Robin L. Brigmon, Denis Altman, Ed Wilde, Christopher J. Berry,
Marilyn Franck, Fatina Washburn, Pam McKinsey, and Cary Tuckfield
Westinghouse Savannah River Company
Aiken, SC 29808

F. Michael Saunders and Kevin Sessions
Georgia Institute of Technology (GT)

Andrine Stanhope
Florida A & M University (FAMU)

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

This report has been reproduced directly from the best available copy.

Available for sale to the public, in paper, from:  U.S. Department of Commerce, National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161,  phone: (800) 553-6847,  fax: (703) 605-6900,  email:  orders@ntis.fedworld.gov   online ordering:  http://www.ntis.gov/support/ordering.htm

Available electronically at  http://www.osti.gov/bridge/

Available for a processing fee to U.S. Department of Energy and its contractors, in paper, from: U.S. Department of Energy, Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831-0062,  phone: (865 ) 576-8401,  fax: (865) 576-5728,  email:  reports@adonis.osti.gov

List of Key Words
Phytoremediation, bioremediation, perchloroethylene, trichloroethylene, and groundwater.

Bioremediation of chlorinated solvents, both natural and accelerated, is exemplified by phytoremediation and biodegradation by rhizosphere microorganisms. Phytoremediation is the use of vegetation for the treatment of contaminated soils, sediments, and water. The potential for phytoremediation of chlorinated solvents has been demonstrated at the Savannah River Site (SRS) Miscellaneous Chemical Basin, Southern Sector of A/M Area and TNX/D-Area. Recent characterization work at the SRS has delineated widespread plumes (1-2 miles) of low concentration (40 ppb –10-ppm range) trichloroethylene (TCE) and perchloroethylene (PCE) contaminated groundwater. Phytoremediation deployments are underway for TCE and PCE phytoremediation in select SRS areas. Phytoremediation appears to be an excellent technology to intercept and control plume migration.

The ongoing Southern Sector treatability study is part of a multi-year field study of SRS seepline-soil systems maintained under saturated conditions. The primary focus is on determining how trees, seepline groundcover, soil microbial communities, and geochemical and surface-volatilization processes affect TCE and PCE in contaminated groundwater that flows through surface seepline areas. Therefore, FY00 represented an initial acclimation phase for soil and plant systems and will facilitate examination of seepline phyto- and bioactivity in subsequent growth season in FY01.

The Southern Sector project has concentrated on groundwater upgradient of the seepline at Tims Branch. The field research has the objective of determining the efficiency of plants and soil for in situ bioremediation of the VOCs, TCE and PCE, under seepline conditions. Well MSB 88C was selected as the groundwater supply because of it’s VOC concentrations (TCE 188 ppb, PCE 55 ppb) and proximity to the seepline. Three phytoreactors were deployed with soil from the seepline. Phytoreactors 1 and 2 were planted with loblolly pines (Pinus taeda) and hybrid poplars (Trichocarpa X deltoides) respectively. Loblolly pine is a native SRS species. Phytoreactor 3 was left non-vegetated as a soil control to evaluate monitored natural attenuation (MNA). The phytoreactors were supplied with a continuous flow of contaminated groundwater. Results to date demonstrate that the loblolly pine phytoreactors can remove up to 90 % of the TCE and 80 % of the PCE. The hybrid poplar phytoreactors demonstrated up to 100% removal of the groundwater contaminants. No detectable amounts of these VOCs were found in transrespiration, soil volatilization, or soil core testing. Microbial activity in the phytoreactors and seepline soils is under investigation using anaerobic microcosms for assessing biotransformation of TCE and PCE.

Drip irrigation has been initiated at four 0.2 acre test plots in D-Area as part of a recently initiated field study to evaluate the use of plants for treating TCE-contaminated groundwater from an aquifer 30-50 feet below the surface. The novel process to be tested involves pumping TCE-contaminated groundwater from the deep subsurface and distributing it via a network of shallow drip emitters to the rhizosphere regions of vegetative test plots where degradation processes can be effectuated by plants and their associated microflora. Phase 1 of the project involves evaluating and optimizing the effectiveness of the delivery system (i.e. getting the maximum possible amount of TCE from the aquifer to the rhizosphere). Subsequent work will focus on comparatively evaluating and optimizing the ability of various plants to detoxify the contaminant and to determine feasible treatment flow rates and operational costs.

It has been estimated that over 13 million pounds of chlorinated degreasing solvents, including trichloroethylene (TCE) and perchloroethylene (PCE) were used at SRS during reactor operations. Although much of the waste volume was reduced by evaporation, over 3 million pounds, including 317,000 pounds of TCE, were discharged to the M-Area Settling Basin and the A-014 outfall. The

M-Area Settling Basin and A-014 outfall were unlined and much of these solvents seeped into the subsurface contaminating the groundwater. The associated groundwater zones in A/M Area (i.e. M-Area and Lost Lake Aquifers) discharge to seeplines adjacent to Tims Branch and Upper Three Runs Creek (WSRC-TR-99-00113). As part of the ongoing compliance and research activities at SRS, evaluations of the nature and extent of groundwater contamination in the A/M-Area are ongoing in the Southern Sector, primarily between recirculation wells and the seepline area. Based on the local hydrogeology and topography, it was predicted that VOC contaminated groundwater would emerge as surface water along a seepline region in the Southern Sector of the A/M-Area. The seepline is presently heavily covered with a variety of vegetation. This area lends itself to the potential of phytoremediation with species known to degrade VOCs and monitored natural attenuation (MNA).

Previous research at SRS has demonstrated the potential for phytoremediation of chlorinated ethenes. A recent investigation of a shallow VOC contaminant plume at the SRS TNX flood plain demonstrated that bald cypress (Taxodium distichum), tupelo (Nyssa aquatica ) and loblolly pine (Pinus taeda) contained significantly higher levels of chlorinated ethenes than adjacent oak (Quercus spp.) and sweet gum (Liquidambar stryaciflua) trees in the same area (Vroblesky, Nietch, and Morris, 1999). Walton and Anderson (1990) previously observed accelerated microbial degradation of TCE in slurries of rhizosphere soil and mineralization of TCE in whole plant systems collected from samples at a former SRS solvent disposal site, the Miscellaneous Chemical Basin (MCB). Two species where enhanced microbial degradation of TCE was observed at the SRS were a legume, Lespedeza cuneata and loblolly pine, Pinus taeda (Anderson and Walton, 1995).

The extent to which rhizosphere VOC remediation varies among soils in this area is uncertain. However, a better understanding of such variability is necessary since rhizosphere responses to seasonal changes including plant succession, rainfall, and temperature can significantly influence potential VOC bioremediation. A microcosm study was undertaken to estimate the potential of Southern Sector rhizosphere soils along the seepline to naturally attenuate TCE. This monitored natural attenuation MNA study demonstrated that sorption to soil was the dominant mechanism removing as much as 90% TCE (Brigmon et al., 1998). A limited amount of TCE aerobic biodegradation and anaerobic reductive dechlorination was observed through the appearance of cis-1, 2-dichloroethylene (c-DCE), and trans-1, 2-dichloroethylene (t-DCE) in microcosm tests. Soils from vegetated areas mineralized TCE several times greater than soils from adjacent non-vegetated areas (Walton and Anderson, 1990). It has been suggested that a possible mechanism for the enhanced microbial mineralization of TCE in the L. cuneata rhizosphere soil is excretion of phenolic compounds in root exudates. Since phenol is a known inducer of toluene monooxygenase, an enzyme responsible for degradation of TCE, the natural plant exudates could play a role in biodegradation of TCE in the rhizosphere (Anderson, Guthrie, and Walton, 1993). Select plants including hybrid poplars are capable of TCE metabolism and transformation (Newman et al., 1997; Schnabel et al., 1997). The two tree species selected for this study based on their phytoremediation potential were the loblolly pine, L. cuneata, and a hybrid poplar, Trichocarpa X deltoides.

One of the primary functions of root exudates is to release natural chelating agents (citric, acetic, and other organic acids) that increase nutrient and contaminant soil mobility. Exudates may also include enzymes, such as nitroreductases, dehalogenases, and laccases that can degrade organic contaminants (Fliermans et < biblio >). Some rhizosphere microorganisms secrete plant hormones that increase root growth and the secretion of root exudates that enhance microbial activity. Exudation of organics by plant roots and turnover of organic root biomass also increases the TCE sorption capacity of soil (Schnabel et al., 1997). The microbial ecology of soil associated with bioremediation in mycorrhizal roots such as pine has not been well characterized even though this environment provides extensive surface area for bacterial colonization. It was previously observed that pine rhizosphere soils in the SRS MCB contained higher quantities of potential TCE-degrading bacteria than the Southern Sector soils that have not been exposed to TCE (Brigmon et al., 1999).

The overall objective of the SRS D-Area project is to evaluate a novel drip irrigation/phytoremediation process for remediating trichloroethylene (TCE)-contaminated groundwater. The phytoremediation application in this project involves two mechanisms: (1) rhizodegradation or the breakdown of organic contaminants by microbial activity enhanced by the presence of plant roots, and (2) phytodegradation (also known as phytotransformation), the breakdown of contaminants by plant metabolic processes. Most of the TCE-contaminated groundwater in D-Area occurs near the bottom of an approximately 30-50 ft. thick aquifer, well below the depth of typical tree root penetration. Thus, the drip irrigation component of the proposed process provides a means to allow plant communities an opportunity to remediate contaminated groundwater from depths otherwise unavailable to plant systems.

The first objectives of the first phase of the study, which is now underway, is to demonstrate that TCE contaminated water can be pumped from depths of 30+ ft. and applied via a drip irrigation system to the rhizosphere region of a plantation. The second objective is to demonstrate that the TCE-contaminated water reaching the rhizosphere can be partially or fully degraded by plants and their associated microflora. The second phase of the study will evaluate seasonal effects, compare the effectiveness of different species plants, and flow limitations for the system. The techniques describe here in conjunction with other applications should provide tools for screening plant species and soils for phytoremediation and MNA activity. Application of phytoremediation should provide significant advantages over conventional remediation techniques for chlorinated ethene-contaminated groundwater. The metabolic actions of the plants and soils in combination with physical reductions of VOCs by volatilization and dilution will enable active remediation at the rhizosphere of the seepline.

A better understanding of the potential mechanisms involved for enhanced biodegradation in the root zone and the interaction between plants, microorganisms, and contaminants can be useful in phytoremediation applications. This information could lead to improved phytoremediation applications including selection of plants, soil amendments, and irrigation systems. Future work based on these techniques could be used to determine phytoremediation deployments and strategies in response to TCE/PCE-contaminated groundwater movement.

Field Treatability Test. After initial site selection (Well MSB 88), an evaluation of Southern Sector soil was undertaken. The area selected was based on the proposed location identified by the Environmental Restoration Department and the well VOC concentrations. The soil (very sandy with red clay) in the area of MSB 88 is not representative of the seepline. Therefore, soils had to be brought to the study site from the seepline. Soils above (to 0.5 m depth below surface) and below the rhizosphere (0.5-1 m depth below surface) were collected in the vicinity of Well MSB 50 (located in the vicinity of the seepline area) for the study. The Georgia Institute of Technology (GT) set up soil columns on 5-28-99 for initial flow and soil-permeability tests. Hydraulic permeability and porosity measurements were made to assist in soil characterization and assessing phytoreactor design and soil placement in the phytoreactors. On 9-28-99, the double insulated Phytoreactors (72"x 48"x30") (Bonar Inc., Atlanta, GA) were brought to the site for set up. Initial startup testing of the Phytoreactors began with groundwater from MSB-88 B that had low (5 ppb) VOC concentrations. Figure 1 shows a diagram of the project field deployment configuration. Figure 2 illustrates the phytoreactor and the set up process. In January 2000 permission was received to use Well MSB 88C that had concentrations in the 120-ppb range for TCE and 60 ppb range for PCE. On 3-20-00, all Phytoreactors were placed in operation receiving contaminated groundwater.

The phytoreactors were developed with an upflow pattern of groundwater flow. Groundwater from MSB-88C is pumped into a 1000 gal polypropylene storage tank that supplies the phytoreactors though a gravity-fed system. A 3-in. layer of gravel in the bottom of the phytoreactors supports a 2-line influent-distribution system in the bottom of each phytoreactor. The gravel layer was then covered with 20 in. of seepline soil. Two separate effluent collection systems were included in each phytoreactor. The effluent collection lines are located 10 in. and 18 in. (i.e., immediately below the soil surface) above the influent lines and parallel to them. This flow pattern allows for simulation of the groundwater upflow through the seepline soils and the collection and removal below the root zone of the plants. The effluent collection system at the 10-in. depth is the one being used in all phytoreactors and provides a 10-in. saturated flow zone and a 10-in. vadose zone for the phytoreactors. Three phytoreactors were set up for the project. Loblolly pine (L. cuneata) was planted in Phytoreactor 1, the hybrid poplar (Trichocarpa X deltoides) in Phytoreactor 2, and Phytoreactor 3 contained only seepline soil as a non-vegetated control. A 1000-gallon steel tank is used for effluent collection downhill from the site.

Collection of Samples. Sampling groundwater from the phytoreactors for chemical and microbial analysis began on March 30, 2000. During April 2000, the phytoreactors were sampled weekly for groundwater influent and effluent microbial activity, VOCs, and ion analysis. Thereafter, influent and effluent groundwater and soils from the phytoreactors were sampled monthly.

D Area. Drip irrigation lines have been installed in four test blocks above the TCE-contaminated groundwater in D-Area. Each test block consists of three adjacent 0.2 acre test plots. The plots were prepared by removing all vegetation in two of the plots and all vegetation except mature pine trees in the third plot. One of the two completely cleared plots in each block serves as non-vegetated control and the other will be planted with cottonwood trees in the spring of 2001. Each treatment plot is 94 ft X 94 ft. Although the entire plot is plumbed for irrigation, sampling is restricted to the interior 66 ft X 66 ft (0.1 acre) region, thus providing a 28 ft buffer along the exterior portion of each plot.

The irrigation design for each plot consists of 23 irrigation feed lines spaced 4 ft apart and running the entire distance of the plot. Groundwater is presently pumped from a well to a 2500 gallon stainless steel tank. The water is then pumped at timed intervals from the tank to the irrigation system of the plots via a 1" diameter manifold line. Branching from the manifold line are the irrigation feed lines. Each feed line is attached to a drip line consisting of emitters spaced at 2 ft intervals. The emitter lines are buried approximately 2 inches below the surface. There are four connections between each feed line and emitter line and these are equipped with pulsators, which help regulate flow among emitters. Currently, flow is restricted to a maximum of 120 gallons per day to only four of the 12 plots due to flow limitations. New wells are being developed to allow for increased flow per plot and usage of all 12 plots.

Flow measurements. The influent flow rates for phytoreactors 1, 2, and 3 were collected using a digital flow meter and a datalogger (Campbell Scientific Inc.) powered by a solar cell. While SRTC assisted in set-up of the system, GT was responsible for weekly monitoring and down loading the flow data. The flow rates for each phytoreactor were logged every fifteen minutes. This fifteen-minute value represented an average flow rate over that particular time period.

Gas Chromatography. Samples were taken from phytoreactor groundwater influent and effluent for VOC analysis. Soil samples were collected by hand auger from four locations in each Phytoreactor, two shallow (.5 m), and two deep (1 m). Each soil sample was collected with a modified plastic syringe and placed directly into a 20 mL glass vial with 5 mL deionized water and immediately sealed for subsequent VOC analysis. Chlorinated ethene analysis was performed on samples in sealed glass vials using headspace gas chromatography (GC). The headspace GC method minimizes sample handling and preparation and measures the bulk (sorbed, dissolved, vapor, and NAPL) TCE content of the sample. Samples were analyzed using a Hewlett-Packard 5890 gas GC equipped with a flame ionization detector (FID) and a 60-m SPB1™ column (0.75-mm ID, 1-΅m thick; Supelco, Bellefonte, PA). Transrespiration and soil volatilization gas samples were collected in the field by the method described by Newman et al. (1997). Manual injections of samples from Tedlar gasbag samples from transrespiration and soil volatilization measurements were made with a 250-m L gastight syringe (Precision Scientific, Baton Rouge) (Newman et al., 1997).

Ion Chromatography. Chloride, nitrite, nitrate, phosphate, and sulfate groundwater concentrations were measured with a Dionex DX500 ion chromatograph equipped with a conductivity detector, and a 250-mm Dionex IonPac AS14 Analytical column (4-mm ID, 16-m m bead; Dionex Corp., Sunnyvale, CA), operated at ambient temperatures. A 3.5 mM sodium carbonate/1 mM sodium bicarbonate buffer solution was used as the eluent (1.2 mL/min). Samples were taken from the supernatant of a solution prepared from groundwater or 5 g of dry soil (dried at 121° C for 24 hours) and 5-mL of deionized water vortexed for 1 minute then centrifuged for 5 minutes at 2500 rpm.

Microbial densities. Comprehensive analysis of specific microbial populations and characterization of the metabolic activity of whole microbial communities can be an effective tool to predict the bioremediation potential of a natural system. These analyses can enable monitoring the activity of specific microorganisms in reducing and/or removing harmful groundwater contaminants. In this project groundwater samples were collected in sterile 50-mL centrifuge tubes and transported to the lab for immediate microbiological processing. Total microbial population densities in phytoreactor influent and effluent groundwater and soils were determined by the Acridine Orange Direct Count (AODC) Method (Balkwill, 1989). The viable microbial population densities of aerobic and facultative heterotrophic bacteria in groundwater and soils were determined using spread plate techniques. Low concentrations (1%) of Peptone-Trypticase-Yeast extract-Glucose (PTYG) media was used (Balkwill, 1989). Community-level physiological analysis using BIOLOG GN2 plates indicates the utilization rate of 95 carbon sources by microorganisms in the groundwater. Individual substrate utilization patterns of groundwater microbial communities were obtained with BIOLOG GN2 plates from each groundwater sample. Groundwater (150 m l) samples were used for direct inoculation of the BIOLOG plates. Autoclaved deionized water was used as a control. All plates were incubated at room temperature and the absorbance (590 nm) of the wells recorded after 1 week. The color intensity of BIOLOG GN2 plates was expressed and calculated as the mean of the 95-absorbance values corrected for the background control.

Microcosm Studies. Microcosm tests were set up at GT to assess the microbial activity and ability of phytoreactors and seepline soils to transform PCE, TCE, and typical degradation products of TCE and PCE. Anaerobic microcosms were established to evaluate the potential for indigenous microorganisms to dechlorinate PCE, TCE, dichloroethene (cis-DCE), and vinyl chloride (VC) to subsequent end-products. Microcosms were established in 20-mL vials. Each microcosm contains 2 grams of wet sediment material, 9 mL of phosphate buffered ground water containing a designated electron donor, and 100 m L of hexadecane containing the chlorinated compound, i.e. PCE, TCE, or cis-DCE. Vinyl chloride was added to designated microcosms in gaseous form (i.e., not with hexadecane). The electron donors used were lactate, acetate, or hydrogen (H₂). Acetate and lactate were added at a concentration of 2mM. In the microcosms containing hydrogen (H₂) as electron donor, 3mL of hydrogen gas were added to each microcosm. Resazurin was used as a redox indicator. This dye stays colorless when reduced and becomes pink when oxidized, thus quickly indicating any oxygen contamination of the microcosm. The electron acceptors (chlorinated compounds) used were PCE, TCE, cis-DCE, and VC. Anaerobic microcosms are being used to assess activity for transformation of TCE and PCE to degradation products, as well as the presence of bacterial populations indicative of other favorable bioprocesses (e.g., halorespiration and methanogenesis).

Microbial Respiration. Metabolic chambers were set up with a respirometer (Columbus Instruments Inc., Columbus, OH) to measure soil microbial respiration in the phytoreactors. This method measures the rates of soil oxygen consumption and carbon dioxide production. These measurements were made in August, 2000, when the trees were fully developed. Two soil samples were taken monthly from each Phytoreactor with a stainless steel hand auger. A shallow sample was taken from the top 0.5 m of soil and a deep sample was taken from the bottom of the Phytoreactor in the saturated zone.

D-Area. Prior to the initiation of flow to the system, a soil characterization study of the soil to a depth of 8 ft in each of the 12 test plots was conducted by the USDA-Natural Resources Conservation Service. Measurements of TCE levels at 2 ft and 8 ft were also made in the 4 plots scheduled for irrigation in phase 1 to obtain baseline data. The soil pH of the samples was also measured (Forster, 1995). Following the initiation of flow to the system on 8/31/00, water and soil samples have been collected on an approximately monthly basis and analyzed by gas chromatography to determine TCE and PCE levels within the distribution system and the surrounding soil. Water samples were collected by filling 20 ml glass head space vials to a mark previously measured to represent 10 ml. Immediately following the collection of water, the vials were sealed by crimping the cap on the vial. Soil samples were collected by hand auger at 6" depth intervals from the surface to a depth of 2 ft. Each soil sample was collected with a modified plastic syringe and placed directly into a 20 mL glass vial with 5 mL deionized water and immediately sealed for subsequent VOC analysis. TCE and PCE analyses were performed on samples in sealed glass vials using GC.

Flow measurements. Groundwater supplying the phytoreactors has been measured with flowmeters interfaced with a datalogger powered by solar cells. While all phytoreactors were set up with an influent of 20 mL groundwater per minute (~7.6 gal/d), the rate constantly changed as a result of soil settling, weather conditions, plant growth and root development, and changes in the supply system (i.e. supply tank water levels) (Figures 1).

Removal of TCE and PCE. When filling the supply tank (Figures 1) it was found that Well MSB 88C has consistent VOC concentrations (TCE 188 ppb, PCE 55 ppb) based on historical quarterly monitoring data history. By the time the test phytoreactors with seepline soils actually receive the continuous flow of contaminated groundwater, the concentration averages around 46 ppb TCE and 48 ppb PCE. The concentration of TCE and PCE in both phytoreactor influent and effluent groundwater is shown in Figures 4 & 5, respectively. All three phytoreactors show a reduction in both TCE and PCE groundwater concentrations in the effluent compared to the influent. Phytoreactor 2 (Hybrid Poplar) has shown no detectable PCE or TCE in the effluent groundwater for the June and July 2000 sampling events, indicating total removal. In July, 2000 samples were taken for soil volatilization and plant transrespiration VOC measurements. No detectable TCE or PCE (<5ppb) was found in soil volatilization from any of the phytoreactors. All measurements from the pine and poplar showed no detectable amounts of TCE and PCE being transrespired. These measurements were repeated in August, 2000 with the same results.

There appears to be significantly larger (P <0.05) TCE and PCE concentrations in the influent than in the effluent in every Phytoreactor (1, 2, and 3). However, there was no significant correlation between the influent and effluent difference and any of the covariates, viz., cumulative mean daily temperature, cumulative daily A/M-area rainfall, or the mean of the day's barometric pressure for the given sampling date. In other words, these covariates were not helpful in partitioning out any variability in the response measure of influent and effluent difference for either TCE or PCE. The same is true for the cumulative maximum daily temperature, cumulative maximum SRS rainfall, or the maximum of the day's barometric pressure. Both the common logarithm and the cube root transformations were also used on the influent and effluent measurements to see if differences in the transformed variables were correlated with the covariates. In no instance, however, was there a statistically significant correlation between a transformed response variable difference and any of the covariates.

Although poplars (Phytoreactor 2) appeared to be the most effective treatment so far as contaminant removal in June and July FY00 (Figures 4 & 5), no statistically significant difference existed between any pair of phytoreactors when comparing the average influent and effluent groundwater TCE and PCE difference among phytoreactors. This suggests that the treatment assigned to each phytoreactor was as effective in the remediation of both TCE and PCE as any other phytoreactor. In order for this interpretation to be convincing, the influent and effluent difference must be demonstrated not to be due in large measure to the groundwater distribution system. It must be shown that the delivery system was not simply overwhelming whatever treatment differences there might have been between pairs of phytoreactors.

Plant tissues (roots, stems, and leaves) from the pine and poplar have recently been taken from the phytoreactors for analysis of PCE, TCE and potential metabolic breakdown products including trichloroacetic acid (TCA), and dichloroacetic acid (DCA). This ongoing analysis will provide useful information on the fate of the chlorinated ethenes in the system.

Ion Chromatography. Table 1 shows the influent and effluent groundwater chloride, nitrite, nitrate, phosphate, and sulfate concentrations. The composite water flows, and resulting flow of soluble ions, for the phytoreactors include influent groundwater, influent rain water, subsurface discharge of groundwater and evaporative losses at the soil surface (Phytoreactors 1, 2 & 3) and evapotranspiration by plants (in Phytoreactors 1 & 2). In addition, the soil placed in the phytoreactors contained pore water moisture with dissolved minerals, as well as minerals sorbed to soil surfaces. These flows and sources need to be considered in the assessment of the ion data to date.

Chloride ion should be conservative in the phytoreactors and, except for an initial perturbation in March for the effluent, the influent and effluent data for chloride appear to be similar. The initial phosphate concentration data in March may represent cross contamination of the influent tank; but thereafter, influent and effluent phosphate concentrations were at trace levels.

Sulfate levels in the effluent of the phytoreactors appear to be elevated, relative to the influent in all cases. Sulfate elution from the soils would appear to be the most plausible assessment of this increase, although it is possible that there is sulfide oxidation taking place in the saturated zone. Nitrogen species in the system are nitrate and nitrite. Nitrate concentrations decrease through the phytoreactors, relative to the influent, and nitrite appears in the effluent despite being at non-detect levels in the influent. The transformations of nitrate and nitrite are indicative of (i) plant uptake of nitrogen species and (ii) denitrification by soil microbes. Plants will use nitrate and a primary source of nitrogen and this decrease is likely related to plant growth in Phytoreactors 1 and 2. Phytoreactor 3 has no trees, so the changes in this phytoreactor would be microbially based. The occasional presence of nitrite in effluents would indicate that anaerobic respiration was in process and that nitrate conversion to nitrite and ultimately to nitrogen (N₂) was occurring in the phytoreactors. These responses need to be further examined in the coming year. Finally, the issue of nutrient addition is supported by these nitrogen and phosphorus data (i.e., phosphorus and nitrogen are at low levels and supplementation is warranted). No significant difference between treatments across time was found.

Microbial densities. In all cases the total microbial densities as measured by Acridine Orange Direct Counts (AODC) were higher in the effluent from the phytoreactor groundwater than the influent groundwater. The source of the influent bacteria is from influent groundwater, and microbial growth in influent tank, filter, and associated supply lines. Bacteria in the effluent groundwater are from soils placed in the phytoreactors, influent groundwater, and environmental origin (air, rain, etc.), as the phytoreactors are open systems. The viable microbial population densities of aerobic and facultative heterotrophic bacteria in groundwater as measured by colony forming units/mL (CFU/mL) were more variable than the AODC density-based data. The viable influent groundwater bacteria were generally in lower concentrations as compared to effluent. Phytoreactor 2 containing the poplar trees appeared to have lower concentrations of viable bacteria in the effluent groundwater relative to the other two Phytoreactors. The results of a Kruskal-Wallis rank sums test indicate a significant difference (p < .05) among phytoreactors (i.e. treatments) means for the difference between the influent and effluent plate counts (CFU/mL) (Daniel, 1978). Note that this test assumes that the microbial counts taken regardless of sampling date are statistically independent. Phytoreactor 2 demonstrated smaller influent and effluent microbial differences. This suggests that the phytoreactor treatment (Poplar trees) is not promoting microbial growth. Microbial data from groundwater using BIOLOG for total amount of substrate utilization demonstrated a similar trend for all three phytoreactors (Figure 4). In Figure 4 the difference in the number of BIOLOG positives represents the delta between total substrates utilized (of 95 possible) in effluent vs. influent groundwater for each phytoreactor. Further analysis of substrate utilization by substrate (i.e. aromatics vs. carbohydrates) is underway.

Microcosm Studies. Microbial activity in the phytoreactors and seepline soils has been investigated and is described above. Anaerobic microcosms are being employed to assess activity for transformation of TCE and PCE to degradation products.

Soil samples were obtained from the seepline and the phytoreactors at the site in May, 2000. Soil samples have been handled in an anaerobic glove box at all times in the laboratory and kept under refrigeration at 4° C prior to analysis. Soil samples were labeled in the following way.

The contaminant concentrations utilized were the following:

PCE at 1.25 m L/100 m l hexadecane ~12m mol in system

TCE at 0.8 m L/100 m l hexadecane ~ 9m mol in system

cis-DCE at 0.3 m L/100 m l hexadecane ~ 4m mol in system

VC at 0.2 mL/ vial ~8m mol in system

Microcosms with SED 1 and SED 2 were established during the last two weeks of May 2000. After the microcosms were established, they were sealed with Teflon-lined butyl rubber stoppers and incubated without agitation at 25° C. Headspace samples (0.1mL) are analyzed monthly (HP 6890 gas chromatograph with a FID detector) to monitor dechlorination activity. Controls include vials containing no electron donor, no electron acceptor and autoclaved (killed) controls. Table 2 summarizes the microcosms established with SED1 and SED 2. Initially, only the PCE containing microcosms were analyzed, because when dechlorination occurs under anaerobic conditions, the more highly chlorinated compounds are more rapidly dechlorinated. A response of no dechlorination in the PCE containing microcosms indicates that most probably dechlorination has not taken place in the rest of the microcosms. After 1, 2, and 3 months all TCE and PCE-containing samples established with SED 1 and SED 2 were analyzed and no intermediates were detected. These data indicate that there are no resident microbial populations transforming PCE or TCE in the phytoreactor soils. This response is compatible with the porewater and effluent groundwater data indicating, in general, an oxidizing environment within the saturated soil zone.

For the soils from the phytoreactors, i.e., SED 1, no dechlorination products were found. However, methane production was detected in some of the microcosms, as shown in Table 3. Table 3 shows that all the SED1 test microcosms, except the TCE microcosms, have some methanogenic activity. Microcosms with SED 2 showed no dechlorination products, and no methane production was observed. These data indicate the potential for the development of methanogenic population, but do not indicate that methanogens are present in the soils of phytoreactors 1, 2 or 3 at any significant levels. In fact, the seepline soils (SED 2) did not demonstrate methanogenesis in the microcosm studies. The microcosm analyses will continue to establish the ability of the sediments to transform TCE and PCE. The current data indicate there is no apparent indigenous activity to transform TCE or PCE neither in the saturated soils of the phytoreactors nor in the seepline soils. These microcosms are continuing and other potential seepline soil nutrient limitations (i.e. nitrogen) will be evaluated as relevant to bioremediation potential.

Soil Respiration. Table 4 shows the phytoreactor soil respiration rates. There was no significant phytoreactor or treatment effect on soil metabolic rate. It is of interest that the oxygen consumption was higher in the deeper soils relative to the shallower soils in all phytoreactors. This information demonstrated that while the bottom zone was saturated with groundwater throughout the project it was not completely anaerobic.

While the phytoreactor groundwater supply tank was filled from Well MSB 88C with consistent VOC concentrations (TCE 188 ppb, PCE 55 ppb), the phytoreactors actually received groundwater through the supply system containing around 46 ppb TCE and 48 ppb PCE (Figures 2 & 3). These groundwater VOC losses within the system are most likely due to volatilization. While no statistically significant difference exists between Phytoreactors when comparing influent and effluent groundwater PCE and TCE concentrations, this evaluation did not take into consideration the overall water budget. As pointed out in the Results section, in comparing the groundwater influent flow rates from the three Phytoreactors to effluent flow rates, Phytoreactor 2 with the poplars appeared to have the highest influent groundwater input and yet the lowest effluent output measured. This is not surprising since the poplar grew on the average over five feet during the year while the pine grew just over one foot. In addition, the mass of roots from poplar was observed to be much larger and extensive in the Phytoreactors relative to the pine.

D-Area. Baseline characterization studies prior to irrigation showed that TCE levels in the soil were <1 ppb. The pH of the soil ranged from 4.17 - 5.0.1. Soil classification data for each of the 12 test plots is summarized in Table 5. Initial results of the TCE analyses of soil and water sample samples from the D-Area drip irrigation/phytoremediation study are summarized in Table 6.

This project is highly significant in that most work in the phytoremediation area has been associated with significantly higher concentrations of VOC’s (Burken and Schnoor, 1998, Newman et al, 1997, Doty et al., 2000). Most compounds in soil (i.e. contaminants) must be in solution to be affected (absorbed, modified, degraded, sequestered, etc.) by either plants or microorganisms. Thus, water movement and nutrient availability in the rhizosphere is a critical factor as plants take up many times more water than is needed for metabolism and growth. This additional water is transpired through the leaves as the final step in nutrient transport. However, all of this water and compounds dissolved in it (the soil solution) moves through the rhizosphere, where it is subjected to processing by microorganisms before it enters the root. In some instances, the magnitude of microbial transformation of TCE can be significantly larger than plant transformation of TCE (Anderson and Walton, 1995). However, this is not always the case (Nichols et al., 1987; Schnabel et al., 1997). It is likely that both processes are useful in applying phytoremediation technology in TCE and PCE for seepline groundwater remediation. Microbial biodegradation of VOCs to be assessed this next year will focus on the established rhizosphere activities relating to TCE/PCE removal.

Phytoremediation enjoys relatively favorable public acceptance, in part because it is perceived to be "natural" or non-intrusive overall, and lower costs (Shimp et al., 1993). Essentially ambient process conditions and the lack of unsightly mechanical equipment also contribute to public acceptance. In agricultural developments the use of genetically engineered plants has lead to increased crop efficiency and production. Use of genetically engineered plants has also been found to be highly efficient for chlorinated ethene biodegradation (Doty et al., 2000), although this technology has not been demonstrated at SRS.

The novel drip irrigation/phytoremediation field test was initiated 8/31/00. Initial sampling by SRTC has identified the need for several modifications to the process thus far. With the original design, substantial quantities of TCE were being lost by volatilization to the atmosphere from a holding tank. This problem was solved by installing a floating glass bead cover on the surface of the water in the tank. Even with better containment, TCE input to the system with the existing well appears to be insufficient to facilitate the movement of sufficient quantities of TCE to the zone of influence for phytoremediation processes (Table 6). Several corrective actions have been initiated in response to this. New wells have been drilled in regions with higher TCE concentrations than the original well. This will permit the application of much higher water flows with water containing much higher TCE levels to the plots. A series of soil moisture sensors with continuous data acquisition have been installed in one of the plots. This will facilitate measuring the movement of irrigation water throughout a plot without having to rely exclusively on relatively expensive VOC measurements. Ancillary soil column studies are also being designed to assess the movement of TCE when applied to the subject soils at the flow rates and TCE concentrations. Subsequent work will focus on comparatively evaluating and optimizing the ability of various plants to detoxify the contaminant and to determine feasible treatment flow rates and operational costs.

Typically at SRS and other sites much of the VOC groundwater contamination, with the exception of source areas, are in lower (ppb) concentrations (WSRC-TR-00113 1999). The results of this project, with concurrent SRS studies, will enable better predictions of the VOC removal at the seepline. This first year, FY00, represented an initial acclimation phase for soil and plant systems and will facilitate examination of seepline phyto- and bioactivity in subsequent growth season in FY01. Initial results indicate that phytoremediation and MNA have considerable potential for the removal of TCE and PCE in the Tims Branch flood plain and seepline at SRS.

Anderson, T. A., E. A. Guthrie, and B. T. Walton. 1993. Bioremediation in the rhizosphere. Environ. Sci. Technol. 27:2630-2636.

Anderson, T. A. and B. T. Walton. 1995. Comparative fate of 14C-trichloroethylene in the root zone of plants from a former solvent disposal site. Environ. Toxicol. Chem 14:2041-2047.

Balkwill, D. L. 1989 Numbers, diversity, and morphological characteristics of aerobic chemoheterotrophic bacteria in deep subsurface sediments from a site in South Carolina, Geomicrobiol. J. 7:33-52.

Brigmon, R. L., N. C. Bell, D. L. Freedman, and C. B. Berry. 1998. Natural Attenuation of Trichloroethylene in Rhizosphere Soils at the Savannah River Site. J. Cont. Soils. 7:433-453.

Brigmon, R.L., T.A. Anderson, and C.B. Fliermans. 1999. "Methanotrophic Bacteria in the Rhizosphere of Trichloroethylene Degrading Plants". Int. J. Phytoremediation. 1:241-253.

Burken, J.G., and J.L. Schnoor. 1998. Predictive Relationships for Uptake of Organic Contaminants by Hybrid Poplar Trees. Environ. Sci. Technol. 32:3379-3385.

Daniel, W. W. 1978. "Applied Nonparametric Statistics". Houghton Mifflin Co. Boston, MA.

Doty, S. L., T. Q. Shang, A. M. Wilson, J. Tangen, A. D. Westergreen, L. A. Newman, S. E. Strand, and M. P. Gordon. 2000. Enhanced metabolism of halogenated hydrocarbons in transgenic plants containing mammalian cytochrome P450 2E1. Proc. Natl. Acad. Sci. 97:6287-6291

Fletcher, J. S. and R. S. Hegde. 1995. Release of phenols by perennial plant roots and their potential importance in bioremediation. Chemosphere. 31:3009-3016.

Fliermans, C.B., Phelps, T.J., Ringelberg, D., Mikell, A.T., and White, D.C. 1988. Mineralization of trichloroethylene by heterotrophic enrichment cultures. Appl. Environ. Microbiol. 54:1709-1714.

Forster, J. 1995. Determination of Soil pH. In Alef, K, and P. Nammipieri. Methods in Applied Soil Microbiology and Biochemistry. Academic Press.

Hegde, R. S. and J. S. Fletcher. 1996. Influence of plant growth stage and season on the release of root phenolics by mulberry as related to development of phytoremediation technology. Chemosphere. 32:2471-2479.

Nelson, M. J. K., S. O. Montgomery, and P. H. Pritchard. 1988. Trichloroethylene metabolism by microorganisms that degrade aromatic compounds. Appl. Environ. Microbiol. 54:604-606.

Newman, L.A., S.E. Strand, N. Choe, J. Duffy, G. Ekuan, M. Ruszaj, B.B. Shurtleff, J. Wilmoth, P. Heilman, and M. P. Gordon. 1997. Uptake and biotransformation of trichloroethylene by hybrid poplars. Environ. Sci. Technol. 31:1062-1067.

Nichols, T.D., D.C. Wolf, H.B. Rogers, C.A. Beyrouty, and C.M. Reynolds. 1997. Rhizosphere microbial populations in contaminated soils. Water, Air, and Soil Poll. 95: 165-178.

Schnabel, W.E., A.C. Dietz, J.G. Burken, J.L. Schnoor, and P.D. Alvarez. 1997. Uptake and transformation of trichloroethylene by edible garden plants. Wat. Res. 4:816-824.

Shann, J. R. 1995. The role of plants and plant/microbial systems in the reduction of exposure. Environ. Health. Perspect. 103(Supplement 5):13-15.

Shimp, J. F., J. C. Tracy, L. C. Davis, E. Lee, W. Huang, L. E. Erickson, and J. L. Schnoor. 1993. Beneficial effects of plants in the remediation of soil and groundwater contaminated with organic materials. Crit. Rev. Environ. Sci. Technol. 23:41-77.

Vroblesky, D.A., C.T. Nietch, and J.T. Morris. 1999. Chlorinated ethenes from groundwater in tree trunks. Environ. Sci. Technol. 33, 510-515

Walton, B. T. and T. A. Anderson. 1990. Microbial degradation of trichloroethylene in the rhizosphere: Potential application to biological remediation of waste sites. Appl. Environ. Microbiol. 56:1012-1016.

WSRC-RP-96-477. 1996. Jackson, D. Development of a vertical recirculation well system for the A/M Area of the Savannah River Site. WSRC-RP-96-477. Westinghouse Savannah River Company. Aiken, SC.

WSRC-TR-99-00113. 1999. Jerome, K.M., J.L. Simmons, D.G. Jackson, and B.B. Looney. Characterization Report to Support Phytoremediation Efforts for the Southern Sector, Savannah River Site, Aiken, South Carolina.

Figure 1. Phytoremediation System in Southern Sector of A/M area demonstrating groundwater
supply system, three phytoreactors containing pine, poplar, soil control, and effluent collection system.

*Diversity = Difference between #substrates utilized by effluent-influent groundwater bacteria.

Table 1. Ion Chromatography of influent and effluent groundwater samples.

Table 2. Microcosms Established with SED 1 and SED 2

Table 3. Methanogenic activity in Microcosms for SED 1 and 2

Table 4. Phytoreactor soil respiration rates.

Table 6. Initial TCE Distribution Measurements: D-Area Drip irrigation/ Phytoremediation Project