Biofiltration of Residual FertilizerNitrate and Atrazine by Rhizobiummeliloti in Saturated andUnsaturated Sterile Soil ColumnsR E Z A M E H M A N N A V A Z , † , ‡

S H I V O . P R A S H E R , †

N A R O M A R K A R I A N , † A N DD A R A K H S H A N A H M A D * , ‡

Department of Agricultural and Biosystems Engineering,Macdonald Campus, McGill University, 21111 LakeshoreRoad, Ste. Anne-de-Bellvue, Quebec, H9X 3V9 Canada, andCentre de Microbiologie et Biotechnologie, INRS-InstitutArmand-Frappier, 245 Boulevard Hymus, Pointe-Claire,Quebec, H9R 1G6 Canada

This study was undertaken to investigate whethermicrobial bioaugmentation of subsurface soil with subsurfaceirrigation could be used as a biofiltration/biocontroltechnology for agricultural pollutants. Nine Plexiglascolumns, 458 mm long × 139 mm in diameter, were packedwith a sterilized sandy loam soil. Subsurface irrigation,through a controlled water table management system, wasused to deliver bacteria, Rhizobium meliloti A-025, to thesoil and to maintain aerobic (unsaturated) or anaerobic(saturated) conditions in the columns. Nitrate and atrazine,a fertilizer and a corn herbicide, were applied to the soilsurface, and leaching was affected by simulated rainfallevents. The soil and drainage waters were analyzed fornitrate and atrazine residues after each rainfall simulationthroughout the experimental period during which thesoil was kept saturated for a total of 80 days and unsaturatedfor a total of 70 days. The monitoring of transport andsurvival of the implanted bacterial strain (A-025) showedthat subsurface irrigation was successful in introducing andtransporting the bacteria throughout the soil columns.During the saturated period, significantly more (95%probability) nitrate-N leached into the drainage watersfrom the control columns than from the bioaugmentedcolumns; the increase being 450% or more for the abioticcontrol columns. The amount of atrazine that leachedinto the drainage waters during the unsaturated periodwas also significantly more from control columns as opposedto bioaugmented columns, with the increase being 262%.

IntroductionAgricultural chemicals play a significant role in the productionand protection of food and feed. It is estimated that cornyields in the second half of the 20th century quadrupled dueto the use of fertilizers and pesticides (1). However, theyhave also gained notoriety for being one of the majornonpoint sources of groundwater pollution as they are

subjected to drainage, surface runoff, and leaching underirrigation and rainfall. Nitrogen fertilizers are used veryextensively in agriculture. It is estimated that 20-60% of thenitrogen fertilizer applied by farmers is lost through runoff,leaching, and denitrification (2) and that 30-60% of thenitrogen fertilizer applied in Quebec leaches into waterwaysand groundwater (3). The effects of water table management(WTM) on agricultural chemicals are also well documented(4-7). It has been suggested that controlling drainageincreases the exposure time of chemicals to the degradingorganisms and prolongs chemical leaching periods, subse-quently decreasing pollution. For example, it was found thaton farms with WTM systems, where water is pumped intothe field for subirrigation and subsurface drainage methodsare used, nitrate-N losses were decreased substantially from34 to less than 20 kg ha-1 yr-1 (8).

The other agricultural chemical investigated in this study,atrazine (2-chloro-4-(ethylamino)-6-isopropylamino-s-tri-azine), is one of the most extensively used herbicides forcrops such as corn, sugarcane, pineapple, and fruit trees. Itspresence in groundwater and its ecotoxicological impactsare well documented (9, 10). Various processes, such ashydrolysis, adsorption, volatilization, and photodegradation,govern its fate in the environment. However, the primarydissipation of atrazine is known to be through biologicaldegradation at neutral pH and by chemical processes in acidicsoils (11-13). Clay, organic matter, temperature, and pH arealso important factors in the adsorption of atrazine. Adsorp-tion increases as the clay content or organic matter contentof the soil increases, whereas increasing temperature, soilwater content, and pH reverses atrazine adsorption (14).Burkhard and Guth (15) reported that the rate of atrazinedegradation by hydrolysis increases as the adsorption rateincreases. Wenk et al. (16) showed that the rate of atrazineremoval is proportional to soil water content.

In the past decade, the microbial inoculation of soils forpest control and as fertilizer has attracted significant attention(17, 18). Rhizobia are used as inocula in many differentcountries for agricultural purposes because these symbionts,in the form of bacteroids, fix N2 in the roots of leguminousplants such as beans, clover, or alfalfa (19). Although thehallmark of rhizobia is N2 fixation, their ability to carry outdenitrification via nitrate respiration during anaerobic growthhas long been known (20) but never exploited. In fact, Garcia-Plazaola et al. (21) have suggested that free-living rhizobiahave the potential to remove fixed nitrogen from soil throughdenitrification under anaerobic conditions. They showed thatoxygen, nitrate, temperature, moisture, and labile organicmatter availability are the main factors that control deni-trification by rhizobia. Denitrification is enhanced by tem-peratures from 15 to 25 °C and by saturation of pore spaces,which helps to reduce oxygen. Such anaerobic conditionsare easily achieved in flooded soils after rainfall or irrigation.Three laboratories, including ours, have recently reportedon the ability of rhizobia to transform atrazine (22-24). Onthe basis of these rationales, we attempted to evaluate oneof our isolates of Rhizobium meliloti, strain A-025 (25), as abiofilter/biocontrol agent for two agricultural pollutants,nitrate-N and atrazine, using subirrigation for the bioaug-mentation procedure.

Materials and MethodsSoil Column Design and Setup. Nine Plexiglas columns, 458mm long × 139 mm i.d., were packed with a sandy loam soil,S-VI (78% sand, 3% silt, 19% clay, and 3.7% organic matter,pH 6.17), excavated from the Macdonald Campus Farm of

* Corresponding author telephone: (514)630-8819; fax: (514)630-8850; e-mail: darakhshan_ahmad@inrs-iaf.uquebec.ca.

† McGill University.‡ INRS-Institut Armand-Frappier.

Environ. Sci. Technol. 2001, 35, 1610-1615

1610 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 8, 2001 10.1021/es0015693 CCC: $20.00 2001 American Chemical SocietyPublished on Web 03/13/2001

McGill University. The columns had a sampling port on theside (298 mm from the top) and were equipped with a deliveryport at the bottom to supply water and bacterial inoculum.A schematic diagram of the setup is shown in Figure 1.

The soil was well mixed and autoclaved for 1 h at 121 °C,three times each day, for two consecutive days before it waspacked into the columns. A 20-mm sterilized gravel filter(size between 9.5 and 2.36 mm) was first packed at the bottomof the columns. The soil was then packed on top of the filterwith 1.97 kg of soil every 100 mm, for a total of 8.44 kg of soiland a bulk density of 1300 kg m-3. Layers of sterilizedcheesecloth were placed on top of the soil to minimize surfaceerosion during rain simulations. All columns, pipes, and tubeswere sterilized by first washing them in 6.0% sodiumhypochlorite (household bleach) and then rinsing with tapwater that was also used in the experiment for irrigation.

Preparation of Bacterial Inoculum. R. meliloti strainA-025 was grown in 5 mL of TYc (25) as a starter culture andincubated at 29 °C in a controlled environment incubatorshaker (Psycrotherm, New Brunswick Scientific). Four-day-old cultures were inoculated into 300 mL of fresh TYc, thensubcultured into six 2-L flasks after 24 h (each with 1 L offresh TYc), and grown for another 24 h. The bacterial cellsfrom the 1-L cultures (approximately 3.3 × 108 cell mL-1)were pooled and harvested by centrifugation (Dupont modelRC5C, Sorvall Instruments) and plated on TYc and TYct (26)agar plates for microbial population counts and verificationof purity. The collected cells were washed with 0.9% salineand resuspended in 200 mL of sterilized, deionized waterand mixed well by vortexing before being introduced intothe soil columns.

Experimental Design. The three treatments, each intriplicate, were as follows: (i) NA: nitrate, atrazine, andbacterial inoculum; (ii) N: nitrate and bacterial inoculum,and (iii) abiotic control: nitrate, atrazine, and no bacterialinoculum. All soil columns first received 300 mL of tap waterfrom the bottom delivery port (subirrigation), followed by200 mL of cell suspension for those columns assigned forbacterial inoculum (i.e., treatment 1, NA; treatment 2, N).The columns were then saturated from the bottom by adding2400 and 2600 mL of water to the treatment and the abioticcontrol columns, respectively. Eight days after the bacterialaugmentation of the soil columns, 300 mg of calcium nitrate4-hydrate, representing approximately 23 kg ha-1 of residualfertilizer nitrate, and 300 µL of atrazine (1000 ppm stocksolution, 90% active) were uniformly applied to the soilsurface. During the first 44 days after the atrazine and nitrateapplication, 20 mm (450 mL) of water was applied as rainfallsimulation on the 9th day and on every 7th day thereafterso that all columns received a total of 120 mm of simulatedrainfall. This is equivalent to the depth of rainfall that isexpected to occur once every 25 yr in the month of May inMontreal, Canada. To observe the effect of a heavy rainfall,on the 80th day after chemical application, the equivalent

of 60 mm of water (1350 mL) was applied, simulating therainstorm that occurred in Montreal on July 14, 1987. On the80th day after nitrate and atrazine application, the columnswere drained, and three more rain simulations of 40 mmwere applied on the 104th, 112th, and 150th days. The stagesof experimental setup and the amounts and times of waterapplications are given in Table 1.

Sample Collection and Analysis. Water samples werecollected at the bottom of the columns after every simulatedrainfall event for analysis of leached nitrate, atrazine, andmicrobial counts. Soil samples (20 g) were collected throughthe sampling ports on the sides of the columns before eachwater application during the unsaturated period to analyzenitrate and atrazine residues. Nitrate was measured by theSoil Testing Laboratory of the Natural Resources ScienceDepartment of Macdonald Campus of McGill University,using a Quikchem automated ion analyzer. Atrazine analysiswas performed as described by Liaghat and Prasher (2) andMasse et al. (27). Water samples were extracted by mixing200 mL of the sample with 50 mL of methylene chloride ina separatory funnel. The mixture was hand-shaken for 5 min,and the organic layer was collected. This process was repeatedthree times, and the extracts were pooled and evaporated todryness. The residues were then dissolved in 10 mL of hexaneand analyzed by gas chromatography (GC). Soil samples wereextracted by shaking 10 g of soil in 100 mL of methanol for60 min and then filtering them under suction. The filtratewas then evaporated to dryness in a rotary evaporator at 35°C. The residues were dissolved in 10 mL of hexane andanalyzed using a GC. The extraction efficiency of sampleswas estimated to be 88% ( 5% (2). The GC was a Varianmodel 3400 equipped with a TSD detector, an autosampler,and an integrator. The GC column was a 0.53 mm i.d. fusedsilica Megabore DB-5. The detector and injector were keptat 290 and 190 °C, respectively. The column temperaturewas maintained at 150 °C for 10 min and then raised to 180°C at a rate of 2.5 °C min-1. The helium carrier gas flow was15 mL min-1.

Bacterial Population Counts. Serial dilutions of the drainwater, up to 10-4, were spread on TYct agar plates (selectivefor R. meliloti) and incubated at 29 °C for a period of 15 days,while the number of colonies was determined periodically.The plates from the control treatment showed no growth ofbacterial colonies on this selective medium.

Statistical Analysis. Statistical analysis for nitrate andatrazine loss were done using the General Linear Models(GLM) procedure, repeated measures analysis of variance,tests of hypotheses for between subjects effects, and multiplepairwise comparison between treatments using the SASSystem release 6.12 for Windows (SAS Institute, 1989).

Results and DiscussionMicrobial Bioaugmentation. The delivery and implantationof strain A-025 were performed through subirrigation. Thepopulation of strain A-025 in drainage water was determinedon TYct agar plates at different times during the experimentalperiod (Figure 2). The results indicated that more cells weretransported during the saturation period (on the 44th and80th days) in treatment 1 (NA, nitrate and atrazine) than thatof treatment 2 (N, nitrate). On the 104th day, the last day ofthe saturated (anaerobic) period, the number of cells leachedfrom treatment 1 was lower than that observed in treatment2. However, during the unsaturated (aerobic) period (on the112th day), there was a large increase in the populationleached from treatment 1. These results suggest that sub-surface irrigation may have good potential for bacterialbioaugmentation of subsurface soil as the bacterial cells weretransported from the bottom of the soil columns, throughoutthe soil profile, up to the soil surface.

Biofiltration of Applied Nitrate. The biofiltration ofapplied nitrate was investigated in this study by first applying

FIGURE 1. Schematic diagram of a soil column.

VOL. 35, NO. 8, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 1611

the GLM procedure to nitrate-N residues in drainage watersof both saturated and unsaturated periods and in soil, aspresented in Table 2. During the saturated period of theexperiment (the first 80 days), the nitrate-N residues in thedrainage waters of all treatments were not significantlydifferent from each other on the days on which the sampleswere taken (Figure 3). The same can also be said for thesamples of the drainage waters or soil samples taken duringthe unsaturated period (Table 2). The only exception is thenitrate-N residues in soil on day 104. This more or less impliesthat, if we compare the nitrate-N residues in the drainage

waters or the soil samples of the various treatments of thesame day during the saturated or unsaturated periods, wemay arrive at the conclusion that there is no treatment effect.However, we would have arrived at this conclusion by makingtreatment comparisons on specific days, and the impact ofrepeated measurements made over time would not have beenconsidered. Thus, a repeated measures analysis of variancewas carried out to investigate the impact of differenttreatments over time (Table 3). Again, there was no significanttreatment effect, although the nitrate-N residues in drainagewaters during the unsaturated period were significantlydifferent with respect to time. The combined effects oftreatment and time for both saturation periods and in soilwere also not significant.

To investigate this further, the nitrate-N concentrationsin drainage waters and soil over the experimental periodwere summed for each saturation period, and a protectedLSD test was performed to investigate variations among thedifferent treatments by doing multiple pairwise comparisonsof the overall means of each treatment (Table 4). During thesaturated period, the summed nitrate-N concentrations indrainage waters were significantly different (R ) 0.05)between the treatment columns and control. The values inthe control columns were significantly higher than those inthe nitrate only and nitrate plus atrazine soil columns. Thesummed concentration of nitrate (seven measured valuesafter the 1st to the 7th rainfall simulations) in drainage waterswas 18 ppm for the control columns, as opposed to 4.0 ppmfor the atrazine-nitrate and 2.0 ppm for the nitrate soilcolumns (Table 4). These results suggest that anaerobicconditions during the saturated period would have initiateddenitrification in the bioaugmented columns, but not in theabiotic control columns. Since all columns were autoclavedin the beginning and the two types of treatment columnswere bioaugmented with R. meliloti A-025, the differencesin nitrate-N residues in drainage waters can be attributedonly to bioaugmentation. It is possible that the bacterialbioaugmentation could be helping either in the denitrificationprocess or in immoblizing the nitrate residues in soil. Sincethere is no significant increase in the nitrate-N residues insoil after the saturated period, this appears to be less likely.In any case, by looking at the results presented in Table 4,it can be stated with confidence that subirrigation was ableto introduce bacteria into soil (Figure 2) and that this

TABLE 1. Experimental Stages and Schedule of Simulated Rainfall Events

treatment

stage application 1 2 3

1 soil packing (kg/column) 8.443 8.443 8.4432 subirrigation (mL) 300 300 3003 subsurface bacterial inoculation (mL) 200 200 0.04 subirrigation (mL) 2400 2400 26005 surface application:

atrazine (mL of 1000 ppm) 0.3 0.0 0.3nitrate (mg) 300 300 300

Stage 6: Rain Simulations

daywater

applied (mL)rainfall

(mm)soil

condition

9 450 20 saturated16 450 20 saturated23 450 20 saturated30 450 20 saturated37 450 20 saturated44 450 20 saturated80 1350 60 saturated

104 900 40 unsaturated112 900 40 unsaturated150 900 40 unsaturated

FIGURE 2. Leached population of bacterial cells, strain A-025, inthe drain water during the experiment (cfu, colony forming unit).

TABLE 2. Results of General Linear Model Procedures forNitrate and Atrazine Concentrations at Different ExperimentalPeriods

Pr > F

nitrate-N atrazine

period daydrainage

water soildrainage

water soil

9 0.3864 0.369316 0.9959 0.116023 0.8622 0.5882

saturated 30 0.5053 0.914137 0.4229 0.233744 0.3859 0.648180 0.1524 0.3800

104 0.6292 0.0204 0.2580 0.9912unsaturated 112 0.9541 0.7844 0.0047 0.2190

150 0.3075 0.4843 0.3245 0.4307

1612 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 8, 2001

introduction significantly reduced nitrate-N pollution overthe saturated period of the experiement. Zablotowicz et al.(20) and Garcia-Plazaola et al. (21) have suggested thefollowing sequence as the likely pathway for denitrificationin anoxic environments by different species of rhizobia:nitrate ion (NO3) f nitrite ion (NO2) f nitric oxide (NO) fnitrous oxide (N2O) f nitrogen gas (N2). However, there isno report of these bacteria oxidizing nitrite ion (NO2) backinto nitrate ion (NO3), or atrazine-NH3 into NO3-, in a

nitrification process, under oxic conditions. It is thereforedifficult to explain the higher concentration of summednitrate-N in the drainage water of treatment 1 in the presenceof atrazine.

Under unsaturated (aerobic) conditions, the summednitrate-N concentrations (3 measured values after threerainfall simulations on days 104, 112, and 150) in the drainagewaters, contrary to those measured during the saturatedperiod, were not significantly different, nor were those in the

FIGURE 3. Concentration of nitrate-N in the drainage waters during (a) the saturated period, (b) the unsaturated period, and (c) in theunsaturated soil after the saturated period. Results represent the mean and SD of three replicates. NA, treatment 1 (bacteria, nitrate, andatrazine), N, treatment 2 (bacteria and nitrate).

VOL. 35, NO. 8, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 1613

soil (Figure 3b; Table 4). This is to be expected sincedenitrification is known to take place under saturatedconditions, and there are significant differences in thesummed nitrate-N concentrations between bioaugmentedand control soil columns.

Biofiltration of Applied Atrazine. Atrazine residues indrainage waters and soil were analyzed along the same linesas the nitrate-N. First the GLM procedure was used tocompare atrazine concentrations in different treatments(Table 2) found on specific days during the saturated andunsaturated periods (Figure 4). Since the measurements wererepeated several times during each saturation period, arepeated measures analysis of variance was also applied tothe measured data. And last, summed atrazine levels fromeach treatment and each experimental period were analyzedby using the protected LSD test in multiple pairwisecomparisons of the overall means of various treatments.

The results of the GLM procedure are given in Table 2.Like nitrate-N, there is no statistical difference in atrazineconcentrations in drainage waters or in soil on measurementdays during both saturated and unsaturated periods, withthe exception of day 112 during the unsaturated period. Therepeated measures analysis of variance was applied next tostudy the effect of different treatments over time. For thesaturated period, there is no treatment effect. In general,there is little microbial activity under saturated conditions,and thus these results are to be expected. The atrazineconcentrations significantly vary over time in all treatments,and this is also to be expected since atrazine will be gettingsorbed/desorbed onto soil particles or organic matter overtime, thus variably leaching out of soil. However, during theunsaturated (abiotic) period, there is a significant treatmenteffect and a highly significant time and treatment × timeeffect (Table 3). Since all columns were autoclaved in thebeginning and there is a significant treatment effect, it clearlymeans that bioaugmentation has worked and that thebacterial strain used in the study, R. meliloti A-025, is causingatrazine degradation in a significant way (Figure 4) therebyleading to significantly different atrazine concentrations indrainage waters during the unsaturated period.

Like nitrate-N, atrazine residues were also summed overtwo saturation periods, and the analysis results are given inTable 4. The protected LSD test was applied to perform themultiple pairwise comparisons of the overall means among

different treatments. As expected, while the summed atrazineconcentrations (seven measured values after the 1st to the7th rainfall simulations) did not vary significantly during thesaturated (anaerobic) period of 80 days, the values in the NA

TABLE 3. Repeated Measures Analysis of Variance of Nitrate-N and Nitrate-N and Atrazine Treatments (r ) 0.05)

Pr > F

nitrate-N atrazine

drainage water soil drainage water soil

source saturated unsaturated unsaturated saturated unsaturated unsaturated

treatment 0.3030 0.2719 0.2290 0.2616 0.0269 0.1471time 0.2603 0.0015 0.1457 0.0242 0.0007 0.1762time × treatment 0.5284 0.5568 0.8953 0.5991 0.0069 0.7171

TABLE 4. Summed Drainage Waters and Soil Concentrations of Nitrate-N and Atrazine during Different Periods of the Experimenta

concn during different periods

chemical treatmentsaturated

(drainage water)unsaturated

(drainage water)unsaturated

(soil)

nitrate-N (ppm) 1 (NA) 4 ( 3a 109 ( 11 2.6 ( 0.62 (N) 2 ( 1a 98 ( 25 2.3 ( 0.83 (control) 18 ( 14b 86 ( 10 1.7 ( 0.4

atrazine (ppb) 1 (NA) 5.9 ( 0.9 4.2 ( 0.8c 19 ( 43 (control) 7 ( 1 11 ( 3d 25 ( 4

a Different letters indicate significant difference between treatments (R ) 0.05) based on the protected LSD test in the multiple pairwise comparisonsof overall means among treatments. a and b for nitrate-N, c and d for atrazine. NA stands for nitrate and atrazine treatment, and N stands for nitratetreatment.

FIGURE 4. Concentration of atrazine in the drainage waters during(a) the saturated period, (b) the unsaturated period, and (c) in theunsaturated soil following the saturated period. Results representthe mean and the SD of three replicates.

1614 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 8, 2001

treatment (Table 4) were significantly lower than those inthe control during the drainage events that took place in theunsaturated (aerobic) period of 70 days. The concentrationswere 11 ppb for the control and 4.2 ppb for the bioaugmentedsoil columns (Table 4), with the largest difference occurringon the 112th day (Figure 4b). During this aerobic period,atrazine concentration in the bioaugmented soil sampleswas also lower (Figure 4c), and the bacterial populationshowed a large increase (Figure 2). The summed concentra-tion of atrazine in the drainage waters from control soilcolumns was 262% higher than that from bioaugmentedcolumns. These results reaffirm that R. meliloti, strain A-025,may be used to decontaminate soils that are contaminatedwith atrazine.

The loss of atrazine in bioaugmented soil columns mighthave been different if nitrate had not been added or if aerobicconditions had been maintained over the entire experimentalperiod. Wilber and Parkin (28) and Crawford et al. (12) haveshown that the transformation of atrazine is decreased inthe presence of nitrate, and Topp et al. (29) reported thatatrazine was not degraded under anaerobic or denitrifyingconditions. Many studies have investigated the transport andfate of agricultural chemicals through saturated and unsat-urated soil profiles (30-32). Saturated conditions are theworst-case scenario for chemical movement in groundwatersystems (33). Therefore, to maximize the leaching of nitrateand atrazine, the initial stage of this experiment wasperformed using saturated soil columns. In these worst-casescenarios, with rain simulations of 20 mm, low concentrationsof atrazine were detected in drainage waters. Atrazine hasbeen shown to have very low leachability through soilcolumns (2, 34). Smith et al. (35) suggested that long periodsof water application are required in order to affect the atrazineconcentration in a soil profile. Therefore, the higher con-centrations of both nitrate and atrazine observed in thedrainwater collected after the 80th day (onset of theunsaturated period) might have resulted from the increaseddepth of rain simulations applied after this date.

In conclusion, subsurface irrigation was successful inintroducing and translocating bacteria in the soil columns.Under saturated conditions, the strain A-025 significantly(95% probability) reduced nitrate-N leaching as comparedto the control. The difference in nitrate-N leaching from thetwo bioaugmented treatments however was not significant.Furthermore, the atrazine loadings from bioaugmented soilcolumns were significantly (95% probability) less than thatfrom the control columns under unsaturated conditions.Other transformation products of atrazine were not measuredin this study, and the soil used had been sterilized and wastherefore devoid of detectable soil microflora and plants.Hence, until further investigations are conducted, these datashould not be extrapolated to predict the fate of nitrate-Nor atrazine in agricultural systems. The overall results of thisstudy indicate that bioaugmentation of agricultural soils withecotoxicologically and ecopathologically safe and suitablebacterial strains, such as Rhizobium, using a WTM-basedsubsurface bioaugmentation system to biofilter/biocontrol/biodegrade agrochemicals such as nitrate and atrazine beforethey reach aquatic systems, may be a feasible, effective, andsustainable solution for the reduction of leaching farmpollutants.

AcknowledgmentsThis work was partly supported by NSERC research grantsto D.A. and S.O.P. and by NSERC and INRS graduatefellowships to R.M.

Literature Cited(1) McRae, B. The characterization and identification of potentially

leachable pesticides and areas vulnerable to groundwater

contamination by pesticides in Canada; Agriculture Canada,Pesticide Directorate: Ottawa, Canada, 1989; p 37.

(2) Liaghat, M.; Prasher, S. O. Am. Soc. Agric. Eng. 1996, 39, 1731-1738.

(3) Miller, P. L.; Mackenzie, A. F. Can. J. Soil Sci. 1978, 58, 153-158.

(4) Evans, R. O.; Skaggs, R. W.; Gilliam, J. W. J. Irrig. Drain. Eng.1995, 121, 271-276.

(5) Madramootoo, C. A.; Dodds, G. T.; Papadopoulos, A. J. Irrig.Drain. Eng. 1993, 119, 1052-1065

(6) Kalita, P. K.; Kanwar, R. S. Am. Soc. Agric. Eng. 1993, 36, 413-422.

(7) Munster, C. L.; Skaggs, R W.; Pemmireddy, V. R. Am. Soc. Agric.Eng. 1996, 39, 55-66.

(8) Skaggs, R. W.; Breve, M. A.; Gilliam, J. W. Crit. Rev. Environ. Sci.Technol. 1994, 24, 1-32.

(9) Solomon, K. R.; Baker, D. B.; Richards, R. P.; Dixon, K. R.; Klaine,S. J.; Lapointe, T. W.; Kendall, R. J.; Weisskopf, C. P.; Gidding,J. M.; Giesy, J. P.; Hall, L. W.; Williams, W. M. Environ. Toxicol.Chem. 1996, 15, 31-76.

(10) Baturo, W.; Lagadic, L.; Caquet, T. Environ. Toxicol. Chem. 1995,14, 503-511.

(11) Blumhorst, M. R.; Weber, J. B. Pestic. Sci. 1994, 42, 79-84.

(12) Crawford, J. J.; Sims, G. K.; Mulvaney, R. L. Appl. Microbiol.Biotechnol. 1998, 49, 618-623.

(13) De Souza, M. L.; Newcombe, D.; Alvey, S.; Crowley, D. E.; Hay,A.; Sadowsky, M. J.; Wackett, L. P. Appl. Environ. Microbiol.1998, 64, 178-184.

(14) Harris, C. I.; Warren, G. F. Weeds 1964, 12, 120.

(15) Burkhard, N.; Guth, J. A. Pestic. Sci. 1981, 17, 241-245.

(16) Wenk, M.; Baumgartner, T.; Dobovsek, J.; Fuchs, T.; Kucsera, J.;Zopfi, J.; Stucki, G. Appl. Microbiol. Biotechnol. 1998, 49, 624-630.

(17) Buchenauer, H. J. Plant Dis. Prot. 1998, 105, 329-348.

(18) Bashan, Y. Can. J. Microbiol. 1998, 44, 168-174.

(19) Zahran, H. H. Microbiol. Mol. Biol. Rev. 1999, 63, 968-989.

(20) Zablotowicz, R. M.; Focht, D. D. J. Gen. Microbiol. 1979, 111,445-448.

(21) Garcia-Plazaola, J. I.; Becerril, J. M.; Arrese-Igor, C.; Gonzalez-Murua, C.; Aparicio-Tejo. P. M. Plant Soil 1993, 157, 207-213

(22) Bouquard, C.; Ouazzani, J.; Prome, J.; Michel-Briand, Y.; Plesiat,P. Appl. Environ. Microbiol. 1997, 63, 862-866.

(23) Topp, E.; Zhu, H.; Lewis, M.; Cupples, D. Abstracts of the AnnualMeeting of Canadian Society of Microbiologists; Guelph,Canada, 1998; p 49.

(24) Labidi, M.; Calveyrac, B.; Mehmannavaz, R.; Chakir, S.; Ahmad,D. Congres de l’ACFAS; Ottawa, Canada, 1999.

(25) Ahmad, D.; Mehmannavaz, R.; Damaj. M. Int. Biodeterior.Biodegrad. 1997, 39, 33-43.

(26) Kinkle, B. K.; Sadowsky, M. J.; Johnston, K.; Koskinen, W. C.Appl. Environ. Microbiol. 1994, 60, 1674-1677.

(27) Masse, L.; Prasher, S. O.; Khan, S. U.; Arjoon, D. S.; Barrington,S. Trans. Am. Soc. Agric. Eng. 1994, 37, 801-806.

(28) Wilber, G. G.; Parkin, G. F. Environ. Toxicol. Chem. 1995, 14,237-244.

(29) Topp, E.; Gutzman, D. W.; Bourgoin, B.; Millette, J.; Gamble, D.S. Environ. Toxicol. Chem. 1995, 14, 743-747.

(30) Kanwar, R. S.; Baker, J. L.; Laflen, J. M. Am. Soc. Agric. Eng. 1985,28, 1802-1807.

(31) Kanwar, R. S.; Baker, J. L.; Baker, D. G. Am. Soc. Agric. Eng. 1988,31, 453-460.

(32) Gish, T. J.; Helling, C. S.; Mojasevic, M. Am. Soc. Agric. Eng.1991, 34, 1699-1705.

(33) Azevedo, A. S.; Kanwar, R. S.; Singh, P.; Pereira, L. S. Am. Soc.Agric. Eng. 1996, 39, 937-945.

(34) Liaghat, M.; Prasher, S. O.; Broughton, R. S. Am. Soc. Agric. Eng.1996, 39, 1329-1335.

(35) Smith, W. N.; Prasher, S. O.; Khan, S. U.; Barthakur, N. N. Am.Soc. Agric. Eng. 1992, 35, 1213-1220.

Received for review August 9, 2000. Revised manuscriptreceived January 30, 2001. Accepted February 2, 2001.

ES0015693

VOL. 35, NO. 8, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 1615