Management and By-Product Utilization Laboratory. 10.1021/es903671n

 2010 American Chemical Society

Published on Web 07/28/2010

nants at these sites, along with dinitrotoluenes and other nitro-substituted explosives (octahydro-1,3,5,7-tetranitro1,3,5,7-tetrazocine (HMX) and C4). These compounds and their metabolites represent environmental hazards, as they exhibit considerable toxicity to aquatic and terrestrial organisms (3), including humans (4). The long-established practice for remediating soil nitroaromatics has been through extensive excavation and subsequent incineration. While effective, this method is not very practical from logistical or economical perspectives. Ex situ slurry anaerobic/aerobic treatment or composting are effective in forming strongly bound residues of TNT, but are also quite expensive compared to in situ processes (5, 6). Alternatively, in situ bioremediation technologies are implemented on-site with minor soil manipulation, have few health and environmental precautions to consider, and are less expensive relative to other procedures (7). Our group is proposing an alternative in situ remediation process which uses plant uptake (phytoextraction) coupled with ruminant processing of the plant biomass to achieve effective remediation of soil TNT residues (8-10). Formation of covalently bound residues of TNT is an effective remediation method in soils (11, 12). Formation of bound residue requires reduction from nitro to amine functional groups, followed by formation of amide bonds with soil (fulvic and humic acids and humin) or plant organic compounds. Plants promote formation of bound residue both in their rhizosphere and their roots and shoots, although some TNT metabolites may remain unbound. Because the rumen represents a highly reductive environment which effectively converts soluble TNT into covalently bound residue (8, 9), a combined plant-rumen bioremediation process offers promise to achieve effective bioremediation of soil contaminated with TNT. The degree of uptake and bioremediation of TNT by plants appears to be dependent on plant species, soil TNT concentration, and TNT bioavailability (13, 14). At least 45 plant species have been studied for transformation of TNT, ranging from poplar trees to grasses to wetland plants (11). An important question to be addressed is finding the plant species that is most adept at absorbing and transforming TNT (15), which will differ depending on soil type and other environmental conditions. In general, grasses appear to be effective at performing this task, including Festuca arundinacea (tall fescue) (13, 16, 17) and Lolium perenne (perennial ryegrass) (18-20). Additionally, grasses are perennial and have well-established cultural practices, making them easy to manipulate and manage for contaminant removal (15). Many phytoremediation studies involving transformation or uptake of explosives have utilized hydroponic systems. While beneficial, these studies avoid a level of complexity in the bioremediation of soil TNT. For example, concentrationdependent effects of TNT in plants do not correlate well between hydroponic and soil matrices (18). In addition, hydroponic studies generally cover short time intervals as opposed to long-term experiments which would more closely mimic conditions for phytoremediation of contaminated soils. Therefore, the objective of this study was to investigate the fate and uptake of [14C]-TNT from soil in three species of cool-season grasses over a one year period in a controlled greenhouse environment.

Experimental Section Chemicals. 2,4,6-Trinitro[U-14C]toluene having a specific activity of 6.12 mCi/mmol was synthesized as described previously (9) with a radiochemical purity of 98%. Unlabeled VOL. 44, NO. 16, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6325

2,4,6-TNT,4-amino-2,6-dinitrotoluene(4ADNT),and2-amino4,6-dinitrotoluene (2ADNT) standards were purchased from ChemService. 2,4-Diamino-6-nitrotoluene (2,4DANT) and 2,6-diamino-4-nitrotoluene (2,6DANT) were purchased from AccuStandard (New Haven, CT). Solvents used were of HPLC grade. Milli-Q (resistance >18 MΩ/cm) quality water was used for all aqueous solutions. Plants. “SR8600” Tall fescue (Festuca arundinacea), “SR4600” perennial ryegrass (Lolium perenne), and “Century” orchardgrass (Dactylis glomerata) seeds were procured from a certified seed source (Seed Research of Oregon, Corvallis, OR). Soil Preparation. Chehalis Silt Loam (fine-silty, mixed, superactive, mesic Cumulic Ultic Haploxerolls) was obtained from the Organic Growers Club farm of Oregon State University; the soil had low organic matter, a pH of 6.5, and 36 ppm phosphorus (bicarbonate extraction method). Soil was dried on a countertop until it was easily sieved (5 mm) and was then placed into pots until soil level was ∼1.5 cm below each rim (average 1433 g soil/pot). Nutrients were added to the soil as follows: 100 mg P/kg (as KH2PO4), 100 mg N/kg (as NH4NO3), 1 mg B/kg (as H3BO3), 50 mg Mg/kg (as MgSO4), 2 mg Zn/kg (as ZnSO4), and 1 mg Cu/kg (as CuSO4). An acetone solution containing [14C]-TNT (2.8 µCi/ mg) and 10 mg/kg cold TNT was dripped onto the soil of each pot via pipet so that treated pots contained a total nominal dose of 11.2 mg/kg TNT, representing the lower end of TNT contamination typically measured on ranges and manufacturing sites. Nutrient and acetone solutions were allowed to dry; the contents of each pot were then placed into Ziploc bags and mixed thoroughly by repeated inversion. The soil was allowed to age for six and a half months at 4 °C, with weekly additions of water and subsequent mixing. The aging process resulted in soil containing TNT, 2ADNT, and 4ADNT at the study’s initiation (see Table 2). Experimental Design and Growth Conditions. Four replicate pots containing aged soil for each of the three grass species were prepared with an equal number of control pots containing soil that was not amended with TNT. From these four replicates, one pot each for the treated and control groups was used for destructive soil sampling, whereby soil core samples (∼10 cm) were taken at plant clipping times from the pots (which also contained plants). Saucers were placed under the pots for watering. On day 0, soil was removed from Ziploc bags and placed into pots. Approximately 100 seeds were mixed with 30 g of soil from the pot and spread on the soil surface. Pots were covered with paper until seedlings emerged to avoid soil crusting. A 16 h light cycle was used to mimic June conditions at approximately 45° N in order to obtain maximum growth rates. Daytime temperatures averaged 26.7 ( 3.8 °C, with night-time temperatures averaging 13.2 ( 2.2 °C. Light was set on for 16 h at 225-445 µmol PAR/m2/s. Pots were placed randomly in the greenhouse to correct for positional inequality and were rotated every two weeks (completely randomized design). Deionized water was added to pots twice a week as follows: on day one, water was added to the saucer until pots stopped taking up water (water balancing all pots). Over the next several waterings, ∼50 mL of deionized water was added to the top of each pot; when plants started to show signs of wilting, all pots were water balanced again. Harvesting Plant Material. Plants were approximately 15 cm tall when first clipped approximately 7 cm above the soil surface, mimicking sheep grazing. Clippings occurred 63, 181, and 369 days after planting. Clippings were immediately weighed and then dried in a hood until a constant dry weight was achieved. After clipping, soil was fertilized with 100 mg N/kg (as NH4NO3) and 100 mg K/kg (as K2SO4) to maintain soil nutrient supply. (Dry weight of shoots harvested was used for N and K determinations to estimate 6326

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 16, 2010

N and K removal from the soil.) At 369 days, the whole plants were collected and separated into shoot, crown, and root sections. The crown and root portions were rinsed with water to remove soil and allowed to dry in a hood. Extraction and HPLC Analyses. Dried soil and plant materials were ground with a mortar and pestle in liquid nitrogen prior to extraction and for combustion analyses. In addition to the day 0 and day 369 soil samples from all 24 pots, soil was removed from the pots designated for destructive soil sampling at 63 and 181 days in order to monitor the fate of TNT over time. Soil was extracted for analysis of TNT and its metabolites as follows: a 2.0 g subsample (dry weight) was placed into 4 mL of acetonitrile and shaken for 18 h at 200 rpm at room temperature. Samples were allowed to settle for 30 min; 3 mL of extract was placed into a test tube with 3 mL of 5.0 g/L CaCl2. After 30 min, samples were filtered through a 0.2 µm PTFE filter, with the first milliliter discarded. A 100 µL aliquot was analyzed by HPLC for TNT and its metabolites based on EPA method 8330A (21) and Fleischmann et al. (2004, see ref 8). Briefly, extracts were analyzed with a Waters 2690 Alliance HPLC and a Waters 996 photodiode array detector (230 nm). Separation was performed using a Waters Nova-Pak C8 column (3.9 mm × 150 mm, 4 µm) and C8 guard column at 35 °C and an isocratic mobile phase of 15/85 isopropanol/water run at 1.0 mL/min for 23 min. Samples were stored at 10 °C in the dark. Results were compared against known standards of TNT, 2ADNT, 4ADNT, 2,6DANT, and 2,4DANT, and appropriate quality control samples were run. Detection limit for HPLC analysis was 0.2 ppm of solution extract injected for TNT and its metabolites. The organic extract containing the soluble residues was also analyzed by liquid scintillation counting; 1 mL of organic extract was added to 10 mL of 3a70B scintillation cocktail (Research Products International Corporation) and analyzed on a Beckman LS6500 multipurpose liquid scintillation counter. Plant material was blended and then extracted by placing 75 mg of dried, ground blade, crown, or root into 300 µL of acetonitrile and shaken for 18 h at 200 rpm at room temperature. The resulting suspension was then centrifuged at 239g for 1 h, and an equivalent volume of 5.0 g/L CaCl2 was added. This mixture was filtered with a 0.2 µm Teflon filter and analyzed by HPLC as detailed above. Combustion Analysis. Ground soil and grass samples were assayed for total radiocarbon content after weighing sample aliquots into cellulose combustion boats capped with cellulose wicks. Samples were combusted using a Packard model 307 sample oxidizer (Perkin-Elmer), and Carbosorb (PerkinElmer) was used to trap [14C]-CO2 released during combustion. The trapped radioactivity was diluted with Permafluor E+ LSC fluid (Perkin-Elmer) and was quantified using liquid scintillation counting (Packard model 1600). Background radioactivity was determined using blank samples processed in an identical manner as the test samples. Autoradiography. Autoradiography of dried plant material was performed using a phosphorimager (Storm, GE Health Care) with ImageQuant software. A color photograph was also taken for orientation. Data Analyses. Multivariate analysis of variance (MANOVA), analysis of variance (ANOVA) (SAS and Statgraphics v5.1), and two-tailed student’s t-tests (Excel 2007) were used to analyze data. A significance level of p < 0.05 was accepted for all statistical analyses.

Results and Discussion This study supports an agricultural solution to munitions remediation, whereby cool-season grasses absorb and partially remediate TNT from the soil and anaerobic ruminal microbes complete the reduction and covalent binding of TNT in the explosive-residue-laden plant material (8-10).

TABLE 1. Disposition of Radiocarbon in Soil and Plant Material in Soil Dosed with 14[C]-TNT recovered activityb grass species

% starting concentrationa

soil, day 369

plant blade, day 63

plant blade, day 181

plant blade, crown, and root, day 369

total recovery (%)

orchard grass perennial ryegrass tall fescue

41.14 ( 13.99 34.16 ( 3.86 25.96 ( 8.99

76.95 ( 9.63 86.05 ( 13.43 113.37 ( 53.88

0.08 ( 0.05 0.11 ( 0.12 0.09 ( 0.02

0.09 ( 0.06 0.06 ( 0.05 0.09 ( 0.04

1.13 ( 0.86 0.70 ( 0.15 0.62 ( 0.58

80.43 ( 10.12 87.77 ( 13.27 126.41 ( 54.47

a Pots were dosed with a nominal concentration of 1.150 mg 14[C]-TNT equivalents/kg of soil. Soil was aged for 6.5 months at 4 °C in Ziploc plastic bags and then placed into pots. As discussed in the text, the recovered concentration on day 0 (planting) was a fraction of the anticipated nominal concentration; the percent recovered is reported here. b Percentages are expressed by normalizing to day 0 soil value and represent average ( standard deviation for n ) 3 pots. Measurements were made via combustion analysis.

Cool season grasses have thick root mats that extend two feet into the soil and lesser rootlets that extend up to 5 ft below the surface (13). Extensive root systems provide grasses with a large absorptive surface area, making these species particularly effective at phytoremediation (13, 16). As the grass matures and the munition is translocated into grass blades, ruminant animals would be introduced and allowed to graze, using a rest and rotation approach. This would allow for repeated plant growth and munition uptake, providing continual low-cost remediation of contaminated lands. The bioavailability of pure TNT in sheep is extremely poor, with no parent TNT measured in urine or tissues of dosed animals (9). Given the transformation capacity of the rumen toward TNT, no adverse affect on grazing ruminants is expected at plant residue levels, giving the additional benefit of increasing the potential productive value of contaminated lands. The current study focused on the phytoremediation portion of our phytoruminal bioremediation solution and examined the uptake and transformation of TNT by orchardgrass, perennial ryegrass, and tall fescue. All three species of grass were able to grow on soil contaminated with TNT. Shoot weights of harvested orchardgrass, perennial ryegrass, and tall fescue did not differ between plants grown in control and TNT amended soils (p > 0.05, see Supporting Information). In addition, no differences in shoot weights were detected across species (p > 0.05). This study was conducted with a nominal concentration of 11.2 mg/kg TNT, aged for 6.5 months; the lack of phytotoxicity confirmed a screening benchmark of 30 mg/kg TNT in soil for toxicity to terrestrial plants (3). In conducting a mass-balance of radioactivity, it was found that only ∼34% of the initial radioactivity dosed to the soil was recovered on day 0 of the experiment (Table 1). As described above, soil was dosed with cold and [14C]-TNT, and then aged at 4 °C for 6.5 months in Ziploc bags, with weekly watering and mixing of the soil. Volatilization of a significant portion of the radiolabel as 14CO2 is unlikely, as demonstrated in previous mass balance studies using [14C]TNT (22-24). Rapid loss of TNT spiked into sediments was seen in an ecotoxicity study, where soil aged for 57 days recovered only 25-40% of the nominal dose of TNT as TNT and its metabolites (25). Another study was conducted with poplars in [14C]-TNT-inoculated fresh soil and soil aged for 280 days; they found that uptake into poplars in the aged soil was much lower than that in the fresh soil system, indicating decreased bioavailability of TNT in aged soil (24). Thompson et al. (1998, see ref 24) conducted a [14C]-TNT hydroponic experiment that subsampled poplar tissues from 2 to 42 days, and although deemed not statistically significant, it showed a general decrease in total recovery of radioactivity over time. Another study which examined the phytotoxicity of TNT showed TNT levels in soil samples measured immediately after blending were lower than anticipated for all concentrations (5-400 mg/kg soil); recovery was particularly low for soils with nominal concentrations of 5 and 25 mg/kg (20%

and 31% recovery, respectively) (26). Thus, there seems to be an unidentified factor with aging of TNT that results in lower total recovery of the nominal dose of TNT, especially when lower concentrations of TNT are used. Perhaps this was more prominently reflected in our study since it encompassed a long aging period of 6.5 months and used a low nominal dose of 11.2 ppm TNT. Thus, all mass balance calculations were conducted by normalizing to total soil radioactivity detected on day 0 of the planting experiment (Table 1). The first two cuttings showed a small amount of radioactivity taken up by each plant species (0.1%) (Table 1). The major portion of recovered activity was in the final harvest at 369 days when radioactivities in the blade, crown, and root were summed. Autoradiographic analyses also showed that radioactivity from soil [14C]-TNT was taken up by all three grass species at each harvest and that, at experiment termination, radioactivity was concentrated in the plant crowns and roots (Supporting Information). These results are in agreement with several other studies which have found TNT-related residues to be predominantly retained in the root mass as bound material (11, 27). A study comparing rhizosphere to bulk soil in pots containing TNT planted with Triticum aestivum found the rhizosphere soil to be significantly reduced in extractable TNT and metabolite concentration, indicating active TNT transformation occurs at the root surface (28). When accounting for total radioactivity in the experiment, 1.3%, 0.9%, and 0.8% of the recovered radioactivity was detected in orchardgrass, perennial ryegrass, and tall fescue plants, respectively. Given this low rate of uptake, additional studies are needed to show the ability of cool-season grasses to adequately sequester TNT for delivery to ruminants using higher nominal concentrations of TNT, performed immediately after soil amendment. However, the mass balance and autoradiography data do demonstrate continual uptake of TNT and its transformation products throughout the year-long study. This bodes well for establishing a perennial plant species which can be repeatedly harvested (or grazed), with TNT removal continuing during subsequent plant growth. Evidence of continual residue uptake is also presented in Figure 1 which shows the distribution of TNT equivalents across species and cutting, and in plant crown and roots. Uptake of radioactivity increased over time such that the TNT equivalents in the blades at 369 days were greater than at 63 or 181 days for all plant species (p < 0.001). This illustrates the importance of chronic exposure studies in addition to acute exposure studies utilizing a few days or weeks of plant growth. At 63 days, no species differences in TNTequivalent accumulations were detected in clippings (p > 0.05). At 181 days, orchardgrass clippings contained more [14C]-TNT equivalents than perennial ryegrass and tall fescue (p < 0.001). Clippings at 369 days from tall fescue had accumulated fewer TNT-equivalents than either orchard or perennial ryegrasses (p ) 0.006). In root crowns, orchardgrass VOL. 44, NO. 16, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6327

FIGURE 1. TNT equivalents (mg/kg) in the blade at each plant harvest and the crown and root at experiment termination. Data are expressed as mg TNT equivalents/kg (dry weight) of plant material. Letters above a bar which are different from each other (A and B) indicate a significant difference (p < 0.05) between grass species within that clipping date and plant part. Error bars represent one standard deviation (n ) 3). accumulated more radioactivity than either perennial ryegrass or tall fescue (p ) 0.005). No species differences in the accumulation of radioactivity within roots were apparent at 369 days (p > 0.05). Within plant species, radioactivity in the crowns and roots was greater than that in the blade (p ) 0.007 and p ) 0.001, respectively); concentrations of radioactivity within crowns and roots were not different (p > 0.05). Again, the concentration of radioactivity by the roots of grasses is consistent with the conclusions of other studies (11, 27, 28). The high concentration of radioactivity in the crown is interesting as this portion of the plant lays on the soil surface and so may provide an opportunity for delivery of concentrated munitions to grazing sheep. Taking all the plant harvests together, a total accumulation in the leaf blades of 2.19 ( 0.77, 2.09 ( 0.58, and 1.02 ( 0.28 mg TNT equivalents/kg plant material was measured for orchard grass, perennial ryegrass, and tall fescue blades, respectively. Uptake of radioactive residues was also expressed as a rate (uptake per day) for the intervals of 0-63, 63-181, and 181-369 days. For the first interval, orchardgrass took up an average of 2.7 ( 0.3 ng 14C-TNT equivalents/g (dry shoots)/ day, while perennial ryegrass took up 2.6 ( 1.2 ng/g/day and tall fescue 2.0 ( 0.2 ng/g/day. No difference was exhibited between species at this time point (p > 0.05). For the interval between 63 and 181 days, residue uptake in orchardgrass (2.1 ( 0.4 ng/g/day) exceeded that in perennial ryegrass (1.0 ( 0.1 ng/g/day) and tall fescue (1.1 ( 0.2 ng/g/day) (p < 0.001). Between 181 and 369 days, accumulation of radioactivity in orchardgrass (9.4 ( 3.8 ng/g/day) and perennial ryegrass (9.6 ( 2.9 ng/g/day) exceeded that in tall fescue (4.1 ( 1.4 ng/g/day) (p ) 0.006). The large jump in rate of uptake of TNT for the third harvest may be explained by the fact that it fell during the warm spring and summer months of the experiment. When broken down into harvest periods, temperature highs and lows differed from each other for each period (p < 0.05), averaging 25.3 ( 2.7 °C and 14.8 ( 1.5 °C for 0-63 days, 22.5 ( 1.3 °C and 11.8 ( 2.0 °C for 64-181 days, and 29.6 ( 2.6 °C and 12.9 ( 2.1 °C for 182-369 days, respectively. Even though plants were grown in a controlled greenhouse environment, it appears that the temperature controls in this facility were not as stringent as we would have hoped. Thus, we theorize that the increased heat associated with the latter harvest period could contribute to the large increase in rate of TNT uptake seen there, although greater active root mass could also cause this outcome. Alternatively, it is possible that continual aging of the soil 6328

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 16, 2010

FIGURE 2. Percent remaining extractable TNT, 2-ADNT, and 4-ADNT in soil amended with 11.2 mg/kg TNT and grown for 369 days with three species of cool-season grasses. All concentrations are based on USEPA method 8330A as described in the Experimental Section. Values represent average ( standard deviation, n ) 3. TNT ) trinitrotoluene, 2ADNT ) 2-aminodinitrotoluene, 4ADNT ) 4-aminodinitrotoluene. provided more 2ADNT and 4ADNT, which the plants were able to take up at a faster rate. Efforts to extract parent [14C]-TNT and/or its metabolites from plant materials using EPA method 8330A were unsuccessful; chromatography of extracts by HPLC with radiochemical detection provided no evidence of parent TNT or metabolites (data not shown). In addition, plant extracts contained no radioactivity. These results strongly suggest that the radioactivity detected in plant tissues by autoradiography and combustion analyses (Supporting Information and Figure 1) was present as bound species. Such a supposition is consistent with results reported by Vila et al. (2007) who determined that more than 70% of the radioactivity present in maize, wheat, and soybean roots and about 50% of that in the aerial parts was unextractable. It is possible that, given the relatively low concentration of TNT used in this study (11.2 mg/kg) and the long time periods between plant harvests, the TNT and/or metabolites were taken up and converted to more polar, bound complexes in the plant material. Other studies have shown that TNT is quickly metabolized to polar products which are compartmentalized in vacuoles or covalently bound to the cell components of plants (11, 29, 30). This lends support to the theory of plants immobilizing TNT through covalent linkages to plant macromolecules. As discussed above, an accepted means of plant remediation of TNT involves the formation of covalent bonds with TNT or TNT metabolite residues which are unextractable using acid and/or enzyme hydrolyses and an array of organic solvents. It is widely accepted that, once these bonds are formed, the residues are nonhazardous, eliminating the environmental risk posed by TNT since they are not bioavailable (31, 32). While looking at plant material in the digested format, we also saw this covalent binding in our previous mass balance study examining TNT disposition in sheep fed [14C]-TNT, as 82% of the material was detected in the feces or gastrointestinal content as bound residue (9). Covalent binding of TNT metabolites occurs in soil as well. Several studies have shown that TNT and its metabolites become irreversibly bound to humic substances (33, 34), thereby achieving a mechanism of remediation. Soil covalent binding of TNT would, of course, influence the bioavailability of TNT to plants in contaminated soil, thus limiting the amount of explosive available to the plant for absorptive processes (14). However, caution has been urged in assuming contaminated soil has reached a steady state for TNT concentration, as TNT may continue to be transformed and become bioavailable for some time depending on changing

TABLE 2. Soil Concentration of Extractable TNT and Metabolites from Three Species of Cool-Season Grasses Grown on Soil Amended with 11.2 mg/kg TNT for 369 Daysa 2,4DANT (mg/kg dry soil)

total TNT and metabolites (mg/kg dry soil)

percent TNT and metabolites remainingc

Orchard Grass 2.47 ( 1.04 NDe 1.42 ( 0.24 ND 1.39 ( 0.15 ND 1.02 ( 0.0 0.80 ( 0.30

ND ND ND ND

6.04 ( 2.01 2.86 ( 0.41 3.36 ( 0.13 3.15 ( 0.25

47.4 55.7 52.2

3.72 ( 0.11 3.97 ( 0.20 3.36 ( 0.02 2.28 ( 0.38

Perennial Ryegrass 2.07 ( 0.86 ND 2.28 ( 0.11 ND 1.27 ( 0.05 ND 1.28 ( 0.27