Environmental Chemistry MICROBIAL TRANSFORMATION of PYRETHROID INSECTICIDES in AQUEOUS and SEDIMENT PHASES

Environmental Toxicology and Chemistry, Vol. 23, No. 1, pp. 1±6, 2004 ᭧ 2004 SETAC Printed in the USA 0730-7268/04 $12.00 ϩ .00

Environmental Chemistry MICROBIAL TRANSFORMATION OF PYRETHROID INSECTICIDES IN AQUEOUS AND SEDIMENT PHASES

SANGJIN LEE,² JIANYING GAN,*² JONG-SIK KIM,² JOHN N. KABASHIMA,³ and DAVID E. CROWLEY² ²Department of Environmental Sciences, University of California±Riverside, Riverside, California 92521, USA ³University of California South Coast Research and Extension Center, Irvine, California 92618, USA

(Received 21 February 2003; Accepted 28 April 2003)

AbstractÐRecent studies showed that synthetic pyrethroids (SPs) can move via surface runoff into aquatic systems. Fifty-six of SP-degrading bacteria strains were isolated from contaminated sediments, of which six were evaluated for their ability to transform bifenthrin and permethrin in the aqueous phase and bifenthrin in the sediment phase. In the aqueous phase, bifenthrin was rapidly

degraded by strains of Stenotrophomonas acidaminiphila, and the half-life (t1/2) was reduced from Ͼ700 h to 30 to 131 h. Permethrin isomers were degraded by Aeromonas sobria, Erwinia carotovora, and Yersinia frederiksenii. Similar to bifenthrin, the t1/2 of cis- and trans-permethrin was reduced by approximately 10-fold after bacteria inoculation. However, bifenthrin degradation by S. acidaminiphila was signi®cantly inhibited in the presence of sediment, and the effect was likely caused by strong adsorption to

the solid phase. Bifenthrin t1/2 was 343 to 466 h for a ®eld sediment, and increased to 980 to 1200 h for a creek sediment. Bifenthrin degradation in the inoculated slurry treatments was not greatly enhanced when compared with the noninoculated system. Therefore, although SP-degrading bacteria may be widespread in aquatic systems, adsorption to sediment could render SPs unavailable to the degraders, thus prolonging their persistence.

KeywordsÐBifenthrin Permethrin Bioavailability Adsorption Biodegradation

INTRODUCTION ments, partitioning into the solid phase could have a profound effect on the transformation rate and hence the persistence of Synthetic pyrethroids (SPs) have been widely used for in- SPs. sect control on agricultural crops, animals, and in households. The main objectives of this study were to identify SP-de- The use of SPs in the United States may increase sharply as grading bacteria from previously contaminated sediments and organophosphate products such as diazinon and chlorpyrifos to compare the degradability of SPs by bacterial isolates in are being phased out for certain uses. Synthetic pyrethroids aqueous systems (BF and PM) and in sediment phases (BF). generally have little or no mammalian toxicity but are effective for controlling many insects when used at low rates [1]. Al- Bifenthrin is a relatively new SP and is characterized by greater though SPs are known for their poor mobility in soil after photostability and insecticidal activity than previous SPs [5,6], application, recent studies have shown that surface runoff may and PM is currently the most widely used SP. Runoff of both transport particle-associated SPs to surface water [2±4]. pesticides to rivers and creeks has been observed in certain Pyrethroids commonly have high acute toxicity to ®sh and watersheds [2,4]. aquatic invertebrates. For instance, the 50% lethal concentration (LC50) of bifenthrin (BF) is 0.07 ␮g/L for Ceriodaphnia MATERIALS AND METHODS dubia and 0.15 ␮g/L for rainbow trout [5,6], while the re- Chemicals and sediments spective values for permethrin (PM) are 0.55 and 9.0 ␮g/L [5,6]. Once in natural water systems, pyrethroids tend to dis- Standards of Z-(cis)-bifenthrin (Ͼ98% purity) and per- appear quickly from the water column due to their strong methrin (20% cis and 78% trans) were purchased from Chem af®nity for sediment [1]. Therefore, the overall risk of SPs in Service (West Chester, PA, USA). Organic solvents used for aquatic systems depends closely on their persistence in the extraction were of pesticide residue or high-pressure liquid sediment phase, which in turn is determined by their suscep- chromatography grade. tibility to microbial degradation. Knowledge of microbial deg- A previously contaminated sediment was obtained from a radation rates should also indicate if bioremediation would be runoff channel at a nursery site in Orange County (CA, USA) feasible for SP-contaminated sediment or water. However, mi- and used for enriching SP-degrading bacteria. Products of BF crobial degradation of SPs has been studied only to a limited and PM had been used at the nursery for over three years prior extent. While earlier studies did show that SP degradation in to the sampling. The moist sediment was stored at 4ЊC before soils was attributable to microbial activity [7±9], only a few use. Two SP-free sediments were collected in the Newport Bay studies attempted to identify SP-degrading bacteria, and soil Watershed in Orange County for use in the biodegradation was the only medium tested [8±11]. Subsequent biodegrada- experiments. A ®eld sediment was taken from a drainage chan- tion was demonstrated only in solid-free microbial cultures nel in an agricultural ®eld, and a creek sediment was collected [9±11]. Because surface waterbodies always contain sedi- from the San Diego Creek near Irvine, California. These sediments were air dried for 48 h and passed through a 0.5-mm * To whom correspondence may be addressed sieve before use. The ®eld sediment contained 86% sand, 7% ([email protected]). silt, 7% clay, and 0.13% organic carbon. The creek sediment

1 2 Environ. Toxicol. Chem. 23, 2004 S. Lee et al. contained 76% sand, 15% silt, 9% clay, and 1.04% organic (BF) or 0.2 ␮g/ml (PM). Twenty milliliters of the spiked so- carbon. lution were transferred to 40-ml screw-top vials. For BF, each vial was inoculated with one of the three strains at 6.5 ϫ 107 SP-degrading bacteria isolation to 3.4 ϫ 108 colony-forming units (CFUs). For PM, vials were Approximately 10 g of the SP-contaminated sediment was inoculated with one of the three selected isolates at 1.8 to 2.6 mixed in 100 ml of mineral salt (MS) medium in a 250-ml ϫ 107 CFUs. All treatments were incubated at room temper- ¯ask, and 10 ␮l of 1.0 ␮g/␮l BF or PM solution (acetone) ature on a platform shaker. At different time intervals, triplicate were added. The MS medium was made by dissolving 0.5 g samples were removed for analyzing SP concentrations. The

K2HPO4, 0. 5 g NaNO3, 0.2 g MgSO4´7H2O, trace FeSO4´7H2O, sample was transferred to a 250-ml separatory funnel and ex- and 15 g agar in 1,000 ml distilled water. The ¯asks were tracted twice with 50 ml of ethyl acetate. The ethyl acetate incubated on a platform shaker at room temperature. After 2 extracts from the same samples were combined, dried with 5 weeks, a 5.0-ml aliquot of the aqueous fraction was removed g of anhydrous sodium sulfate, and then concentrated to near from the culture and transferred to a new ¯ask containing 100 dryness on a rotary evaporator at 60ЊC. The residues were ml fresh MS medium, and was treated with a fresh spike of recovered in 5.0 ml of hexane-acetone (1:1 v/v), and then BF or PM. The same enrichment step was repeated for a total analyzed by GC. Preliminary experiments showed that the of three times. The bacterial growth was monitored by ob- recovery of BF or PM for the above extraction and analysis serving turbidity on a spectrophotometer (Beckman DU 640; procedures was Ͼ90%. Beckman, Fullerton, CA, USA). After serial dilution, 100 ␮l An Agilent 6890N GC system with electron capture detec- of solution was removed and spread onto triplicate plates that tor (Agilent, Wilmington, DE, USA) was used for analysis of were incubated for 10 d at 28ЊC. BF and PM. An Agilent-5 capillary column (30 m ϫ 0.32 mm ϫ 0.25 ␮m) was used for separation, with helium as the carrier Identi®cation of SP-degrading bacteria isolates gas at 2.1 ml/min. The other GC parameters were as follows: The agar plate was divided into four quadrants, and 28 Inlet temperature, 250ЊC; detector temperature, 300ЊC; oven bacterial colonies from one quadrant were subjected to iden- temperature, initially 150ЊC for 1.0 min, ramped to 280ЊCat ti®cation using the MIDI (Microbial ID, Newark, DE, USA) 15ЊC/min, and kept at 280ЊC for 5.0 min; and injection volume, fatty acid methyl ester system [12]. The isolates were streaked 1.0 ␮l. Samples were introduced in the splitless mode. on trypticase soy-broth agar in four quadrants, and the plates were incubated at 28ЊC for 24 h. A loopful of cell material of Biodegradation experiments in sediment slurry late log-phase cells was harvested. Fatty acids were extracted Degradation of BF by the three selected strains was further and methylated according to the procedures given by the man- compared between aqueous systems in the presence or absence ufacturer. Samples were analyzed on a Hewlett-Packard 6890 of a sediment solid phase. Two grams of the ®eld sediment or gas chromatography (GC) (Palo Alto, CA, USA) and were creek sediment were added into 40-ml vials containing 20 ml identi®ed using the MIDI microbial identi®cation software. of MS solution and 10.0 ␮g of BF. Bacterial strains were Three of the fastest growing bacteria isolates were selected inoculated into each treatment at a level similar to that used after BF or PM enrichment and were further subjected to DNA in the aqueous phase experiments. The spiked vials were in- sequence analysis [13,14]. The same strains were subsequently cubated on a platform shaker at room temperature. At different used as inoculants in pure culture degradation experiments. time intervals, triplicate samples were removed and centri- Bacterial DNA was extracted using a FastDNA kit (Bio101, fuged at 2,500 rpm for 20 min to separate the aqueous and Vista, CA, USA), after which the 16S rDNA sequences were solid phases. The supernatant was transferred to a 250-ml se- ampli®ed by polymerase chain reaction using the primer set paratory funnel and extracted twice with ethyl acetate, using 27f and 1492r [14]. The polymerase chain reaction conditions the same procedure as given above. The sediment phase was were 95ЊC for 1 min, 60ЊC for 1 min, and 72ЊC for 1 min for then vigorously mixed with 10 ml acetone:hexane (1:1) and 30 cycles, followed by elongation at 72ЊC for 10 min. The 10 g of anhydrous sodium sulfate on a shaker for 1 h. The ampli®ed polymerase chain reaction products were directly mixture was then centrifuged at 2,500 rpm for 20 min and an ligated into the pGEM-T Easy Vector System II (Promega, aliquot of the supernatant was removed for analysis by GC Madison, WI, USA) and were then transformed into E. coli under conditions given above. Concentrations of BF in the JM109. The rDNA inserts in pGEM-T were sequenced with aqueous and solids phases were combined to calculate the an ABI PRISM Dye Terminator Cycle Sequencing Kit (Ap- overall transformation rate in the slurry system. Preliminary plied Biosystems, Foster, CA, USA). Nucleotide sequence data experiments showed that the recovery of BF for the above were submitted to the GenBank nucleotide sequence databases extraction and analysis procedure was Ͼ90%. (National Center for Biotechnology Information, Bethesda, Following the addition of BF to the culture medium, the MD, USA) and were assigned the accession numbers bacterial cell densities were determined periodically using the AY171915, AY171916, AY171917, AY172967, AY172968, dilution plate counting method. The cell counts from three and AY172969. replicate plates from each treatment were used to determine The bacteria selected for BF and PM degradation experi- the cell densities as CFUs. ments were also examined using electron microscopy. Cell ®xation was achieved by staining with 2% uranyl acetate. Elec- RESULTS AND DISCUSSION tron microscopy was carried out on a Philips CM 300 transmission electron microscope (Eindhoven, The Netherlands). Synthetic pyrethroid degraders The GC chromatograms from the fatty acid analyses were Biodegradation experiments in aqueous phase compared with a large database [12] of well-known reference The mineral salt medium was spiked with 1.0 ␮g/␮lBFor cultures previously grown on trypticase soy-broth agar. Spe- PM acetone solution to give an initial concentration of 0.1 cies names of the organisms with the most similar chromato- Biodegradation of pyrethroids in sediments Environ. Toxicol. Chem. 23, 2004 3

Table 1. Culturable bifenthrin (BF)-degrading bacteria identi®ed in Table 2. Culturable permethrin (PM)-degrading bacteria identi®ed in sediments sediments

Sample Species Division Sample no. Species Division

BF-1 Xanthobacter agilis Alpha-Pro PM-1a Aeromonas sobria Gamma-Pro BF-2 Stenotrophomonas acidaminiphila Gamma-Pro PM-2a Erwinia carotovora Gamma-Pro BF-3 Burkholderia cepacia Beta-Pro PM-3 Aeromonas salmonicida Gamma-Pro BF-4 Stenotrophomonas acidaminiphila Gamma-Pro PM-4 Aeromonas salmonicida Gamma-Pro BF-5 Methylobacterium mesophilicum Alpha-Pro PM-5a Yersinia frederiksenii Gamma-Pro BF-6a Stenotrophomonas acidaminiphila Gamma-Pro PM-6 Aeromonas sobria Gamma-Pro BF-7 Stenotrophomonas acidaminiphila Gamma-Pro PM-7 Erwinia carotovora Gamma-Pro BF-8 Xanthobacter agilis Alpha-Pro PM-8 Aeromonas sobria Gamma-Pro BF-9 Stenotrophomonas acidaminiphila Gamma-Pro PM-9 Yersinia frederiksenii Gamma-Pro BF-10 No match PM-10 Aeromonas sobria Gamma-Pro BF-11 Stenotrophomonas acidaminiphila Gamma-Pro PM-11 Erwinia carotovora Gamma-Pro BF-12 No match PM-12 Aeromonas caviae Gamma-Pro BF-13 No match PM-13 Aeromonas salmonicida Gamma-Pro BF-14 Stenotrophomonas acidaminiphila Gamma-Pro PM-14 Erwinia carotovora Gamma-Pro BF-15 Stenotrophomonas acidaminiphila Gamma-Pro PM-15 Yersinia frederiksenii Gamma-Pro BF-16 Stenotrophomonas acidaminiphila Gamma-Pro PM-16 Aeromonas veronii Gamma-Pro BF-17 Agrobacterium radiobacter Alpha-Pro PM-17 Aeromonas salmonicida Gamma-Pro BF-18 No match PM-18 Aeromonas caviae Gamma-Pro BF-19 Stenotrophomonas acidaminiphila Gamma-Pro PM-19 Aeromonas salmonicida Gamma-Pro BF-20 No match PM-20 Aeromonas salmonicida Gamma-Pro BF-21 Gluconobacter asaii Alpha-Pro PM-21 Aeromonas salmonicida Gamma-Pro BF-22 Gluconobacter asaii Alpha-Pro PM-22 Yersinia frederiksenii Gamma-Pro BF-23 Xanthobacter agilis Alpha-Pro PM-23 Aeromonas caviae Gamma-Pro BF-24a Stenotrophomonas acidaminiphila Gamma-Pro PM-24 Erwinia carotovora Gamma-Pro BF-25 Burkholderia cepacia Beta-Pro PM-25 Aeromonas sobria Gamma-Pro BF-26 No match PM-26 Yersinia frederiksenii Gamma-Pro BF-27 No match PM-27 Aeromonas sobria Gamma-Pro BF-28a Stenotrophomonas acidaminiphila Gamma-Pro PM-28 Erwinia carotovora Gamma-Pro a Denotes isolates that were selected as inoculants for pure culture a Denotes isolates that were selected as inoculants for pure culture degradation experiments. degradation experiments.

Transformation in aqueous phase grams from the library are given in Table 1 for BF degraders and Table 2 for PM degraders. Among the 28 BF-degrading The three fastest growing isolates were selected as test or- isolates, the predominant bacteria were from the Proteobac- ganisms for evaluating BF degradation in solution and sedi- teria division. Twelve isolates were from the gamma Proteo- ment phases, and the isolates were labeled as S. acidaminiphila bacteria (Stenotrophomonas sp.), two from the beta Proteo- BF6, BF24, and BF28 (Table 1). The decline of BF over time bacteria, and seven from the alpha Proteobacteria. Overall, was ®tted to a ®rst-order decay model, and better model ®t six genera were represented, including, Xanthobacter, Sten- otrophomonas, Burkholderia, Methylobacterium, Agrobac- terium, and Gluconobacter, with the remaining seven isolates not assigned to any described taxa (no match). In contrast, all of the 28 isolates examined for PM degradation were identi®ed as gamma Proteobacteria. Six species were identi®ed, including Aeromonas salmonicida, Aeromonas sobria, Erwinia carotovora, Yersinia frederiksenii, Aeromonas caviae, and Aeromonas veronii. Using these enrichment and culture methods, the Aeromonas sp. isolates were predominant among the PM-degrading bacteria. In a previous study, Grant et al. [11] found that Pseudomonas species were among the most dom- inant SP-degrading organisms in garden and farm soils that were previously treated with SP products. These ®ndings suggest that SP-degrading bacteria might be widespread in soils and sediments exposed to SP products. The strains selected for pure culture incubation experiments were further con®rmed by DNA sequence analysis to belong to S. acidaminiphila (BF) and A. sobria, E. carotovora, and Y. frederiksenii (PM). The microscopic image of the S. acidaminiphila isolates showed polar ¯agella (Fig. 1). This suggests that S. acidaminiphila may move in the sediment phase and degrade SP in the aqueous phase and in water-saturated soil and sediment pore spaces that are large enough to provide Fig. 1. Transmission electron micrograph of Stenotrophomonas aci- access to the bacterium. daminiphila. Length of photo ϭ 2 ␮m. 4 Environ. Toxicol. Chem. 23, 2004 S. Lee et al.

Table 4. Kinetic parameters of ®rst-order kinetics in permethrin (PM) biodegradation of soil-free phase; see Table 2 for species identi®cation

Ϫ1 2 Bacterial isolate k (h ) t½ (h) r

cis-Permethrin A. sobria 1.24 ϫ 10Ϫ2 56 0.88 E. carotovora 1.13 ϫ 10Ϫ2 61 0.91 Y. frederiksenii 8.66 ϫ 10Ϫ3 80 0.80 Control 1.43 ϫ 10Ϫ3 485 0.83 trans-Permethrin A. sobria 1.50 ϫ 10Ϫ2 45 0.81 E. carotovora 1.51 ϫ 10Ϫ2 46 0.91 Y. frederiksenii 1.85 ϫ 10Ϫ2 37 0.95 Control 2.68 ϫ 10Ϫ3 259 0.87

trans-PM was consistent with ®ndings in other media, in- Fig. 2. Biodegradation of bifenthrin (BF) in soil-free medium after inoculation with Stenotrophomonas acidaminiphila strains. Vertical cluding soils [8]. bars are standard deviations of three replicates. In previous studies, transformation of SPs was shown to occur with bacterial isolates of soil origin, and the rate of transformation was generally substrate speci®c. Deltamethrin was found for the S. acidaminiphila BF28 and control treat- was degraded by bacterial isolates, and the t1/2 was one to two ments than for the S. acidaminiphila BF6 and BF24 treatments weeks in solid-free media [9]. Permethrin was preferentially (Table 1). The k values estimated from the regression were transformed over the other SPs by bacterial strains in the pres- compared at 95% con®dence interval. Disappearance of BF in ence of a surfactant, and the shortest t1/2 of PM was about 5 solution was signi®cantly enhanced after inoculation of S. aci- d [10]. Strains from the genera Pseudomonas and Serratia daminiphila strains when compared with the uninoculated con- were shown to degrade cypermethrin faster than ¯umethrin, trol (Fig. 2). After 100 h, Ͼ80% of the added BF was trans- and t1/2 of cypermethrin was Ͻ20 d [11]. These studies suggest formed by S. acidaminiphila BF6 and BF28, while only Ͻ15% was unaccounted for in the control treatment without inoculation. Degradation of BF by S. acidaminiphila BF6 and BF24 appeared to consist of two phases, with the initial phase dem- onstrating a more rapid degradation than the second phase. Among the three S. acidaminiphila strains, S. acidaminiphila BF24 degraded BF at a slower rate than S. acidaminiphila

BF6 and BF28. The estimated t1/2 was Ͼ700 h for BF in the uninoculated solution, but was reduced to 131 h for the S. acidaminiphila BF24 treatment and further to 30 to 36 h for the S. acidaminiphila BF6 and BF28 isolates (Table 3). The three strains used for PM biodegradation experiments were identi®ed as A. sobria, E. carotovora, and Y. frederiksenii. Relatively good ®t (r2 Ն 0.80) to the ®rst-order decay model was found for all the treatments (Table 4). The k values estimated from the regression were compared at 95% con®- dence interval. Inoculation of these strains resulted in signif- icantly more rapid degradation of cis-ortrans-PM than in the uninoculated control (Fig. 3 and Table 4). The t1/2 of cis-PM was 485 h in the uninoculated solution, which decreased to 56 to 80 h in the inoculated treatments. With trans-PM, inoculation with the bacterial strains shortened the t1/2 from 260 h in the control treatment to only 37 to 46 h in the inoculated treatments. In the same treatments, trans-PM was transformed more rapidly than cis-PM. The preferential transformation of

Table 3. Kinetic parameters of ®rst-order kinetics in bifenthrin (BF) biodegradation of soil-free phase; see Table 2 for species identi®cation

Ϫ1 2 Bacterial isolate k (h ) t½ (h) r

S. acidaminiphila BF6 2.30 ϫ 10Ϫ2 30 0.86 S. acidaminiphila BF24 5.3 ϫ 10Ϫ3 131 0.84 Fig. 3. Biodegradation of permethrin (PM) isomers in soil-free me- S. acidaminiphila BF28 1.9 ϫ 10Ϫ2 36 0.99 dium after inoculation with pesticide-degrading bacterial species. (A) Control 9.6 ϫ 10Ϫ4 719 0.98 cis-permethrin and (B) trans-permethrin. Vertical bars are standard deviations of three replicates. Biodegradation of pyrethroids in sediments Environ. Toxicol. Chem. 23, 2004 5

5). Inoculation with S. acidaminiphila resulted in signi®cantly (p ϭ 0.01) enhanced dissipation of BF when compared with the uninoculated control for both of the sediments (Fig. 4 and Table 5). In the ®eld-sediment mixture, the persistence of BF was reduced by 1.8- to 2.7-fold with bacterial inoculation when compared with the uninoculated control. In the creek-sediment mixture, BF transformation was enhanced by 9.8- to 11.7-fold. Among the three inoculants, transformation by S. acidaminiphila BF24 was slightly faster than the other two strains (Table 4). This trend differed from that in the solution phase, where S. acidaminiphila BF24 was found to degrade BF at the slowest rate. This suggests that introduction of sediment particles may have altered the activity of the inoculated strains or that S. acidaminiphila BF24 had a greater capability to transform BF that was adsorbed to the sediment phase. Transformation of BF by the bacterial isolates was consis- tently much slower in the sediment slurry system than in the sediment-free solution for the same strain (Figs. 2 and 4). Compared with the aqueous phase treatments, addition of the ®eld sediment inhibited BF transformation by S. acidaminiphila strains by 2.4- to 16.4-fold, and the reduction caused by the creek sediment was 3.8- to 19.2-fold (Tables 3 and 4). One explanation for the sediment-induced slow transformation could be that BF was removed from the aqueous phase due to adsorption to the sediment, thus limiting the availability of BF to S. acidaminiphila. The effect of adsorption on bioavailability was given previously by Zhang and Bouwer [15] in the following relationship: Fig. 4. Biodegradation of bifenthrin (BF) in the presence or absence 1 B ϭ of a sediment solid phase after inoculation with Stenotrophomonas f K ´m/v acidaminiphila strains. (A) Field sediment and (B) creek sediment. d Vertical bars are standard deviations of three replicates. where Kd is the linear adsorption coef®cient (ml/g) and m/v is the solid content in a batch slurry system (g/ml). This mea- surement was used in previous studies to evaluate the inter- that SP-degrading microbes are widely present in soils. Our action of adsorption with the bioavailability of contaminants, ®ndings show that sediments also contain numerous SP-de- with the presumption that adsorbed contaminants are not avail- grading bacterial species, which may play a major role in the able for transformation [15,16]. In a preliminary study, the K fate of SPs in contaminated aquatic systems. d of BF was measured to be 4,200 ml/g for the ®eld sediment Transformation in sediment±water mixtures and 41,000 ml/g for the creek sediment. Using these Kd values, Ϫ3 the Bf index in the slurry system was 2.4 ϫ 10 for the ®eld Transformation of BF by the S. acidaminiphila BF6, BF24, sediment and only 2.5 ϫ 10Ϫ4 for the creek sediment. Due to and BF28 was further evaluated in sediment±water mixtures. the higher content of clay and organic matter, adsorption of Concentrations of BF from both the solution and sediment BF to the creek sediment was about 10 times stronger than phases were simultaneously measured and then pooled to es- that to the ®eld sediment. The stronger adsorption apparently timate the overall BF loss in the slurry system (Fig. 4). Dis- rendered even less BF available for bacterial transformation sipation of BF over time was ®tted to the ®rst-order decay in the creek-sediment mixture. Chapman et al. [7] demonstrat- model to estimate k and t1/2 (Table 5). The ®t was better for ed that fenvalerate degraded slower in organic soils than in treatments showing more extensive BF transformation than for mineral soils, suggesting that the adsorption to organic matter treatments with limited BF transformation, e.g., the control possibly increased its persistence. treatments and the inoculated creek-sediment treatments (Table Transformation of BF in both sediment systems appeared to consist of a rapid initial phase and a slow or stagnant second Table 5. First-order parameters for bifenthrin (BF) degradation in phase. For the ®eld sediment, rapid transformation continued sediment slurry (total); see Table 1 for species identi®cation until 120 h after the treatment, whereas in the creek-sediment mixture, rapid transformation only lasted for the ®rst 48 h Ϫ1 2 Sediment k (h ) t½ r (Fig. 4). The progressive decrease in the transformation rate may be attributed to different adsorption sites or increased Field sediment S. acidaminiphila BF6 1.42 ϫ 10Ϫ3 488 0.83 S. acidaminiphila BF24 2.18 ϫ 10Ϫ3 318 0.90 sorption over time after the pesticide was added [17]. The S. acidaminiphila BF28 1.65 ϫ 10Ϫ3 420 0.83 effect of adsorption was also re¯ected in the fact that fast Control 8.0 ϫ 10Ϫ4 870 0.69 bacterial growth occurred only during the initial hours when Creek sediment S. acidaminiphila BF6 1.19 ϫ 10Ϫ3 582 0.60 rapid BF transformation occurred (Fig. 5). This suggests that S. acidaminiphila BF24 1.43 ϫ 10Ϫ3 485 0.60 the inoculated bacterial strains were able to use BF as the Ϫ3 S. acidaminiphila BF28 1.01 ϫ 10 686 0.62 carbon source for growth only at the beginning when a fraction Control 1.77 ϫ 10Ϫ4 3,915 0.66 of the added BF was in the dissolved phase and/or weakly 6 Environ. Toxicol. Chem. 23, 2004 S. Lee et al.

AcknowledgementÐWe thank D. Haver and M. Poole for assistance in sample collection.

REFERENCES 1. Hill IR. 1989. Aquatic organisms and pyrethroids. Pestic Sci 27: 429±465. 2. California Department of Pesticide Regulation. 2002. Preliminary results of pesticide analysis of monthly surface water monitoring for the red imported ®re ant project in Orange County, March 1999 through August 2002. Sacramento, CA, USA. 3. Werner I, Deanovic LA, Hinton DE, Henderson JD, de Oliveira GH, Wilson BW, Krueger W, Wallender WW, Oliver MN, Zalom FG. 2002. Toxicity of stormwater runoff after dormant spray application of diazinon and esfenvalerate (Asana (R)) in a French prune orchard, Glenn County, California, USA. Bull Environ Contam Toxicol 68:29±36. 4. Kabashima JN, Lee SJ, Haver DL, Goh K, Wu L, Gan J. 2003. Pesticide runoff and mitigation at a commercial nursery site. In Gan J, Zhu P, Lemley A, Aust S, eds, Pesticide Decontamination and Detoxi®cation. ACS Series. American Chemical Society, Washington, DC. 5. Mokrey LE, Hoagland KD. 1990. Acute toxicities of ®ve synthetic pyrethroid insecticides to Daphnia magna and Ceriodaphnia dubia. Environ Toxicol Chem 9:1045±1051. 6. Worthing CR, Hance RJ. 1991. The Pesticide Manual. British Crop Protection Council, Surrey, UK. 7. Chapman RA, Tu CM, Harris CR, Cole C. 1981. Persistence of ®ve pyrethroid insecticides in sterile and natural mineral and organic soil. Bull Environ Contam Toxicol 26:443±469. 8. Kaufman DD, Haynes SC. 1977. Permethrin degradation in soil and microbial culture. Am Chem Soc 42:147±161. 9. Khan SU, Behki RM, Tapping RI, Akhtar MH. 1988. Delta- methrin residues in an organic soil under laboratory conditions and its degradation by a bacterial strain. J Agric Food Chem 36: 636±638. 10. Maloney SE, Maule A, Smith ARW. 1988. Microbial transformation of the pyrethroid insecticides: Permethrin, deltamethrin, fastac, fenvalerate, and ¯uvalinate. Appl Environ Microbiol 54: 2874±2876. Fig. 5. Stenotrophomonas acidaminiphila growth and population den- 11. Grant RJ, Daniell TJ, Betts WB. 2002. Isolation and identi®cation sities, in colony-forming units (CFUs), during biodegradation of bi- of synthetic pyrethroid-degrading bacteria. J Appl Microbiol 92: fenthrin (BF) in sediment±water mixture. (A) Field sediment and (B) 534±540. creek sediment. 12. Mahaffee WF, Kloepper JW. 1997. Temporal changes in the bacterial communities of soil, rhizosphere, and endorhiza associated with ®eld-grown cucumber (Cucumis sativus L.). Microb Ecol adsorbed. As adsorption and transformation depleted the bio- 34:210±223. available fraction, bacterial cell densities stabilized and started 13. Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular Cloning: to decline slightly (Fig. 5). A Laboratory Manual, 2nd ed. Cold Spring Harbor, New York, NY, USA. The persistence of a chemical in sediment is closely related 14. Lane DJ. 1991. 16S/23S rRNA sequencing. In Stackebrandt E, to its availability to microorganisms. Weber and Coble [18] Goodfellow M, eds, Nucleic Acid Techniques in Bacterial Sys- have shown that microbial decomposition of diquat was pro- tematics. John Wiley, New York, NY, pp 115±175. portionally decreased as more and more clay mineral was add- 15. Zhang W, Bouwer E. 1997. Biodegradation of benzene, toluene, ed to test systems. Subba-Rao and Alexander [19] reported a and naphthalene in soil-water slurry microcosms. Biodegradation 8:167±175. similar effect on the mineralization of benzylamine; increases 16. Feng Y, Park J, Voice TC, Boyd SA. 2000. Bioavailability of in clay concentration generally decreased the rate of substrate soil-sorbed biphenyl to bacteria. Environ Sci Technol 34:1977± mineralized. Several researchers also have shown that bio- 1984. degradation can be limited by the strong adsorption or slow 17. Koskinen WC, Cox L, Yen PY. 2001. Changes in sorption/bio- desorption of organic compounds [20,21]. availability of imidacloprid metabolites in soil with incubation time. Biol Fertil Soils 33:546±550. Reduced microbial transformation of BF in sediment±water 18. Weber JB, Coble HD. 1968. Microbial decomposition of diquat mixtures suggests that, although SP degraders may be prev- adsorbed on montmorillonite and kaolinite clays. J Agric Food alent in the environment, the actual persistence of these pes- Chem 16:475±478. ticides may be prolonged because of their high af®nity to the 19. Subba-Rao RV, Alexander M. 1982. Effect of sorption on mineralization of low concentrations of aromatic compounds in lake sediment phase. The relative persistence may depend closely water samples. Appl Environ Microbiol 44:659±668. also on the properties of speci®c sediments and may be much 20. Pignatello JJ. 1989. Sorption dynamics of organic compounds in longer in sediments with higher clay and organic matter con- soils and sediments. In Sawhney BL, Brown K, eds, Reactions tents. The strong effect of adsorption also implies that reme- and Movement of Organic Chemicals in Soils. Soil Science So- diation treatments using SP degraders may be effective only ciety of America, Madison, WI. 21. Al-Bashir B, Hawari J, Samson R, Leduc R. 1994. Behavior of for contaminated water (e.g., ef¯uents) or sandy sediments. nitrogen substituted naphthalenes in ¯ooded soilÐPart II: Effect The ef®ciency may decrease for clayey sediments or sediments of bioavailability on biodegradation kinetics. Water Res 28:1827± with aged pyrethroid residues. 1833.