Developments in bioremediation of soils and sediments pollutedwith metals and radionuclides: 2. Field research on bioremediation of metalsand radionuclides

Terry C. Hazen1,* & Henry H. Tabak21Lawrence Berkeley National Laboratory, Virtual Institute for Microbial Stress and Survival, Berkeley, CA,USA; 2US EPA, ORD, National Risk Management Research Laboratory, Cincinnati, OH, USA; (*authorfor correspondence e-mail: tchazen@lbl.gov)

Key words: heavy metal, radionuclide, field test, bioremediation, biotransformation, biodegradation,natural attenuation, treatment train, bioaccumulation, biosorption, volatilization

Abstract

Bioremediation of metals and radionuclides has had many field tests, demonstrations, and full-scale im-plementations in recent years. Field research in this area has occurred for many different metals andradionuclides using a wide array of strategies. These strategies can be generally characterized in six majorcategories: biotransformation, bioaccumulation/bisorption, biodegradation of chelators, volatilization,treatment trains, and natural attenuation. For all field applications there are a number of critical bio-geochemical issues that most be addressed for the successful field application. Monitoring and character-ization parameters that are enabling to bioremediation of metals and radionuclides are presented here. Foreach of the strategies a case study is presented to demonstrate a field application that uses this strategy.

1. Introduction

Bioremediation technology uses microorganismsto reduce, eliminate, contain, or transform to be-nign products contaminants present in soils, sedi-ments, water, or air. Bioremediation is not a newtechnology. Both composting of agriculturalmaterial and sewage treatment of householdwaste are based on the use of microorganisms tocatalyze chemical transformation. Such environ-mental technologies have been practiced byhumankind since the beginning of the recordedhistory. Evidence of kitchen middens and com-post piles dates back to 6000 B.C, and the more‘‘modern’’ use of bioremediation began over100 years ago with the opening of the first bio-logical sewage treatment plant in Sussex, UK, in1891. However, the word ‘‘bioremediation’’ isfairly new. Its first appearance in peer-reviewedscientific literature was in 1987 (Hazen 1997).

The last 15 years have seen an increase in thetypes of contaminants to which bioremediation isbeing applied, including solvents, explosives, poly-cyclic aromatic hydrocarbons (PAHs), and poly-chlorinated biphenyls (PCBs) (McCullough et al.1999; NABIR 2004). Now, microbial processes arebeginning to be used in the cleanup of radioactiveand metallic contaminants, though these contami-nants present special problems since they cannotbe destroyed, only transformed or contained.

There are a number of ex-situ and in-situbioremediation methods currently available (Fig-ure 1). Ex-situ methods have been around longerand are better understood, and they are easier tocontain, monitor, and control. However, in-situbioremediation has several advantages over ex-situ techniques. In-situ treatment is useful forcontaminants that are widely dispersed in theenvironment, present in dilute concentrations, orotherwise inaccessible (e.g., due to the presence of

Reviews in Environmental Science and Bio/Technology (2005) 4:157–183 � Springer 2005DOI 10.1007/s11157-005-2170-y

buildings or structures). This approach can beless costly and less disruptive than ex-situ treat-ments because no pumping or excavation is re-quired. Moreover, exposure of site workers tohazardous contaminants during in-situ treatmentis minimal. Broadly, bioremediation strategiescan be further divided into natural attenuation,biostimulation, and bioaugmentation strategies.Bioaugmentation being the most aggressive, sinceorganisms are added to the contaminatedenvironment. Biostimulation can be aggressive orpassive, in that electron donors, electronacceptors, and trace nutrients can be injected intothe environment to stimulate indigenous organ-isms to increase biomass or activity to affect thecontaminant. Passive biostimulation techniquesinclude simple infiltration galleries or simplyspreading fertilizer on surface without any pump-ing or mixing. Natural attenuation relies on the

intrinsic bioremediation capabilities of that envi-ronment. Environments high in organic carbonand energy sources, low contaminant concentra-tions, and without significant nutrient deficienciesmay be able to degrade or transform the contami-nants of concern without any intervention. Ide-ally, the most cost effective and efficient approachto treat most large contaminant plumes is to usemore aggressive approaches, e.g., bioaugmenta-tion or even excavation and removal, at thesource, grading into natural attenuation at theleading edge, or over time as the contaminantconcentration declines. There are no bioaugmen-tation candidates yet for metals and radionuclidesthat we are aware of. Rarely is a single remedia-tion approach completely effective or costefficient. Indeed, combining aggressive physicaland chemical treatment techniques like chemicaloxidation/reduction, thermal desorption with bio-

Figure 1. Bioremediation technologies.

158

remediation can provide advantages to sometypes of contaminants and allows bioremediationto be an effective polishing or sentinel strategy forthe cleanup.

1.1. Bioremediation strategies for metalsand radionuclides

Over the past few years, interest in bioremedia-tion has increased. It has become clear that manyorganic contaminants such as hydrocarbon fuelscan be degraded to relatively harmless productslike CO2 (the end result of the degradationprocess). Wastewater managers and scientistshave also found that microorganisms can interactwith metals and convert them from one chemicalform to another. Laboratory tests and ex-situbioremediation applications have shown thatmicroorganisms can change the valence, or oxi-dation state, of some heavy metals (e.g., chro-

mium and mercury) and radionuclides (e.g.,uranium) by using them as electron acceptors. Insome cases, the solubility of the altered speciesdecreases and the contaminant is immobilized insitu, e.g., precipitated into an insoluble salt in thesediment. In other cases, the opposite occurs—thesolubility of the altered species increases,increasing the mobility of the contaminant andallowing it to more easily be flushed from theenvironment. Both of these kinds of transforma-tions present opportunities for bioremediation ofmetals and radionuclides—either to lock them inplace, or to accelerate their removal. Microor-ganisms can do much more than biotransformcontaminants. They can also influencecontaminant behavior by changing the acidity ofthe system in the immediate vicinity of thecontaminant, or by altering the form of organiccompounds that influence radionuclide and metalmobility (Figure 2 & 3). A significant number of

Figure 2. Abiotic and biotic mechanisms that influence the fate of metals in the subsurface.

159

field studies and successful deployments usingbioremediation strategies for metals andradionuclides have occurred recently. Under thegeneral categories of biotransformation, bioaccu-mulation/biosorption, biodegradation of chela-tors, volatilization, treatment trains & naturalattenuation we will provide some example casestudies for a variety of metals and radionuclides.Detailed examples of laboratory studies andfundamental research on the nature of bioreme-diation of metals and radionuclides can be foundin the companion chapter in the book by Tabaket al. (2005).

1.2. Critical biogeochemistry

The state and fate of metals in all environmentsis highly dependent on the redox or valence stateof the metal. The redox potential of theenvironment will control the direction of chemi-cal equilibria and whether the metal is reducedor oxidized. This in turn controls the possiblecompounds that the metal can form and therelative solubility of these metals in the environ-ment. To stimulate microbes to produce condi-tions that are appropriate for remediation ofspecific contaminants requires a through knowl-edge of the geochemistry of that environment.Since electron acceptors vary greatly as to theenergy that can be derived from their use in

respiration, the most common terminal electronacceptors (TEA) will be utilized in a set order,according to the energy that can be derived(Figure 4). Thus, oxygen is the preferred TEAand first TEA to be utilized, followed by nitrate,iron(III), sulfate, and carbon dioxide. Sincereduction of Cr and U is not favored until theredox potential is in iron reducing conditions,these two TEAs would have to be depleted first.Indeed, for sites that also have PCE/TCE theiron(III) and the sulfate would have to bedepleted before sustained methanogenesis andsubsequently halorespiration can occur. For fieldapplications, this means that enough electrondonor would have to be added to deplete all theoxygen and nitrate present, at a minimum. Bymonitoring the TEA and their daughterproducts, it provides an excellent measure of theredox conditions at the site and the potential fordegradation of the contaminants of concern.

As an example of the importance of pH andEh, Uranium forms a number of differentcompounds in the environment with U(VI) beinggenerally more soluble and thus mobile thanU(IV). As seen in Figure 5 the U(IV) stabilityregion extends to higher Eh as pH decreases.Thus U bioreduction for immobilization of U(IV)is expected to be more sustainable under slightlyacidic conditions. The trade off is that the U(VI)sorption isweaker andU(IV) is soluble atpH<4.

Figure 3. Three possible non-reducing mechanisms of bacterial influence on U(VI) geochemistry.

160

1.3. Characterization and monitoringconsiderations

The success of any bioremediation applicationwill be highly dependent on the characterizationand monitoring that is done before and duringthe field deployment. For any field remediation,

the first step is to form a conceptual model ofthe contaminant plume in the environment andhow that environment effects that plume. Theuncertainties in this conceptual model providethe drivers for the characterization and monitoringneeds. For example, characteristics of the aquiferwill have a profound impact on the remediation

Figure 5. pH-Eh U predominance diagrams for equilibrium with amorphous UO2, (a) without Ca2+, and (b) in equilibrium with

calcite. UC, UDC, and UTC denote UO2CO3�, UO2(CO3) 22�, and UO2(CO3) 3

4�, respectively.P

U(VI) = 10)6 M. Boundariesare shown for PCO2 ¼ 10�1:5 atm (black) and 10)3.5 atm (red). The yellow rectangular area denotes conditions at the Oak RidgeY-12 Field Research Center (Tokunaga et al., in press).

Figure 4. Competing electron acceptors.

161

strategy (Table 1). The largest part of the expense ofany remediation project is the characterization andmonitoring. Hydraulic conductivities can have asevere effect on your ability to deliver nutrients tothe subsurface (Table 2) and can be the most limit-ing part of the environment. However, as discussedabove if bioreduction was the strategy for a metalcontaminated site and the site had a hydraulicconductivity of only 10)8 cm/s with very highnitrate and sulfate levels and high pH it may not becost effective to use bioreduction at this site. Theseissues also suggest why bioaugmentation has notlived up to its hope. Though bioaugmentationpromises ‘‘designer biodegraders,’’ it has not provento be better then biostimulation in repeated fieldtrials over the last 2 decades. Indeed, there is onlyone bacterium that has demonstrated that it canperform better then biostimulation in situ on someoccasions, Dehalococcoides ethenogenes for deha-lorespiration of chlorinated solvents. At least twoproducts are commercially available and have beenwidely used in theUnited States that are proprietarystrains of this organism (Regenesis andGeosyntec).

We suspect the reason that this microbe has beensuccessful is that it is a strict anaerobe, chlorinatedsolvent dehalorespiration requires establishedmethanogenic redox potentials, and the organism isvery small irregular coccus (0.5 lM) so it canpenetrate the subsurface more easily (Loffler et al.2000). Patchy distributions of this organism innature are also common, so bioaugmentation mayprovide a couple of advantages. Fortunately, newadvances in geophysics and hydraulic push technol-ogy (Geoprobe) has enabled us to characterize sitesin a fraction of the time and cost. Once we haveestablished the hydrology and basic geochemistry atthe site and used that data to refine our conceptualmodel, a base line characterization of themicrobiol-ogy is essential to establish that the right microor-ganisms are present, that they can be stimulated,and that no undesirable reactions with the stimu-lants or daughter products from the stimulation willoccur. This usually requires some treatability andsoil compatibility studies and monitoring ofmicrobial community structure and function toestablish the base conditions prior to stimulation(Plaza et al. 2001). For example, some metals likearsenic actually increase solubility under the sameredox potentials that precipitate Cr and U. Table 3provides an example list of the types of measure-ments that should be performed from eithertreatability slurries, soil columns or in-situ sampling(Hazen 1997). This data and the refined conceptualmodel provide the functional design criteria for the

Table 1. Aquifer profile

Site characteristics Impact on remediation

(a) Soil type (a) Level of difficulty

Homogeneity

Permeability

Chemistry

(b) Aquifer type and use (b) Remediation goals, urgency, level of difficulty, treatment strategy

Confined, perched

Drinking water, agriculture, etc.

(c) Groundwater flow (c) Urgency

(d) Sustainable pumping rate (d) Duration

(e) Water table location (e) Level of difficulty

Current depth to water

Depth to water

Water table fluctuation (seasonal and extreme)

(f) Recharge (f) Level of difficulty, treatment strategy

Location

Seasonal rainfall

Table 2. Critical hydraulic conductivities for bioremediation

Minimum gas injection 10)9 cm/s

Minimum liquid injection 10)7 cm/s

Minimum particle injection 10)5 cm/s

*General values for consideration of applicability of differentinjection strategies for bioremediation.

162

remediation and can be used to develop a numericalmodel to predict the remediation rates, stability andlegacy management needs, e.g., monitoring,especially if the remediation is an immobilizationstrategy.

2. Biotransformation (bioreduction

and biooxidation)

Unfortunately, metals and radionuclides cannotbe biodegraded. However, microorganisms caninteract with these contaminants and transformthem from one chemical form to another bychanging their oxidation state through the add-ing of (reduction) or removing of (oxidation)electrons. In some cases, the solubility of thetransformed metal or radionuclide increases, thusincreasing the mobility of the contaminant andallowing it to more easily be flushed from theenvironment. In other cases, the opposite willoccur, and the transformed metal or radionuclidemay precipitate out of solution, leading toimmobilization. Both kinds of transformationspresent opportunities for bioremediation of metaland radionuclides in the environment—either toimmobilize them in place or to accelerate their

removal. All of the field studies and deploymentsto date that we could find have used immobiliza-tion. Mobilization has been traditionally difficultto justify to regulators and stakeholders atremediation sites because it represents an inher-ently greater risk if the capture zone does notadequately capture the mobile and usually moretoxic contaminant. However, since this strategyremoves the contaminant from the environmentit is a better long-term strategy especially whereepisodic changes in the environment due tostorm surges, etc. occur. Another critical issuefor all biotransformation studies are how stablethe transformation is, i.e., does the bioreducedmetal reoxidize with exposure to oxygen orchanges in pH, this especially critical for Ura-nium which has a fairly narrow range where itremains insoluble.

2.1. Case study 1: Aquifer chromium bioreductionat hanford 100 h using polylactate biostimulation

The US Department of Energy produced nuclearmaterials at the Hanford site for more than40 years, Chromium was used to preventcorrosion in the cooling towers at the site and asan oxidizer in the nuclear fuel production

Table 3. Bioremediation characterization and monitoring parameters

Measurements Parameter

Biomass

Viable Counts Plate counts, most probable number (MPN), enrichments, BIOLOGTM

Direct counts Acridine orange direct count (AODC), fluorescien isothiocyanate (FITC),

direct fluorescent antibody (DFA)

Signature compounds Phospholipid fatty acid (PLFA), DNA, RNA, 16S rDNA microarrays

Bioactivity and bioremediation

Daughter products Cl, CO2, CH4, stable isotopic C, reduced contaminants

Intermediary metabolites Epoxides, reduced contaminants

Signature compounds PLFA, ribosome probes, BIOLOGTM, phosphatase, dehydrogenase,

iodophenyl-nitrophenyl, tetrazolium chloride (INT), acetylene reduction,

recalcitrant contaminants

Electron acceptors O2, NO3, SO4 (microrespirometer)

Conservative tracers He, CH4, Cl, Br

Radiolabeled and stable

isotope mineralization

14C-, 13C-, 3H-labeled contaminants, acetate, thymidine

Sediment

Nutrients PO4NO3, NH4, O2, total organics, SO4

Physical/chemical Porosity, lithology, cationic exchange, redox potential, pH, temperature, moisture,

heavy metals

Toxicity MicrotoxTM, MutatoxTM

163

process. Consequently, the site has a large plumeof low concentration Cr(VI) that is impacting theColumbia River. Previous studies by our group(Tokunaga et al. 2001a, 2001b, 2003a, 2003b)demonstrated that simple organic carboncompounds, like lactate, could stimulate ironreducers in the soil to reduce enough Fe(III) toFe(II) that the Fe(II) would reduce Cr(VI) toCr(III) and precipitate in the soil. In August2004, we injected 40 lbs of 13C-labeled polylac-tate into a single well after doing pump tests, tra-cer tests, treatability studies, and base linegeophysics. The complete project design, meth-ods, and results are given at http://esd.lbl.gov/ERT/hanford100h/. Figure 6 shows the Hanfordplume and the injection well design and geology.

The polylactate (Hydrogen Release Com-pound, HRCTM) hydrolyzes to lactate in theaquifer, which is readily utilized by the indige-nous bacteria, the rate of hydrolyzation iscontrolled by the degree of esterization of thepolylactate. The HRC was labeled with 13C sowe could trace via a stable isotope if themicroorganisms that were utilizing the HRC andto measure rates of daughter product formation.Within 2 weeks, the total density of bacteria had

increased more than 2 orders of magnitudefrom<105 cells/ml, to more than 107 cells/ml. Asexpected, the oxygen was depleted first, thenthe nitrate, and then the Fe(III) with a redoxpotential of )130 mv (Figure 7). The sulfatestarted to go down with production of H2S, butwas never depleted. Methane was never detected.Within 3–4 weeks of injection, the Cr(VI) camedown in the monitoring wells and stayed wellbelow the levels for drinking water for severalmonths, even after the nitrate and oxygenreturned to their original concentrations(Figure 7). Microbial community analyses with16S rDNA microarrays for the entire knownribosome database showed that the diversityincreased dramatically. Analysis of thecommunity structure after the injection showedan increase in denitrifiers, followed by increasesin iron reducers and sulfate reducers. Eventhough nitrate was depleted, and iron was re-duced, sulfate depletion as a TEA was nevercomplete and subsequently methanogens werenot observed in any of the samples. After2 months, the bacteria densities slowly returnedto their original densities prior to stimulationwith HRC. Drilling and pump tests are being

Figure 6. Field injection of HRC for Cr(VI) bioreduction.

164

Figure 7. Biogeochemical changes after HRC injection.

165

carried out now to determine if the attenuationseen in the geophysical measurements that coin-cided with the reduction in Cr(VI) was caused byCr(III) precipitates. Long-term studies at this sitewill be necessary to determine if reducing condi-tions must be maintained to prevent reoxidationof Cr(III) to Cr (VI).

2.2. Case study 2: Using mercury resistantbacteria to treat chloralkali wastewater

Mercury resistant bacteria have been used recentlyto detoxify Hg(II)-contaminated water at lab andpilot scale (Figures 8a and b). Wagner-Dobleret al. (2000a) at the German Research Centre for

Figure 8. Flow scheme (a) and photograph (b) of pilot plant for removal of mercury from wastewater by mercury resistantbacteria. Flow scheme (a) and photograph (b) of a pilot plant for removal of mercury from wastewater by mercury resistantbacteria. The plant includes pH adjustment to pH 7, nutrient amendment, the bioreactor (volume 1 m3), a buffering tank and apolishing carbon filter. Continuous automated Hg measurement is performed at the inflow, after the bioreactor and at the outflow.pH is measured before and after adjustment to pH 7. Other parameters determined continuously are chlorine concentration (Cl2),oxygen concentration (O2), redox potential (R), conductivity (C) and temperature (T). Figures kindly provided by Dr. IreneWagner-Dobler.

166

Biotechnology in Braunschweig captured reducedelemental Hg in a 20-ml immobilized cellbioreactor, inoculated with a mercury resistantPseudomonas putida, and subsequently colonizedwith other mercury resistant strains (Wag-ner-Dobler et al. 2000b). A companion studydemonstrated successful removal of Hg2+ fromchloralkali electrolysis water at laboratory scale(Von Canstein et al. 1999), prior to developmentof a pilot-plant for Hg(II) removal using thistechnology (Wagner-Dobler et al. 2000c). In thelatter study, a 700-L reactor was packed withpumice granules of particle size 4–6 mm andinoculated with seven mercury resistantPseudomonas species. Acidic wastewater from achloralkali factory was neutralized and amendedwith sucrose and yeast extract prior to introduc-tion into the bioreactor. Concentrations of up to10 mg/L Hg were successfully treated with aremoval efficiency of 95%, although influentspikes above this concentration had a deleterious(if reversible) effect on the reactor performance.When operated in combination with an activatedcarbon filter, which also became colonized bybacteria, further removal of Hg to below 10 lg/Lwas reported. Very high loadings of Hg wereretained in the reactor, conservatively estimated at31.5 kg for the 700-L vessel.

Long-term performance of the reactors hasbeen studied, with no loss of the entrapped Hg(0)from the system over 16 months (Von Canstein etal. 2001). Although the reactors were sensitive tomechanical and physical stresses (e.g., shear fromgas bubbles or increased temperature over 41�C),the system seems robust and able to adjust to ele-vated Hg(II) concentrations (up to 7.6 mg/L)within several days (Von Canstein et al. 2001).With a continuous selection pressure for mercuryresistance, a stable and highly active mercury-reducing microbial community is establishedwithin the bioreactors; confirmed using PCR-based techniques targeting the intergenic spacerregion of 16S–23S rDNA, and a functional genetarget for Hg(II) reduction, merA (Von Cansteinet al. 2001). The performance of the reactor sys-tem has also been studied in response to the oscil-lation of the mercury concentration in thebioreactor inflow (Von Canstein et al. 2002). Atlow-mercury concentrations, maximum Hg(II)reduction occurred near the inflow at the bottomof the bioreactor. At higher concentrations, the

zone of maximum activity migrated to the upperhorizons. Molecular analysis of the microbialcommunities showed an increasing microbialdiversity along a gradient of decreasing mercuryconcentrations (Von Canstein et al. 2002).

2.3. Case study 3: Ex-situ bioremediationof metals using sulfate-reducing bacteria

The ability of sulfate-reducing bacteria to precip-itate metals as insoluble metal sulfides has beenused by Paques BV of the Netherlands (http://www.paques.nl) in ex-situ bioreactors for thetreatment of metal-contaminated water. The pat-ented reactor configurations, marketed under theregistered trademark ‘‘Thiopaq�’’ can also beadapted to treat other waste streams containingsulfur compounds including hydrogen sulfide.

Early development work focused on the BudelZinc BV refinery at Budel-Dorplein in theNetherlands. Over 200,000 tons of zinc are pro-duced annually at the refinery, which has beenoperated since 1973. However, zinc was refined byvarious companies at this site for more than100 years, resulting in contamination of soil andgroundwater with heavy metals and sulfate. In 1992Paques designed and installed a system to treat wa-ter extracted from strategically located wellsaround a geohydrological containment system in-stalled to protect local drinking water supplies(Barnes et al. 1994). The bioreactor system isshown in Figure 9a, with a flow sheet of the processshown in Figure 9b. In the first stage, water is pas-sed to an anaerobic bioreactor containing sulfate-reducing bacteria that couple the oxidation of etha-nol to the reduction of sulfate to sulfide. This leadsto the precipitation of insoluble metal sulfides.

H2Sþ ZnSO4 ! ZnSðprecipitateÞ þH2SO4

Excess toxic sulfide is then oxidized to ele-mental sulfur in an aerobic reactor, and tiltedplate settlers (TPLS) and sand filters are used asfinal polishing steps to remove solids. Metalsulfides and elemental sulfur are returned to theplant for metal recovery and sulfuric acidproduction respectively. Performance of this sys-tem is summarized in Table 4.

Since 1999, this type of technology has alsobeen employed by Budel-Dorplein to treat processstreams containing sulfate and zinc produced bythe conventional roast-leach-electrowin process

167

operated at this site. These streams were previ-ously treated conventionally by neutralizationwith lime, resulting in the production of 18 tons/day gypsum. However, recent legislation prohibitedfurther production of solid residues from July2000. The high rate Thiopaq� biological sulfatereduction bioreactor, supplied with hydrogen asthe electron donor was, however, able to convertzinc and sulfate into zinc sulfide (10 tons/day),which is recycled at the refinery.

Table 4. Performance of Thiopaq� metal/sulphate treatmentprocess at Budel-Dorphein

Component Unit Influent Effluent

Flow m3/h 300 300

Zinc mg/L 100 <0.3

Sulfate mg/L 1000 <200

cadmium mg/L 1 <0.01

Figure 9. Aerial photograph (a) and flow sheet (b) of the Paques BV Thipaq zinc-sulfate treatment process at the Budel Zinc B.V.refinery at Budel-Dorplein in the Netherlands. With Permission from Paques BV.

168

Paques have also used THIOPAQ� to removemetals from an alkaline slag dump leachate atKovohute Pribram lead waste recycling facility inthe Czech Republic. An alkaline carbonate buf-fered sodium sulfate leachate, containing lead,zinc, tin, and high concentrations of arsenic andantimony is treated using hydrogen sulfide that isproduced in a separate bioreactor from the reduc-tion of elemental sulfur (Weijma et al. 2002). Eth-anol is used as the electron donor for sulfurreduction. The hydrogen sulfide is passed intogas–liquid contractors where it reacts with leach-ate that has been acidified by waste battery acid,leading to the precipitation of arsenic and anti-mony as sulfides. In a second stage, the remainingmetal sulfides are precipitated at neutral pH. Pa-ques report that this technology significantly out-performs lime treatment due to the lower solubil-ity of metal sulfides as opposed to hydroxides.

2.4. Case study 4: In-situ uranium bioremediationthrough bioreduction

Laboratory studies of uranium-contaminatedaquifer sediment collected from uranium milltailings remedial action (UMTRA) sites in Colo-rado and New Mexico indicated that acetateaddition stimulated anaerobic conditions and theloss of soluble U(VI) fromsolution(Finneranetal.2002). Loss of soluble U(VI) occurred in live sed-iments only, coincident with Fe(II) productionand prior to observed losses of sulfate (Finn-eran et al. 2002). These results are consistentwith the loss of U(VI) from solution occurringunder stimulated Fe(III)-reducing conditions.More detailed analyses of the stimulated micro-bial community using 16S rDNA-based tech-niques revealed that the stimulated loss of U(VI)from solution occurred as the microbial commu-nity shifted towards organisms known to reduceboth Fe(III) and U(VI). In these studies Geobact-eraceae, known Fe(III)- and U(VI)-reducing,microorganisms were greatly enriched (up to40% of the detected bacterial community) in sed-iments exhibiting a loss of soluble U(VI) relativeto control sediment incubations (Holmes et al.2002). These results indicated that the additionof acetate to the subsurface of uranium-contami-nated aquifers would result in the removal of sol-uble U(VI) from groundwater under Fe(III)-reducing conditions consistent with the know

ability of Geobacteraceae to reduce soluble U(VI)to insoluble U(IV) (Figure 10). This hypothesiswas tested at the field scale at the Old Rifle UM-TRA site in Rifle, Colorado (NABIR 2004).

Acetate addition to the subsurface of the OldRifle UMTRA site stimulated the loss of U(VI)from groundwater. A test plot consisting of anacetate injection gallery composed of 20 injectionwells in tow offset rows of 10 wells each and a to-tal of 18 monitoring wells were installed within a16 m � 24 m portion of the Old Rifle site(Figure 11; Anderson et al. 2003). Initialgroundwater sampling indicated U(VI) concentra-tions of approximately 0.4–1.4 lM, well abovethe established UMTRA contaminant limit of18 lM for this site. A sodium acetate solution(100 mM) containing a bromide tracer (10 mMKBr) was prepared from site groundwater,sparged with nitrogen gas and stored anaerobi-cally under nitrogen pressure (0.1 atm) within astainless steel tank (2081 L capacity) housed with-in a storage shed erected on site. Acetate solutionflowed from the storage tank to a manifold span-ning the entire width of the injection gallery to 60injection ports within the 20 injection wells (3ports per well) delivering acetate to three differentdepths within the saturated subsurface (Andersonet al. 2003). Each injection port was fitted with aflow meter set to provide acetate to the subsur-face at a rate of 1–3 mL/min, which corre-sponded to a calculated volume addition to theaquifer of 1–3% per day (in-situ acetate concen-tration 1–3 mM). Upon the initiation of acetateinjection, soluble uranium concentrations de-creased rapidly within the monitoring well fieldresulting in removal percentage averaging 70% ofinitial concentrations over a period of approxi-mately 50 days. Loss of soluble U(VI) occurredcoincident with the arrival of acetate, the produc-tion of Fe(II) and prior to any observed loss ofsulfate. Furthermore, 16S rDNA-based analysesof the groundwater indicated a microbial commu-nity greatly enriched in Geobacteraceae, up to89% of the detected bacterial community. Phos-pholipid fatty acid (PLFA) analyses of ground-water using ‘‘Geobacteraceae-specific’’ lipids alsoindicated an increase in Geobacter biomass. Theseresults are consistent with the previous laboratorystudies indicating a stimulated removal of solubleU(VI) from groundwater via the in-situ stimula-tion of Fe(III)- and U(VI)-reducing Geobactera-

169

ceae (Anderson et al. 2003; Finneran et al. 2002;Holmes et al. 2002).

Metal-reducing conditions were not sustainedwithin the Old Rifle site over 50 days, and it wasthought that acetate-oxidizing sulfate- reducingbacteria became dominant when Fe(III) was de-pleted. In the vicinity of the injection gallery andthe terminal electron accepting process shifted tosulfate reduction. Indeed, a complete loss of ace-tate (limiting under sulfate-reducing conditions inthis aquifer) was accompanied by a nearly stoichi-ometric loss of sulfate from the groundwater.Analyses of the microbial community detectedwithin the groundwater also indicated a shift froma community dominated by Fe(III)-reducingorganisms to a community dominated by organ-isms known to reduce sulfate, i.e., Desulfobactera-ceae (Anderson et al. 2003). The results stress theimportance of maintaining metal reduction withinthe subsurface or encouraging the growth and

activity of sulfate-reducing bacteria capable ofU(VI) reduction; acetate-oxidizing sulfate-reduc-ing bacteria have not been shown to reduce U(VI),although there is ample evidence that lactate-oxi-dizing sulfate-reducing bacteria are able to reduceU(VI) using lactate or hydrogen as electron do-nors (Lovley & Phillips 1992; Lovley et al. 1993).Thus, addition of these electron donors to the sub-surface may stimulate U(VI) reduction in situ.

3. Bioaccumulation and biosorption

Bioaccumulation and biosorption strategies forremediation of metals and radionuclide contami-nated soil are based on the ability of bacteria andplants to concentrate metals within the cells atconcentrations 1000 s of times higher then theambient concentrations. By forming phosphate–metal, organo–metal or metal–sulfide complexes,

Figure 10. Schematic showing the removal of soluble U(VI) from contaminated groundwater under Fe(II)-reducing conditions,stimulated by the addition of acetate to the subsurface. Courtesy of Jonathan Lloyd.

170

they become insoluble or at least not bioavailableto the target risk group, i.e., less toxic. Biosolidsapplications and phytoextraction have proven tobe possible applications of this approach.

3.1. Use of biosolids for treatment of metalcontaminated soils

3.1.1. Biosolids for remediaton of metalcontaminated soilsThe conventional remedial approach to metal con-taminated soils within the US EPA Superfund pro-gram involves stabilization and replacement of thesoil with clean material or capping the soil with animpermeable material to reduce potential exposureto the contaminants. Standardized tests exist to

evaluate the contaminated soils as well as to mea-sure the success of the remedial action, but the testsare largely engineering based and do not considerecosystem function. Tests commonly used includemeasures of total metal concentrations and of thepotential for metals to leach into groundwater (i.e.,toxicity Characteristic Leaching Procedure, SW-846 Method 1311, Multiple Extractions with dif-ferent molar acid solutions). Human exposure tocontaminated groundwater is the driving factorboth in identifying contaminants of concern aswell as in setting appropriate concentration limits(National Research Council 2003).

Alternative remedial technologies are currentlybeing developed that involve leaving the contami-nated materials in place and using soil amend-

Figure 11. Test plot for U(VI) remediation at the Old Rifle UMTRA site, consisting of and acetate injection gallery composed of20 injection wells and 18 monitoring wells installed with a 16 m � 24 m area. Courtesy of Jonathan Lloyd.

171

ments to reduce the bioavailability to humans(Brown et al. 2004; Ryan et al. 2004). The use ofmunicipal biosolids for restoration of disturbedlands is well documented (Sopper 1993). Recentstudies show that amending soils with municipalbiosolids and lime to reduce the bioavailability ofcontaminants, restores ecological function to soilto enable a vegetative cover for large-scale metalcontaminated sites (Basta et al. 2001; Brown et al.2003a, 2003b; Conder et al. 2001; Li et al. 2000).

3.1.2. BioavailabilityWhile conventional extraction tests have been usedto evaluate the success of in-situ technologies, addi-tional assays are necessary to measure restorationof ecosystem function. In particular, bioavailabilityhas to be more broadly considered to include arange of ecological receptors and relevant pathways(National Research Council 2003).

Several procedures have been developed to mea-sure the bioavailable, rather than total, fraction ofcontaminants in soils and sediments and these aregenerally based on the exposure pathway for themost sensitive ecological receptor. In many cases,toxicity is the defined endpoint. For example, soilextracts are routinely used to determine the phyto-available fraction of total nutrient concentrationsin soils (McLaughlin et al. 2000; Sparks et al. 1996).In cases of contaminated soils, extracts have beenaltered to better mimic the behavior of plants inthese environments. Diffusive gradients in thinfilms have been found to better mimic soluble metalbioavailability for assessing the potential forphytotoxcity (Sauve 2002; Zhang et al. 2001). Ex-tracts have also been correlated with reductions inmicrobial activity, as measure by microbial lux bio-sensors (Shaw et al. 2000; Vulkan et al. 2000).

Direct toxicity tests and animal feeding trails arealso used. Earthworm mortality has been used as ameasure of the effectiveness of soil amendments toreduce bioavailability in mime tailings (Conder etal. 2001). Both in vivo and in vitro extracts havebeen used to predict the bioavailability of soil Pb tohumans (Ruby et al. 1996). In each case, the testwas developed to focus on a particular endpoint orreceptor. None of the tests attempts to evaluate thecollective ecosystem function.

3.1.3. Ecosystem functioningMethods to assess ecosystem function are rare.Techniques have been developed to assess the

health of the soil microbial population, includingmeasure of soil function through respiration,N cycling, and ability to utilize added substrates(Cela & Sumner 2002; Chang & Broadbent 1982;McGrath 2002; Sauve 2002). One example is theBiolog extraction (Kelly & Tate 1998), whichattempts to evaluate the functionality of the soilmicrobial population through its ability to utilize arange of carbon sources. The procedure has beencriticized for difficulty of interpretation, i.e., organ-ism presence can falsely suggest a robust microbialcommunity, and the ability of different groups ofmicrobes to utilize the same substrates (NationalResearch Council 2003). Other tests assess bioavail-ability of contaminants by measuring the reactionsof single types of organisms to exposure to reme-diated soils (Geebelen et al. 2003).

For the amended tailings in Leadville, analternative testing protocol was developed. Inaddition to conventional engineering criteria,standard principles of ecological function wereused to develop a series of tests to assess amend-ment impacts on ecosystem function. For exam-ple, the ability of the system to decomposeorganic matter and recycle nutrients can indicatethe stability of the restored system. In addition,examining the health of, and contaminants from,the amended soil through the food chain.

From the studies on the use of biosolids totreat metal contaminated soil, it is clear thatconventional tests alone will not provide anappropriate assessment of the ability of an in-situamendment to restore ecosystem function tometal contaminated soils. These tests need to becombined with analyses of ecosystem functionand measures of the potential for contaminatedtransfer through the food chain.

3.2. Case study 5: Ecosystem function in alluvialtailings after biosolids and lime addition

Municipal biosolids and agricultural limestonewere incorporated into the surface of alluvial high-ly acidic, metal contaminated mine tailings inLeadville, CO in 1998 (Brown et al. 2005). Amen-ded sites were seeded, and a plant cover subse-quently established. A range of chemical andbiological parameters were measured over time todetermine if treatment was sufficient to restoreecosystem function. An uncontaminated upstreamcontrol (UUC), a contaminated vegetated area

172

(CVA), and soils collected from the tailings depos-its prior to amendment addition were used forcomparison. Standard soil extracts showed de-creases in extractable Pb, Zn, and Cd in the amen-ded soils. Increased CO2 evolution, reduced N2Oand elevated NO3

) in the amended tailings, indi-cated an active microbial community (Table 5).Levels of CO2 and NO3

) were elevated in compar-ison to the CVA and the UUC. Rye grass (Loliumperenne) and earthworm (Eisenia foetida) survival,as well as metal uptake values were similar inamended tailings to a laboratory control soil. Ryegrass and worms in unamended tailings died. Fieldplant diversity was lower in amended areas than inCVA or UUC, with a higher percentage of thevegetative cover consisting of grasses. Small mam-mal analysis showed a low potential for elevatedbody Cd and Pb in the amended tailings. A re-entrainment study using fathead minnows (Pimep-hales promelas) showed no danger for re-sus-pended amended tailings, survival of fish wassimilar to the laboratory control. Data suggest(Figure 12) that ecosystem function has been re-stored to the amended tailings, but these systemsare not yet in equilibrium (Brown et al. 2005).

3.3. Case study 6: In-situ soil treatments toreduce the phyto- and bioavailability of lead,zinc, and cadmium

A study was established near a former Zn and Pbsmelter to test the ability of soil amendments toreduce the availability of Pb, Zn, and Cd in situ.Soil collected from the field was amended in thelab with P added as 1% P-H3 PO4, biosolidscompost added at 10% (referred to hereafter as‘‘compost’’), and a high-Fe by-product (referredto hereafter as ‘‘Fe’’) + P-triple superphosphate(TSP) (2.5% Fe +1% P-TSP) and incubatedunder laboratory conditions at a constant soil pH.Changes to Pb bioavailability were measured withan in vitro test and a feeding study with weanlingrats. Field-amended and incubated soils usingthese plus addition treatments was evaluated usingthe in vitro extraction and tall fescue (Festucaarundinacea Schreb. cv. Kentucky-31) metalconcentration. Reductions were observed acrossall parameters but were not consistent. In thefeeding study, the 1% P- H3PO4 treatment causedthe greatest reduction in vitro extractable Pb fromfield samples (pH 2.2) with a measured reduction

of 66%, while the compost treatment has a 39%reduction and the 2.5% Fe + 1% P-TSP treat-ment a 50% reduction. The in vitro extraction(pH 1.5) run on field samples showed no reduc-tion in the compost or Fe treatments. The 1% P-H3PO4 treatment was the most effective at reduc-ing plant Pb, Zn, and Cd (Brown et al. 2004).

3.4. Case study 7: Using municipal biosolids incombination with other residuals to restore metal-contaminated mining areas

High metal waste materials from historic miningat the Bunker Hill, Idaho (ID) Superfund site wasamended with a range of materials including mu-nicipal biosolids, woody debris, pulp and papersludge, and compost (Brown et al. 2003b). Theexisting soil or waste material elevated metal con-centrations with total Zn, Pb, and Cd rangingfrom 6000 to 14700, 2100 to 27000 and 9 to 28 mgkg)1, respectively. Surface application of certainamendments including biosolids mixed with woodash resulted in significant decreases in subsoilacidity as well as subsoil extractable metals. Thismixture was sufficient to restore a plant cover tothe contaminated areas. At the Bunker Hill site, asurface application of high N biosolids (44 or66 tons ha)1) in combination with wood ash(220 tons ha)1) was able to restore a vegetativecover to the metal contaminated of the vegetationindicated that plans were within normal concen-trations for the 2 years that data were collected.Surface application of amendments was also ableto reduce Ca (NO3)2 extractable Zn in the subsoilfrom about 50 mg kg)1 in the control to less than4 mg kg)1 in two treatments. Use of conventionalamendments including lime alone and microbialstimulants were not sufficient to support plantgrowth. These results indicate that surface appli-cation of biosolids in combination with otherresiduals is sufficient to restore a vegetative coverto high metal mine wastes (Brown et al. 2003b).

Table 5. Microbial Function at the Leadville, CO site

CO2–C respiration Ratio NO3/NH4

Tailings 4.7±0.6 0.01

Upstream 16.9±9 1.1

Control Biosolids

amended tailings

28.2±7.2 12.7

173

3.5. Case study 8: In-situ remediation andphytoextraction of metals from hazardouscontaminated soils

Mining and smelting of Pb, Zn, and Cd oreshave caused widespread soil contamination in

many countries. In locations with severe soil con-tamination, and strongly acidic soil or minewaste, ecosystems are devastated. Research hasshown that Zn phytotoxicity, Pb-induced phos-phate deficiency, Cd risk through uptake by riceor tobacco, and Pb risk to children, livestock or

Figure 12. Leadville, CO Biosolids application comparison 1997–2000.

174

wildlife which ingest soil are the common adverseenvironmental effects at suck contaminated sites.Improved understandings of soil metal risks tothe environment have been developed whichexamine risk to all possible exposed organismsthrough soil, plants, animals, or water exposures.

Soil Cd risk to food-chains only occurs whenCd is present at the usual 0.005–0.02 ratio to Znin the contaminated soil, only rice and tobaccoallow Cd to be transferred from the soil inamounts which can harm humans over theirlifetime. Zn inhibits plant uptake of Cd andinhibits intestinal absorption of Cd, protectinganimals from Cd in most situations. Pb risk tochildren or other highly exposed mammals resultsfrom ingestion of the contaminated soil, andabsorption of Pb from the soil into the bloodwhere adverse health effects occur at 1.0–1.5 lgPb/L blood. Soil Pb has much lower bioavailabil-ity than water Pb, and if ingested with food it haseven lower bioavailability. Research has shownthat if high phosphate levels are added to Pbcontaminated soils, an extremely insoluble Pbcompound, chloropyromorphite, is formed insoils from all known chemical species of Pb whichoccurs in contaminated soils. It had earlier beenlearned that adding adsorbents such as hydrousFe oxides and phosphate to Pb contaminatedsoils inhibited Pb uptake by crops, and combinedwith the evidence that these materials couldreduce the bioavailability of soil Pb to children,feeding tests were conducted with rats and pigs inseveral laboratories (Chaney et al. 1999).

A new approach to remediation of severely dis-turbed Pb/Zn/Cd contaminated soils has beendeveloped which uses mixtures of limestone equiv-alent from industrial byproducts such as wood ash(to make soil calcareous and prevent Zn phytotox-icity), phosphate and Fe from biosolids and by-products (to precipitate Pb and with Fe, increasePb adsorption), organic-N from biosolids andmanures and other beneficial components whichcorrect the infertility of contaminated and erodedsoils. Composting can stabilize the organic matterand slow N release to allow higher test locationswhere this approach was tested (Palmerton in PA;Bunker Hill in ID; Leadville in CO in US, and Ka-towice, Poland). All plants tested were readilygrown on the amended soil even when soils con-tained over 1% Zn and 1% Pb. Plant analysisindicates that these plants may be consumed safely

by wildlife and livestock, although soil ingestionshould be minimized at such sites. Although min-ing and smelting contamination has caused severeenvironmental harm in many locations, this meth-od of soil metal remediation allows effective andpersistent remediation at low cost, and should beapplied to prevent further dispersal of thecontaminated soil materials at many locations.

The potential use of metal hyperaccumulatorplants to phytoextract soil metals is a newmethod of remediation under development.Combining improved cultivars of these accumu-lator plants, agronomic management practices tomaximize yield and metal accumulation, burningthe biomass to generate power, and recovery ofmetals from the ash appear to offer an economictechnology compared to soil removal andreplacement (Chaney et al. 1999).

4. Biodegradation of chelators and biosurfactants

A number of chelators have been used to increasethe solubility of metals and radionuclides andfacilitate desired chemical or microbial reactions.However, when the chelators are disposed of withthe metals/radionuclides or were disposed of asthe chelator–metal complex it makes the metalmore mobile in the environment and potentiallymore toxic. The US Department of Energy used anumber of chelators in nuclear productionprocesses some of which are readily biodegradableand others that are fairly recalcitrant (NABIR2004). Citric acid, nitrilo acetic acid (NTA), andethylene diamine disuccinate (EDDS) are rela-tively biodegradable; while ethylene diaminetetraacetate (EDTA) is much more difficult to bio-degrade (Meers et al. 2005). Several studies haveshown that EDTA is biodegradable, but it doestend to persist longer in the environment than theothers (Bohuslavek et al. 2001; Gorby et al. 1998).We are unaware of any applications that pro-moted chelator degraders for bioremediation of acontaminated site, but it certainly would be a pos-sibility, especially using a bioaugmentation appli-cation ex situ. Biodegradation of chelators is alsoof interest for phytoextraction, since the chelatorpromotes the uptake of the metal in the plant(Meers et al. 2005). Persistence of chelator in therhizosphere could lead to groundwater contami-nation down gradient. Several compounds that

175

are naturally produced by microbes in the soilhave been found to be chelators. Thus, it raisesthe possibility that stimulation of certain micro-bial populations could lead to the production ofchelators and increase metal mobility, either adesirable or undesirable outcome, depending onwhether you are trying to mobilize or immobilizethe metal/radionuclide.

Bacteria can also produce biosurfactants,which combine complexation activity with physicalsequestration of the complexed ions. As a surface-active agent, the biosurfactant can concentratemetals from the soil into critical aggregates thatwill form at critical micelle concentrations a 50 nmparticle that can escape most filtration processes(Meers et al. 2005). Biosurfactants would only beused as a strategy for mobilization of metals com-bined with phytoextraction or pump and treat.

5. Biologically assisted soil washing

and bioleaching

Biologically assisted soil washing for metal reme-diation can occur when indigenous microbes arestimulated via the washing process by addingspecific electron donors or electron acceptors thatwill encourage bacteria that produce acids orsurfactants. Microbial leaching is a process ofpromoting leaching of metals from rock bychemolithotrophs that oxidize iron and sulfides,generating sulfuric acid and thus releasing theassociated metals (Brierley 1982). Bioleaching forcopper is one of the oldest known biotechnologyapplications. One of the earliest records ofthepractice of leaching is from the island ofCyprus. Galen, a naturalist and physicianreported in AD 166 the operation of in situleaching of copper. Surface water was allowed topercolate through the permeable rock, and wascollected in amphorae. In the process of percola-tion through the rock, copper minerals dissolvedso that the concentration of copper sulfate insolution was high. The solution was allowed toevaporate until copper sulfate crystallized. Pliny(23–79 AD) reported that a similar practice for theextraction of copper in the form of copper sulfatewas widely practiced in Spain. This same processalso causes acid mine drainage from mining wasteand abandoned mines. The pH of acid mine wastecan get down less than 1.0 and generate

temperatures in excess of 50�C. Recently the IronMountain mine in California acid mine drainagehas been the sight of some ground breakingresearch in metagenome analysis and metaprote-ome analysis (Ram et al. 2005; Tyson et al. 2004).Because of the extreme nature of this environmentthey were able to sequence and sort out thegenomes of all of the dominant microbes presentwithout ever culturing them, even more recentlythe same group has analyzed directly all theproteins present to determine the relationshipbetween the bacteria in this community (Ramet al. 2005). This promises new insights as to howthis community functions and the interdependen-cies of metabolic pathways and biogeochemistryof this ecosystem.

Zn, Au, Co, and a variety of other metals aremined using bioleaching usually with an irrigationleach or a stirred tank biooxidation (Figure 13). In-deed, over 20% of the world’s copper is producedin this way. Bioleaching can also generate a signifi-cant waste, e.g., final gold extraction process usescyanide. The cyanide/metal waste liquor that re-mains is then treated in a bioreactor to degrade thecyanide (Brierley 1982; Brierley 1990).

Bioleaching has been used in Germany forremoval of metals from dredged sediments(Seidel et al. 2004). About 62% of the Zn, Cd, Ni,Co, and Mn were removed from the sediment in120 days if the sediment was oxic and of goodpermeability. However, only 9% was removedwhen the sediment were freshly dredged andanoxic. Similar processes have been reported in theNetherlands and Switzerland (Tichy et al. 1998).

6. Volatilization

Biovolatilization occurs with several metals thatundergo methylation when they are taken up bythe plant or microbial cell. Unfortunately, thismakes the metal, e.g., Hg and Se, many timesmore toxic than the elemental form. Thus, themain worry of regulators and stakeholders hasbeen that the methylated metal would accumulatein the food chain and potentially be spread over awider area around the treatment zone unless fugi-tive air emissions were trapped. The central valleyof California has a notorious problem with Sebuildup in ponds in wetlands in the area, which haveproventobetoxic for localwildlife. Studies by Terry

176

et al. (De Souza et al. 2001) and Frankenberger etal. (Frankenberger & Arshad 2001; Frankenberger& Karlson 1994, 1995) demonstrated that from 30to 70% of the Se coming into the wetlands arevolatilized. Microalgae and bacteria were shownto be responsible in a number of studies that ad-ded fungicides and bacteriocides to the wetlands

water. The fungicides had no effect on dimethyl-selinite releases, while the bacteriocides greatly re-duced the rate of dimethylselinite production.While a number of investigators have proposedvolatilization as a bioremediation strategy forthese ponds and wetlands because it is as safe asanything else proposed and lot cheaper, the regu-

Figure 13. (a) Irrigation style bioleaching. Crushed copper ore is piled and dilute H2SO4 is trickled through the pile. Bacteria oxi-dize the sulphide, producing acid. As the matrix containing the metal is destroyed the metal dissolves in the acidic solution to pro-duce the ‘‘leach liquor.’’ This liquid is then subjected to electroplating to remove the metal ions from solution. (b) Stirred tankbiooxidation. The slurry of crushed ore and liquid ore is moved from tank to tank to ensure adequate time for arsenopyrite oxida-tion. The last tank is a settling tank where the solids containing gold are removed and extracted using cyanide. Courtesy of JaneWang http://bioteach.ubc.ca/.

177

lators, and stakeholders have not yet funded afull-scale treatment facility.

7. Treatment trains and natural attenuation

Treatment trains or multiple treatmentapproaches are clearly the best way to keep costsdown and remediate a large plume as quickly aspossible. Unfortunately, there are fewdocumented field examples of this, especiallythose that incorporate a physical/chemicaltreatment followed by biostimulation, passive bio-remediation, and finally natural attenuation. Ourunderstanding of reduction/oxidation biogeo-chemistry in the subsurface is still too primitive to

have good predictive models of how even metalslike Cr and U would behave on the long-term.Thus, a monitored natural attenuation, while fea-sible, is practically unheard of. The following twocase studies show multiple component systemsand how they were linked together to solve aproblem and reduce cost, and improve efficiency.

7.1. Case study 9: Algal high-rate pondcombinations for removal of se in agriculturedrainage water

Monthly maximum discharge limits (MMDL)have been established for selenium in irrigationdrainage by the State of California and the US

Figure 14. An algal High-Rate Pond combined with a bacterial Reduction Pond, Dissolved Air Flotation Unit, and a slow sandfilter for removal of Se in Agriculture drainage water.

178

EPA following observations of avian teratogenesisat the Kesterson Reservoir in the San JoaquinValley of California (Green et al. 2003). As a

result of these and other adverse effects, farmers,and drainage districts on the western side of theSan Joaquin Valley must reduce selenium con-

Figure 15. Ex-situ and in-situ treatment train for nitrate removal and in-situ bioreduction of U(VI).

179

centrations in irrigation, drainage discharged tothe San Joaquin River (Green et al. 2003).Drainage treatment will be required in the nearfuture to meet existing MMDL and future totalmaximum discharge limits (TMDL) for the SanJoaquin River.

A pilot-scale 0.4-ha Algal Bacterial SeleniumRemoval (ABSR) Facility was designed andconstructed at the Panoche Drainage District in1995 and 1996 using the Advanced IntegratedWastewater Pond Systems(R) or AIWPS(R)Technology (Green et al. 2003). Each of twophysically identical systems combined aReduction Pond (RP) with a shallow, peripheralalgal High Rate Pond (HRP) (Figure 14). ADissolved Air Flotation (DAF) unit and a slowsand filter were used to remove particulate sele-nium from the effluent of each system. The twosystems were operated under different modes ofoperation and the bacterial substrate varied ineach system. Microalgae were harvested usingDAF and used as a carbon-rich substrate fornitrate- and selenate-reducing bacteria. Massremovals of total soluble selenium of 77% orgreater were achieved over a 3-year period. Ni-trate and selenate were removed by assimilatoryand dissimiliatory bacterial reduction, andnitrate was also removed by algal assimilation.The removal of particulate selenium increasedthe overall removal of selenium to over 90%and would allow farmers and drainage districtsto discharge irrigation drainage in compliancewith regulatory discharge limits. A full-scale sys-tem is currently under construction at the samesite.

7.2. Case study 10: In-situ groundwaterbioreduction of U(VI) with upgradientpreconditioning using fluidized bed bioreactor

At the US Department of Energy Oak Ridge Y12Field Research Center, pilot scale test has beenrun for the last year to demonstrate the utility of atreatment train (Figure 15), both ex-situ and in-situ for treatment of groundwater with low pH(�3.0), high nitrate (11,000 ppm) and high ura-nium (10 lM) (see http://www.stanford.edu/group/evpilot/Reasearch/Oakridge/oakridgepics/oakridgeintro.html). Since U(VI) cannot bereduced when high concentrations of nitrate are

still present and the nitrate would potentiallyproduce enough gas to plug the subsurface if en-ough electron donor was added to deplete the ni-trate, it was decided to create a recirculation zoneand remove a large part of the nitrate in thegroundwater prior to adding enough electron do-nor to reduce the U(VI). The above ground sys-tem removes aluminum, calcium, and nitrate andthen the pH is adjusted to �6.2 to prevent thealuminum from co-precipitating the U and othermetals when the pH is adjusted, the water is thenreinjected to establish an inner treatment zoneand out protection loop. During the first100 days of in-situ biostimulation, the U(VI) con-centration went from 10 lM to less then 3 lM inthe groundwater. Microbial analysis of thegroundwater indicated growth of denitrifiers, sul-fate-reducing bacteria, and iron-reducing bacte-ria. Control of pH/carbonate levels is anengineering tool for the management of U(VI)bioavailability, to limit U(VI) escape from thetreatment zone, and to prevent growth of metha-nogens, which interfere with U(VI) reduction (seehttp://www.lbl.gov/NABIR/generalinfo/annual-mtg/ 05_ann_mtg_pstr1.html)[TCH1].

8. Summary

Bioremediation of metals and radionuclides is afairly new technology for waste site remediation,though bioleaching for metal recovery has beenpractised for nearly 2,000 years. Since metalsand radionuclides are not destroyed, but onlytransformed, it makes both in situ and ex situstrategies more difficult than organic com-pounds. This also makes the biogeochemistry(especially Eh/pH) much more critical for con-trolling the long-term stability and controllingthe remediation process itself. Characterization,monitoring and development of a conceptualmodel are important to minimize errors and todevelop sound functional design criteria for theremediation effort. Bioreduction has a field pre-sence and is starting to show real promise, espe-cially as it relates to immobilization strategies.Biooxidation and mobilization strategies couldbe a better long-term solution since it can becoupled with removal from the environment.Continued research on critical biogeochemistry

180

will provide essential background needed forlong-term confidence by regulators and stake-holders. Bioaccumulation, biosorption, and bio-volatilization have demonstrated utilityespecially with biosolids applications and phy-toextractions. The exact mechnisms of biosolidsapplications are poorly understood and may in-clude both sorption and reduction/oxidation.Together these processes might work quite ni-cely for a longer-term solution that neithercould supply by itself, thus long-term studies inthis area are area also needed. In some casesphytoextraction has been found to be too slowand thus may represent an unacceptable risk;however, application of certain chelating agentsjust before harvesting have been found to makethe risk acceptable. The greatest problem withphytoextraction especially with highly toxic hea-vy metals and radionuclides is not that it won’twork, but rather the secondary waste issues cre-ated by the metal containing biomass. Disposalcosts and overall life cycle costs can keep thistechnology from reaching its full potential.However, composting, extraction process, andincineration, have been found to provide accep-table alternatives at many sites for some metals.Chelators and bio surfactants and their con-trolled biodegradation, is an important nichemarket with some growth potential, especiallywhen linked to other technologies, e.g. phytoex-traction and biooxidation. Biologically-assistedsoil washing and bioleaching of metal-con-taminated dredged sediments show good possi-bilities, but can be limited by organic co-contaminants and may require a treatmenttrain, e.g., dredged sediment ripening beforebioleaching, like is being done in several partsof Europe. Volatilization of most metals, viabiomethylation usually increases metal toxicityand risk and thus may have to be combinedwith fugitive air emission control methods; how-ever, this may be an intiguing solution for avariety of aquatic metal contaminants. Treat-ment trains and natural attenuation are just get-ting started but will probably become dominantstrategies over the next decade. Bioremediationof metals and radionuclides shows great promisebut will benefit greatly from more integrationwith other remediation technologies, and long-term field studies to reach its full potential.Indeed, metals and radionuclides represent our

most retractable contaminant problems thatcould relegate vast areas to become permanentwastelands. The power of control of biologicalprocess that is now emerging from genomicssuggest that bioremediation may be our besthope for remediation of metals and radio-nuclides in the environment.

Acknowledgements

The authors are extremely grateful to Dr. SallyBrown, University of Washington, and Dr. JonLloyd, University of Manchester, who providedgenerous background information about fieldsites they were working on. We are also gratefulto the support of the US DOE Natural andAccelerated Bioremediation Research (NABIR)program in the Office of Biological and Environ-mental Research and the Genomics:GTL pro-gram in the same office. This work wassupported in part by Contract No. DE-AC02-05CH11231.

References

Anderson RT, Vrionis HA, Ortiz-Bernad I, Resch CT, LongPE, Dayvault R, Karp K, Marutzky S, Metzler D.R, PeacockA, White DC, Lowe M & Lovley DR (2003) Stimulating thein situ activity of Geobacter species to remove uranium fromthe groundwater of a uranium-contaminated aquifer. Appl.Environ. Microbiol. 69(10): 5884–5891

Barnes LJ, Scheeren PJM & Buisman CJN (1994) MicrobialRemoval of Heavy Metals and Sulfate from ContaminatedGroundwaters. Emerging Technology for Bioremediation ofMetals, pp. 38–49

Basta NT, Gradwohl R, Snethen KL & Schroder JL (2001)Chemical immobilization of lead, zinc, and cadmium insmelter-contaminated soils using biosolids and rock phos-phate. J. Environ. Qual. 30(4): 1222–1230

Bohuslavek J, Payne JW, Liu Y, Bolton H & Xun LY (2001)Cloning, sequencing, and characterization of a gene clusterinvolved in EDTA degradation from the bacterium BNC1.Appl. Environ. Microbiol. 67(2): 688–695

Brierley CL (1982) Microbiological Mining. Scientific American247: 44–53

Brierley CL (1990) Bioremediation of metal-contaminatedsurface and groundwaters. Geomicrobiol J 8(3): 201–223

Brown S, Chaney R, Hallfrisch J, Ryan JA & Berti WR (2004)In situ soil treatments to reduce the phyto- and bioavailabilityof lead, zinc, and cadmium. J. Environ. Qual. 33(2): 522–531

Brown S, Sprenger M, Maxemchuk A & Compton H (2005)Ecosystem function in alluvial tailings after biosolids andlime addition. J. Environ. Qual. 34(1): 139–148

Brown SL, Chaney R, Halfrisch J & Xue Q (2003a) Effect ofBiosolids Processing on Lead Bioavailability in an UrbanSoil. J. Environ. Qual. 32: 100–108

181

Brown SL, Henry CL, Chaney R, Compton H & DeVolder PS(2003b) Using municipal biosolids in combination with otherresiduals to restore metal-contaminated mining areas. Plantand Soil 249(1): 203–215

Cela S & Sumner ME (2002) Critical concentrations of copper,nickel, lead, and cadmium in soils based on nitrification.Commun. Soil Plant Anal. 33(1–2): 19–30

Chaney RL, Brown SL, Stuchynski TI, Daniels WL, Henry CL,Li Y-M, Siebelec G, Malik M, Angle JS, Ryan JA &Compton H (1999) Progress in remediation of soils contam-inated by mining and smelting of lead, zinc and cadmiumusing tailor-made biosolids mixtures and composts. In:Innovative Clean-up Approaches: Investments in TEchnoogyDevelopment, Results & Outlook for the Future

Chang FH & Broadbent FE (1982) Influence of trace metals onsome soil nitrogen transformations. J. Environ. Qual. 11: 1–4

Conder JA, Lanno RP & Basta NT (2001) Assessment of metalavailability in smelter soil using earthworms and chemicalextractions. J. Environ. Qual 30: 1231–1237

Souza M.P de , Amini A, Dojka MA, Pickering IJ, Dawson SC,Pace NR & Terry N (2001) Identification and characteriza-tion of bacteria in a selenium-contaminated hypersalineevaporation pond. Appl. Environ. Microbiol. 67(9): 3785–3794

Finneran KT, Anderson RT, Nevin KP & Lovley DR (2002)Potential for Bioremediation of uranium-contaminated aqui-fers with microbial U(VI) reduction. Soil Sediment. Contam.11(3): 339–357

Frankenberger WT & Arshad M (2001) Bioremediation ofselenium-contaminated sediments and water. Biofactors14(1): 241–254

Frankenberger WT & Karlson U (1994) Soil-managementfactors affecting volatilization of selenium from dewateredsediments. Geomicrobiol. J. 12(4): 265–278

Frankenberger WT & Karlson U (1995) Volatilization ofselenium from a dewatered seleniferous sediment – a field-study. J. Indust. Microbiol. 14(3): 226–232

Geebelen W, Vvander Lelie D, Mench M, Carleer R, ChijstersH & Vangronsveld J (2003) Selected bioavailability assays totest the efficacy of amendment-induced immobilization oflead in soils. Plant and Soil 249: 217–228

Gorby YA, Caccavo F & Bolton H (1998) Microbial reductionof cobalt(III)EDTA(-) in the presence and absence ofmanganese(IV) oxide. Environ. Sci. Technol. 32(2): 244–250

Green FB, Lundquist TJ, Quinn NWT, Zarate MA, ZubietaI.X & Oswald W.J (2003) Selenium and nitrate removal fromagricultural drainage using the AIWPS (R) technology.Water Scie. Technol. 48(2): 299–305

Hazen TC (1997) Bioremediation. In: Haldeman PAaD (Ed),Microbiology of the Terrestrial Subsurface (pp 247–266).CRC Press, Boca Raton, FL

Holmes DE, Finneran KT, O’Neil RA & Lovley DR (2002)Enrichment of members of the family Geobacteraceaeassociated with stimulation of dissimilatory metal reductionin uranium-contaminated aquifer sediments. Appl. Environ.Microbiol. 68(5): 2300–2306

Kelly JJ & Tate RL (1998) Use of BiOLOG for the analysis ofmicrobial communities from zinc-contaminated soils. J. Envi-ron. Qual. 27: 600–608

Li Y, Chaney R, Siebielec G & Kershner BA (2000) Responseof four turfgrass cultivars to limestone and biosolids compostamendment of a zinc and cadmium contaminated soil atPalmerton, PA. J. Environ. Qual. 29: 1440–1447

Loffler FE, Sun Q, Li J & Tiedje JM (2000) 16S rRNA Gene-Based Detection of Tetrachloroethene-Dechlorinating Des-ulfuromonas and Dehalococcoides Species. Appl. Environ.Microbiol. 66(4): 1369–3445

Lovley DR & Phillips EJP (1992) Reduction of uranium byDesulfovibrio desulfuricans. Appl. Environ. Microbiol.58(3): 850–856

Lovley DR, Widman PK, Woodward JC & Phillips EJP (1993)Reduction of uranium by cytochrome c3 of Desulfovibriovulgaris. Appl. Environ. Microbiol. 59(11): 3572–3576

McCullough J, Hazen TC, Benson SM, Metting FB &Palmisano AC (1999) Bioremediation of metals and radio-nuclides.What it is and how it works (pp. 58). LBNL-42595,Lawrence Berkeley National, Berkeley, California

McGrath SP (2002) Critical Metals. Society of EnvironmentalToxicology and Chemistry

McLaughlin M, Zarcinas BA, Stevens DP & Cook N (2000)Soil testing for heavy metals. Commun. Soil Sci. Plant Anal.31: 1661–1700

Meers E, Ruttens A, Hopgood MJ, Samson D & Tack FMG(2005) Comparison of EDTA and EDDS as potential soilamendments for enhanced phytoextraction of heavy metals.Chemosphere 58(8): 1011–1022

NABIR (2004) Bioremediation of metals and radionuc-lides.What it is and how it works, 2nd ED. (ed. T. C.Hazen), pp. 74. LBNL-42595 2nd Ed., Lawrence BerkeleyNational Laboratory, Berkeley, California

National_Research_Council (2003) Bioavailability of contam-inants in soils and sediments. National Academy of Sciences

Plaza G, Ulfig K, Hazen TC & Brigmon RL (2001) Use ofmolecular techniques in bioremediation. Acta Microbiol. Pol.50(3–4): 205–18

Ram RJ, VerBerkmoes NC, Thelen MP, Tyson GW, Baker BJ,Blake Ii RC, Shah M, Hettich RL & Banfield JF (2005)Community Proteomics of a Natural Microbial Biofilm.Science 308: 1915–1920

Ruby MV, Davis A, Schoof R, Eberle S & Sellstone CM (1996)Estimation of lead and arsenic bioavailability using aphysiologically-based extraction test. Environ. Sci. Technol.30(2): 422–430

Ryan JA, Scheckel KG, Berti WR, Brown SL, Casteel SW,Chaney RL, Hallfrisch J, Doolan M, Grevatt P, MaddaloniM & Mosby D (2004) Reducing children’s risk from lead insoil. Environ. Sci. Technol. 38(1): 18A–24A

Sauve S (2002) Speciation of metals in soil. Society ofEnvironmental Toxicology and Chemistry

Seidel H, Loser C, Zehnsdorf A, Hoffmann P & Schmerold R(2004) Bioremediation process for sediments contaminatedby heavy metals: Feasibility study on a pilot scale. Environ.Sci. Technol. 38(5): 1582–1588

Shaw L.J, Beaton Y, Glover L.A, Killham K & Meharg A.A(2000) Interactions between soil, toxicant, and a lux-markedbacterium during solid phase-contact toxicity testing. Envi-ron. Toxicol. Chem. 19(5): 1247–1252

Sopper W. (1993) Municipal Sudge Use in Land Restoration.Lewis

Sparks DL, Page AL, Helmke PA, Loeppert RH, SoltanpourPN, Tabatabai MA, Johnston CT & Sumner ME (1996)Methods of Soil Analysis: Part 3 Chemical Methods. SoilScience Society of America

Tabak HH, Lens P, van Hullebusch E & Dejonghe W (2005)Developments in bioremediation of soils and sedimentspolluted with metals and radionuclides - 1. Microbial

182

processes and mechanisms affecting bioremediation of metalcontamination and influencing metal toxicity and transport.Reviews in Environmental Science & Bio/Technology 4: 115–156

Tichy R, Rulkens W.H, Grotenhuis J.T.C, Nydl V, Cuypers C& Fajtl J (1998) Bioleaching of metals from soils orsediments. Water Sci. Technol. 37(8): 119–127

Tokunaga T.K, Wan J.M, Firestone M.K, Hazen T.C, OlsonK.R, Herman D.J, Sutton S.R & Lanzirotti A (2003a) In situreduction of chromium(VI) in heavily contaminated soilsthrough organic carbon amendment. J. Environ. Qual. 32(5):1641–1649

Tokunaga T.K, Wan J.M, Firestone M.K, Hazen T.C,Schwartz E, Sutton S.R & Newville M (2001a) Chromiumdiffusion and reduction in soil aggregates. Environ. Sci.Technol. 35(15): 3169–3174

Tokunaga T.K, Wan J.M, Firestone M.K, Hazen T.C,Schwartz E, Sutton S.R & Newville M (2001b) Chromiumdiffusion and reduction in soil aggregates. Environ. Sci.Technol. 35(15): 3169–3174

Tokunaga T.K, Wan J.M, Hazen T.C, Schwartz E, FirestoneM.K, Sutton S.R, Newville M, Olson K.R, Lanzirotti A &RaoW (2003b) Distribution of chromium contamination andmicrobial activity in soil aggregates. J. Environ. Qual. 32(2):541–549

Tyson GW, Chapman J, Hugenholtz P, Allen EE, Ram RJ,Richardson PM, Solovyev VV, Rubin E.M, Rokhsar D.S &Banfield J.F (2004) Community structure and metabolismthrough reconstruction of microbial genomes from theenvironment. Nature 428(6978): 37–43

von Canstein H, Li Y, Leonhauser J, Haase E, Felske A,Deckwer WD & Wagner-Dobler I (2002) Spatially oscillatingactivity and microbial succession of mercury-reducing bio-films in a technical-scale bioremediation system. Appl.Environ. Microbiol. 68(4): 1938–1946

von Canstein H, Li Y, Timmis KN, Deckwer W-D & Wagner-Dobler I (1999) Removal of Mercury from ChloralkaliElectrolysis Wastewater by a Mercury-Resistant Pseudomo-nas putida Strain. Appl. Environ. Microbiol. 65(12): 5279–3445

von Canstein H, Li Y & Wagner-Dobler I (2001) Long-termperformance of bioreactors cleaning mercury-contaminatedwastewater and their response to temperature and mercurystress and mechanical perturbation. Biotechnol. Bioeng.74(3): 212–219

Vulkan R, Zhao FJ, Barbosa-Jefferson V, Preston S, PatonG.I, Tipping E & McGrath SP (2000) Copper speciationand impacts on bacterial biosensors in the pore water ofcopper-contaminated soils. Environ. Sci. Technol. 34(24):5115–5121

Wagner-Dobler I, Von Canstein H, Li Y, Timmis KN &Deckwer W-D (2000a) Removal of mercury from chemicalwastewater by microorganisms in technical scale. Environ.Sci. Technol. 34(21): 4628–4634

Wagner-Dobler I, Von Canstein H, Li Y, Timmis KN &Deckwer W-D (2000b) Removal of mercury from chemicalwastewater by microorganisms in technical scale. Environ.Sci. Technol. 34(21): 4628–4634

Wagner-Dobler I, Lunsdorf H, Lubbehusen T, vonCansteinHF & Li Y (2000c) Structure and species composition ofmercury-reducing biofilms. Appl. Environ. Microbiol. 66(10):4559–4563

Weijma J, Hoop Kde , Bosma W & Dijkman H (2002)Biological conversion of anglesite (PbSO4) and lead wastefrom spent car batteries to galena (PbS). Biotechnol. Prog.18(4): 770–775

Zhang H, Zhao F-J, Sun B, Davison W & McGrath SP (2001)A new method to measure effective soil solution concentra-tion predicts copper availability to plants. Environ. Sci.Technol. 35(12): 2602–2607

183