Pest Management Science Pest Manag Sci 60:645–654 (online: 2004)DOI: 10.1002/ps.826

Degradation and leaching potentialof pesticides in biobed systemsPaul Fogg,1∗ Alistair BA Boxall,1 Allan Walker2 and Andrew Jukes2

1Cranfield Centre for EcoChemistry, Shardlow Hall, Shardlow, Derby, DE72 2GN, UK2Horticulture Research International, Wellesbourne, Warwickshire, CV35 9EF, UK

Abstract: Biobeds provide a potential solution to pesticide contamination of surface waters arising fromthe farmyard. Previous work has shown that biobeds can effectively treat spills and splashes of pesticide.This study investigated the potential for biobeds to treat much larger volumes and amounts of pesticidewaste not only arising from spills but also from washing processes. Two systems were assessed using arange of pesticides at the semi-field scale, ie a lined biobed system and an unlined system. Studies usingthe lined biobeds demonstrated that water management was crucial, with biobeds needing to be coveredto exclude rain-water. Once covered, the top of the biobed became hydrophobic, restricting moisture lossand resulting in saturated conditions at depth. The drying out of the top layer coincided with a measureddecrease in microbial biomass in the treated biobeds. Applied pesticides were effectively retained withinthe 0–5 cm layer. Whilst all pesticides tested degraded, low moisture content and microbial activity meantdegradation rates were low. Studies using unlined biobeds showed that only the most mobile pesticidesleached, and for these >99% was removed by the system, with a significant proportion degraded within9 months. Peak concentrations of the two most mobile pesticides did however exceeded the limits that arelikely to be required by regulatory bodies. However, it is thought that these limits could be reached byoptimisation of the system. 2004 Society of Chemical Industry

Keywords: biobeds; pesticide waste; leaching potential; degradation; water management

1 INTRODUCTIONSurface waters and groundwaters have been shownto be contaminated with a range of pesticides.1,2

In order to meet the standards set by, for example,the European drinking water Directive 80/778/EEC,treatment is required before these water resources canbe used for drinking water, and such treatment canbe expensive. Research over recent years has focusedon the non-point sources of pesticide contaminationresulting from application to agricultural land.3–6

However, contamination arising from other sourcessuch as non-approved use, poor practice, illegaloperations, accidental releases from the farmyardand inputs of washings is increasingly recognisedas contributing to water contamination.7–10 Forexample, a number of workers have indicated thatpoint sources (ie the spills and washings from thefarmyard) can contribute between 18 and 84% of thepesticide load measured in individual catchments.11–16

Better training of sprayer operators and goodmachinery maintenance can reduce the number ofspills, and, by following appropriate codes of practice

and regulations, releases to the farmyard due to spraytank washings should be minimised.17–19 However,even with well-trained operators, small drips andspills are still likely to occur,10,11 and due to timeconstraints and other pressures, Codes of Practicemay not always be followed. Inputs from equipmentwashings and residues that remain in the sprayersump after infield tank rinsing are also an unavoidablefeature of the sprayer operation.7,20 For example, itis reported that between 0.5 and 25 litres of dilutespray solution remains within the sprayer sump afterin-field tank washing and disposal,10,21 and thatover the course of a normal spray season a typicalspray applicator can produce between 3800 and15 000 litres of pesticide-contaminated waste water.22

The concentration of pesticide can vary greatly,however, and concentrations of up to 450 mg litre−1

have been reported for tank washings following twointernal tank rinses.10

Additional methods for preventing pesticide wastearising from farmyard washings and spills fromreaching both surface and ground waters are therefore

∗ Correspondence to: Paul Fogg, Cranfield Centre for EcoChemistry, Shardlow Hall, Shardlow, Derby, DE72 2GN, UKE-mail: p.fogg@cranfield.ac.ukContract/grant sponsor: Department for the Environment Food and Rural AffairsContract/grant sponsor: Environment AgencyContract/grant sponsor: Crop Protection AssociationContract/grant sponsor: Monsanto Agricultural Company(Received 16 March 2003; revised version received 11 August 2003; accepted 25 September 2003)Published online 16 February 2004

2004 Society of Chemical Industry. Pest Manag Sci 1526–498X/2004/$30.00 645

P Fogg et al

required. Biobeds appear to provide a low-costalternative for treating pesticide waste and washings,providing a matrix to absorb the pesticide(s) andfacilitate biodegradation. In its simplest form a biobedis a clay lined hole in the ground filled with amixture of topsoil, peat and straw.23,24 The biobed iscovered with grass and equipped with a ramp enablingthe tractor and sprayer to be driven over the bed.Studies in Sweden have demonstrated that biobedscan effectively retain and degrade pesticide wastearising from accidental spillages of concentrate andprepared pesticides.25 However, studies performedin Denmark have shown that the clay membraneat the base of the biobed could not retain all ofthe leachate draining through the biobed.26 Studieshave also shown that whilst less mobile pesticides areeffectively retained within the biobed matrix significantamounts of the more mobile pesticides can leach fromthe biobed.26–28 A number of modifications to thebasic biobed design have been suggested in order toremove the leaching risk from biobeds, and these haveincluded the inclusion of an impermeable membraneunderneath the biobed29 and the use of activatedcarbon filters to remove any pesticide present inleachate draining from the biobed.28

Recent laboratory-based studies show that biobedsmay be able to degrade the high concentrations andcomplex mixtures of pesticides that are likely to arisein washings as well as spills,30 although the use ofbiobeds for treating larger amounts of waste (ie spillsand washings) has not yet been established.

This study was therefore performed to determinewhether biobeds could be used to treat pesticidesarising from spillages on the farmyard as well asfrom tank and sprayer washing activities. The specificobjectives were (1) to compare the performance ofboth lined and unlined biobeds; and (2) on thebasis of the results, provide recommendations on theconstruction and operation of a biobed system. Studieswere performed at the semi-field scale.

2 MATERIALS AND METHODS2.1 Preparation of biomixBiomix was prepared by mixing topsoil (69% sand,13% silt, 18% clay, 1.95% organic carbon, pH 6.15,maximum water holding capacity 37% w/w), peat-free compost (Levington Peat-Free Universal) andunchopped winter barley straw in the proportions of1 + 1 + 2 by volume.

2.2 Test chemicalsTest pesticides were selected on the basis oftheir physico-chemical properties and average annualusage,31 (Table 1). Isoproturon 500 g litre−1 SC(Alpha Isoproturon 500), pendimethalin 400 g litre−1

SC (Stomp 400 SC) and chlorpyrifos 480 g litre−1

EC (Dursban 4), were used to make up a stock sus-pension in tap water containing 11 140 mg AI litre−1,

Table 1. Study compounds and their reported physico-chemical

charactersisticsa

Active substance Koc (ml g−1)DT50 (soil)(days)

Watersolubility(mg litre−1)

Isoproturon 100 6–28 65Pendimethalin 5000 90–120 0.3Chlorpyrifos 6000 60–120 1.4Chlorothalonil 1600–14 000 6–43 0.81Epoxiconazole 957–2647 60–90 6.6Dimethoate 16–52 7–16 22 300

a Values taken from Wauchope et al46 and Tomlin.47

8000 mg AI litre−1 and 5825 mg AI litre−1 of isopro-turon, pendimethalin and chlorpyrifos, respectively.Chlorothalonil 500 g litre−1 SC (Cropgard), epox-iconazole 150 g litre−1 SC (Opus) and dimethoate400 g litre−1 EC (Rogor L40) were used tomake up a stock suspension in tap water con-taining 6533 mg AI litre−1, 756 mg AI litre−1 and2438 mg AI litre−1 of chlorothalonil, epoxiconazoleand dimethoate respectively.

2.3 Lined biobedsForty biobed lysimeter cores were prepared usingunplasticised polyvinyl chloride (PVC-u) piping(19 cm internal diameter ×75 cm length) with oneend of the cut pipe sealed using a socket. Cores werefilled with 15 cm of washed sand and a 50-cm layerof biomix (organic matter 12.36%, pH 7.5, maxi-mum water holding capacity 121% w/w), packed toa density (measured 427 days after construction) of1.27 g cm−3 in the top 0–10 cm layer and 0.53 g cm−3

for the bottom layer and placed into the ground infive groups of eight. The biobed columns were freeof any vegetation. Four of the five groups of coreswere treated with 50 ml of the pesticide mixture con-taining isoproturon, pendimethalin and chlorpyrifos inDecember 1998 and January 1999 in order to achieve afinal treatment rate of 1114 mg (isoproturon), 800 mg(pendimethalin) and 583 mg (chlorpyrifos). Treat-ment rates gave nominal concentrations in the 0–5-cm layer of 618 mg kg−1 (isoproturon), 443 mg kg−1

(pendimethalin) and 323 mg kg−1 (chlorpyrifos). Theremaining group of eight cores was left untreated andacted as a control. The four treated groups of coreswere treated with 50 ml of the pesticide mixture con-taining chlorothalonil, epoxiconazole and dimethoatein April 1999 and June 1999 in order to achieve afinal treatment rate of 653 mg (chlorothalonil), 76 mg(epoxiconazole) and 244 mg (dimethoate). Treatmentrates gave nominal concentrations in the 0–5cm layerof 361 mg kg−1 (chlorothalonil), 42 mg kg−1 (epoxi-conazole) and 135 mg kg−1 (dimethoate). Applicationrates were based on theoretical worst-case disposalrates (ie two applications of 100 litres of full strengthdilute pesticide). A roof was constructed over thecores following the first treatments with isoproturon,pendimethalin and chlorpyrifos to exclude rainfall.

646 Pest Manag Sci 60:645–654 (online: 2004)

Degradation and leaching potential of pesticides in biobeds

To simulate runoff from an impermeable pesticidehandling area connected to a biobed, artificial irriga-tion was applied in February, May, July, August andSeptember at the rate of 314 ml per core, equivalentto 11.1 mm of rainfall.

Two untreated cores were collected prior to thefirst pesticide treatment and sectioned into threeapproximately equal parts. Sub-samples were obtainedfrom each section for microbial biomass determina-tion. Following treatment, cores were collected oneight occasions over a 12-month period (Table 2). Oneach sampling occasion, three treated cores and oneuntreated control were collected, the cores were thensectioned (0–5 cm, 5–10 cm, 10–20 cm, 20–30 cmand 30–50 cm). Sections down to 20 cm depth werehomogenised in a food processor and stored at −15 ◦Cprior to chemical analysis. With the exception of sam-ples taken at T = 0 and T = 3, sub-samples werecollected (0–10 cm, 10–30 cm and 30–50 cm) forbiomass and moisture content determinations.

2.4 Unlined biobedsTwo sets of four lysimeters were prepared using PVC-u piping (19 cm internal diameter ×75 cm length) withone end of the cut pipe sealed using a socket fitted witha drain outlet. Cores were filled with 2–3 cm of gravelfollowed by 15 cm of washed sand. A 50-cm layer ofeither biomix (organic matter 12.36%, pH 7.5, max-imum water holding capacity 121% w/w) or topsoil(69% sand, 13% silt, 18% clay, 1.95% organic car-bon, pH 6.15, maximum water holding capacity 37%w/w) was then packed into each lysimeter. A density(measured 316 days after construction) of 1.67 g cm−3

in the 0–5cm layer down to 0.21 g cm−3 at the base wasachieved for the biomix compared with 1.68 g cm−3 to0.59 g cm−3 for topsoil. The base of each core drainedvia the drain outlet through Teflon tubing to a 2.5-litre amber glass bottle located in a central collectionpit.32 Three of the biomix-filled lysimeters and threesoil-filled lysimeters received split applications of iso-proturon, pendimethalin, chlorpyrifos, chlorothalonil,epoxiconazole and dimethoate. Treatment rates andtimings were the same as for the lined biobed cores.A potassium bromide (KBr) tracer was also applied

Table 2. Sampling time points for lined biobedsa

Time point

Days afterapplication

1

Days afterapplication

2

Days afterapplication

3

Days afterapplication

4

T = 0 1T = 1 36T = 2 105 68T = 3 123 86 1T = 4 165 128 43T = 5 260 223 138 89T = 6 322 285 200 151T = 7 365 328 243 194

a Applications 1 and 2: isoproturon, pendimethalin and chlorpyrifos;applications 3 and 4: chlorothalonil, epoxiconazole and dimethoate.

(628 mg core−1) to check the hydrological integrity ofthe lysimeters, as well as to determine breakthroughtiming of the infiltrating water. Collection vessels weremonitored after all rainfall events and the total volumeof leachate recorded. Volumes in excess of 500 mlwere collected and stored at 0–10 ◦C prior to analysis.Where possible, a 60-ml sub-sample was also taken forKBr analysis. At the end of the study, (254 days afterapplication 1) all cores were sectioned (0–5, 5–10,10–20, 20–30 and >30 cm), homogenised and storedat −15 ◦C prior to analysis.

2.5 Analysis2.5.1 Water analysesWater samples were either analysed directly using highperformance liquid chromatography (HPLC) or anal-ysed by HPLC or gas chromatography (GC) afterliquid/liquid extraction. For leachate collected prior tothe application of chlorothalonil, epoxiconazole anddimethoate, samples (500 ml) were extracted withHPLC grade dichloromethane (2 × 30 ml) in a 1-litre glass separating funnel. The dichloromethaneextracts were combined and evaporated to drynessusing a rotary evaporator at 40 ◦C. The resultingresidue was then re-dissolved in acetonitrile + water(60 + 40 by volume; 2 ml). Concentrations of isopro-turon and pendimethalin were determined by HPLCusing a Kontron Series 320 pump linked to a Kon-tron Series 332 UV detector. Samples (20 µl) wereinjected using a Kontron Series 360 autosampler. Sep-aration was achieved using a Lichrosorb RP18 column(250 mm × 4 mm ID) and a flow rate of 1 ml min−1.For isoproturon determinations, the mobile phasewas acetonitrile + water at 75 + 25 by volume, andfor pendimethalin determinations, at 90 + 10 by vol-ume. The detection wavelength for both compoundswas 250 nm. Quantification was achieved by com-paring peak areas with results from known standards.For chlorpyrifos determinations, sub-samples (1 ml) ofthe acetonitrile/water extracts were mixed with water(25 ml) and extracted into hexane (1 ml). Concentra-tions of chlorpyrifos were then determined by GC witha nitrogen/phosphorus detector (GC method 1). Sep-aration was achieved using 3% OV1 on ChromosorbWHP column (1 m × 3 mm IS), with nitrogen flow of50 ml min−1, hydrogen flow 2 ml min−1 and air flow450 ml min−1. The column temperature was 220 ◦C,the injector temperature 225 ◦C and the detector tem-perature 230 ◦C. Quantification was again achievedby comparison of peak areas with results from knownstandards. Recoveries checks for all three compoundswere >93%.

For leachate collected following application ofall six pesticides, samples (200 ml) were extractedwith dichloromethane (3 × 30 ml) in a 500-ml glassseparating funnel. The dichloromethane extracts werepassed through anhydrous sodium sulfate and thenevaporated to dryness at 40 ◦C. The resulting residueswere re-dissolved in methanol + dichloromethane(10 + 90 by volume; 2 ml). Concentrations of each

Pest Manag Sci 60:645–654 (online: 2004) 647

P Fogg et al

pesticide were determined on a Hewlett–PackardHP5890 gas chromatograph fitted with a split/splitlessinjector, 12 m × 0.53 mm BPX5 column (SGE) anda nitrogen/phosphorus detector (GC method 2).The carrier gas (helium) flow rate was 7 ml min−1

and detector-gas flow rates were 100 ml min−1 (air)and 4 ml min−1 (hydrogen). Oven temperature wasraised from 90 ◦C to 190 ◦C (40 ◦C min−1) andthen to 220 ◦C (10 ◦C min−1) and finally to 245 ◦C(15 ◦C min−1). Samples (2 µl) were injected usinga Hewlett–Packard HP7673 autosampler. Underthese conditions all six pesticides were baselineseparated with retention times of 3.1 (dimethoate), 3.5(chlorothalonil), 3.9 (isoproturon), 4.2 (chlorpyrifos),4.7 (pendimethalin) and 7.2 minutes (epoxiconazole).Quantification was achieved by comparison of peakareas with results from external standards. Recoverieswith dichloromethane extraction of water spiked at0.01 mg litre−1 were >94% for all compounds.

Concentrations of potassium bromide were deter-mined using ion chromatography. Water samples(0.5 ml) were filtered (0.2 µm) and analysed usinga Dionex DX-100. Samples (25 µl) were injected neatwith a typical retention time of 2.3 min. The systemwas calibrated using a series of standards with knownconcentrations of bromide with a limit of detection setat 1.1 mg litre−1.

2.5.2 Soil analysisFollowing solvent extraction, soil and biomix sampleswere analysed either by HPLC or GC.

Samples of T = 0 to T = 3 (Table 2) solidmaterial (25 g) were mixed with methanol (50 ml),shaken for 50 min using a wrist-action shaker, andthen allowed to stand until the solid materialhad settled. Aliquots (2 ml) of the clear methanolwere transferred directly to glass HPLC vials fordetermination of isoproturon and pendimethalin usingthe HPLC method described above. Sub-samples(either 1 ml or 5 ml) of the methanol extract weretaken for chlorpyrifos determination. For chlorpyrifosdetermination, the methanol extracts were mixedwith water (50 ml) and the methanol/water mixtureextracted into hexane (5 ml). The hexane extract wasdried using anhydrous sodium sulfate (5 g) prior toGC analysis using the GC method 1 (Section 2.5.1).

For all other soil and biomix, samples (40 g) wereplaced into 250-ml glass bottles. Anhydrous sodiumsulfate (40 g) and dichloromethane + 10% methanol(90 + 10 by volume; 160 ml) were added, with sam-ples shaken for 1 h using an end-over-end shaker.Samples were allowed to stand until clear, with analiquot of the solution taken for determination of iso-proturon, pendimethalin, chlorpyrifos, chlorothalonil,epoxiconazole and dimethoate using the GC method2, Section 2.5.1. With the exception of chlorothalonil(82%) the recovery of all six pesticides exceeded 95%.

2.5.3 Microbial biomassTotal microbial biomass was determined by fumi-gation extraction.33 Chloroform (2 ml) was added

to triplicate samples (20 g) of soil and biomix, anda control sample was left untreated. Treated anduntreated samples were sealed and incubated at 30 ◦Cfor 7–10 days. Following incubation, fumigated sam-ples were evacuated four to six times in a vacuumdesiccator to remove the chloroform and then shakenfor 50 min with potassium chloride (2 M; 50 ml). Sam-ples were then centrifuged, an extract (1 ml) takento which ninhydrin (0.5 ml) was added. The sam-ples were then immersed in a boiling water bathfor 20 min. After cooling, samples were made upto 10 ml using ethanol + water (50 + 50 by volume),transferred to plastic cuvettes and the absorbancemeasured using a spectrophotometer at 570 nm. Theabsorbances were corrected for the unfumigated con-trols and the amounts of ninhydrin-reactive nitrogenderived from a calibration curve produced using dif-ferent concentrations of L-leucine. The results werecorrected for moisture content and the total biomasscarbon (mg kg−1) calculated.33

3 RESULTS3.1 Lined biobeds3.1.1 Moisture status and microbial biomassPrior to being covered, lined biobeds intercepted156 mm of rainfall, equivalent to 4.42 litres, with anadditional 1.5 litres applied in the form of artificialirrigation over the course of the experiment. Themeasured maximum water holding capacity for thebiomix was 127% w/w, equivalent to approximately8.2 litres of water per core. Moisture content inthe 0–10 cm layer remained relatively static (average52%) throughout the study period. Below 10 cmdepth a gradual increase was measured with saturatedconditions observed by the end of the study(Fig 1). Total microbial biomass in the untreatedcores (0–50 cm) ranged from 264 to 5310 mg kg−1

carbon and in the treated (0–50 cm) cores 141 to3164 mg kg−1 carbon. Despite considerable variationin measurements in the 0–10-cm layer, biomass inthe treated cores declined over the study (ANOVAP < 0.05, F 13.28, df 1), whereas in the untreatedcores the measured biomass remained relatively static,

0

100

200

300

400

500

600

700

800

Pre-treatment

T=1 T=2 T=4 T=5 T=6 T=7

Sampling Time Points

Mo

istu

re C

on

ten

t %

(m

ass) 0-10 cm

10-30 cm30-50 cmMWHC

Figure 1. Measured water contents (±1 SE) in lined biobeds.MWHC: maximum water holding capacity.

648 Pest Manag Sci 60:645–654 (online: 2004)

Degradation and leaching potential of pesticides in biobeds

(Fig 2). In the 10–30-cm and 30–50-cm layers therewere no significant differences between the treated anduntreated cores.

3.1.2 Pesticide residuesThe highest concentrations of all pesticides weremeasured in the 0–5-cm layer of the biobed, (Fig 3).Concentrations in the deeper layers were significantlylower, indicating little downward movement ofthe study compounds, (Fig 4). Concentrations ofisoproturon, pendimethalin and chlorpyrifos in the0–5-cm layer remained static for the first 100 days.During the next 100 days rapid degradation was

0200400600800

100012001400160018002000

Pre-treatment

T=1 T=2 T=4 T=5 T=6 T=7

Untreated 0-10 cmTreated Mean 0-10 cm

Sampling time point

Bio

mas

s C

(m

g k

g-1

)

Figure 2. Mean microbial biomass (±1 SE) in the 0–10 cm layer ofthe lined biobed columns.

observed, after which residues persisted at a lowlevels until the end of the experimental period,(Fig 3a,b,c). For chlorothalonil and epoxiconazole therate of degradation was slow, such that the amount ofeach pesticide recovered at the end of the study wassimilar to that measured at the beginning, (Fig 3d,e).However, for dimethoate the pattern of degradationwas relatively fast and <10% of the applied dose wasrecovered at the end of the study, (Fig 3f).

3.2 Unlined biobedsUnlined topsoil and biomix lysimeters received 116%of the long-term average rainfall, equivalent to16.9 litres of water per lysimeter, with 13 samplesof leachate being collected over a nine-month period.With one exception, cumulative leachate volumes weresimilar, with approximately 10 litres collected fromboth topsoil and biomix lysimeters. The maximumwater holding capacity of the biomix lysimeterswas approximately twice that of topsoil (8.1 litrescompared with 4.1 litres). Rapid breakthrough ofbromide was observed from topsoil lysimeters,with highest concentrations observed 35 days aftertreatment (DAT). Movement of bromide throughbiomix-filled cores was much slower, with maximumconcentrations not being observed until 102 DAT.Generally, data suggested chromatographic water

0

20

40

60

80

100

0 100 200 300 400

Days after treatment

% o

f ap

plie

d d

ose

(a)

0

20

40

60

80

100

0 100 200 300 400

Days after treatment

% o

f ap

plie

d d

ose

(b)

0

20

40

60

80

100

0 100 200 300 400

Days after treatment

% o

f ap

plie

d d

ose

(c)

0

20

40

60

80

120

100

0 100 200 300

Days after treatment

% o

f ap

plie

d d

ose

(d)

0

20

40

60

80

0 100 200 300

Days after treatment

% o

f ap

plie

d d

ose

(e)

0

10

20

30

50

40

0 100 200 300

Days after treatment

% o

f ap

plie

d d

ose

(f)

Figure 3. Amounts (±1 SE) of (a) isoproturon, (b) pendimethalin, (c) chlorpyrifos, (d) chlorothalonil, (e) epoxiconazole and (d) dimethoate remainingin the 0–5-cm layer of the lined biobeds expressed as percentage of the applied dose. For isoproturon, pendimethalin and chlorpyrifos the secondapplication was made 37 days after treatment 1. For chlorothalonil, epoxiconazole and dimethoate the second application was made 49 days aftertreatment 1.

Pest Manag Sci 60:645–654 (online: 2004) 649

P Fogg et al

0

10

20

30

40

50

0 100 200 300 400

Days after treatment

% o

f ap

plie

d d

ose

(a)

0.0

1.0

2.0

3.0

4.0

0 100 200 300 400

Days after treatment

% o

f ap

plie

d d

ose

(c)

0.00.20.40.6

1.00.8

0 100 200 300

Days after treatment

% o

f ap

plie

d d

ose

(d)

0.0

0.2

0.4

0.6

1.0

0.8

0 100 200 300

Days after treatment

% o

f ap

plie

d d

ose

(f)

0.0

0.5

1.0

1.5

2.0

0 100 200 300

Days after treatment

% o

f ap

plie

d d

ose

(e)

0

5

10

15

20

25

0 100 200 300 400Days after treatment

% o

f ap

plie

d d

ose

(b)

Figure 4. Amounts (±1 SE) of (a) isoproturon, (b) pendimethalin, (c) chlorpyrifos, (d) chlorothalonil, (e) epoxiconazole and (f) dimethoate remainingin the 5–10-cm layer of the lined biobeds expressed as percentage of the applied dose. For isoproturon, pendimethalin and chlorpyrifos the secondapplication was made 37 days after treatment 1. For chlorothalonil, epoxiconazole and dimethoate the second application was made 49 days aftertreatment 1.

movement in both the topsoil and biomix filledlysimeters, indicating hydraulic integrity of the testsystem.

With the exception of pendimethalin, concen-trations of pesticide in leachate from biomix-filledlysimeters were significantly lower than in leachatefrom topsoil (Fig 5). Peak concentrations of activeingredient in leachate from biomix ranged from0.15 µg litre−1 (epoxiconazole) to 127 µg litre−1 (iso-proturon), whereas from topsoil cores concentra-tions ranged from 0.47 µg litre−1 (pendimethalin) to3845 µg litre−1 (isoproturon). By the end of thestudy (ie 254 days after the first application), con-centrations of isoproturon, pendimethalin, chlorpyri-fos, chlorothalonil, epoxiconazole and dimethoate inleachate from the biomix columns had dropped to lessthan 1.0 µg litre−1 (Fig 5).

With the exception of dimethoate in soil, nopesticide was detected in the soil or biomix matrixbelow 30 cm depth, with the majority being retained inthe top 10 cm (Fig 6). By the end of the study, between7% (isoproturon) and 30% (epoxiconazole) remainedin the biomix, whereas between 0.7% (isoproturon)and 38% (pendimethalin) remained in the topsoilcores (Table 3). This indicates that, in biomix, only asmall proportion of the applied dose (<1%) is leachedand between 70 and 93% is degraded.

4 DISCUSSIONBiobeds have been in use in Sweden since 1993,with more than 1000 in practical use on farms andother places where pesticide sprayers are filled.25 Thebasic design of a 0.6-m deep hole lined with clay andfilled with biomix with an access ramp has remainedlargely unchanged,24 with reliable performance beingmeasured for up to years.25 Whilst the Swedish systemhas been shown to be able to treat pesticide spills,the use of biobeds for treating the large volumesof waste and high amounts of pesticide associatedwith washings as well as spillages has not yet beenestablished. If a system could be developed to deal withthese types of input, then it is possible that incidencesof contamination of surface waters by pesticides couldbe greatly reduced.

In this study, two systems were investigated: a linedsystem where the biomix was enclosed in a sealedcolumn and an unlined system where leachate wasable to percolate from the bottom of the biomix. Theuse of a lined system was considered attractive as itminimises the potential for leachate to contaminategroundwaters and is hence likely to be more attractiveto regulatory authorities.

The lined biobed columns had to be covered fol-lowing the first herbicide application to exclude cleanrain-water from being intercepted by the biobed itself.

650 Pest Manag Sci 60:645–654 (online: 2004)

Degradation and leaching potential of pesticides in biobeds

0

1000

2000

3000

4000

5000

6000

0(a)

100 200 300Days after treatment 1

Co

nce

ntr

atio

n (µ

g li

tre-1

)

(b)

0

2

4

6

8

10

0 100 200 300Days after treatment 1

Co

nce

ntr

atio

n (µ

g li

tre-1

)

(c)

0

2

4

6

8

10

0 100 200 300Days after treatment 1

Co

nce

ntr

atio

n (µ

g li

tre-1

)

(d)

0

50

100

150

200

300

250

0 50 100 150 200Days after treatment 1

Co

nce

ntr

atio

n (µ

g li

tre-1

)

(e)

05

101520

353025

0 50 100 150 200Days after treatment 1

Co

nce

ntr

atio

n (µ

g li

tre-1

)

0

50

1000

1500

2000

2500

0(f)

50 100 150 200Days after treatment 1

Co

nce

ntr

atio

n (µ

g li

tre-1

)

Figure 5. Concentrations (±1 SE) of (a) isoproturon, (b) pendimethalin, (c) chlorpyrifos, (d) chlorothalonil, (e) epoxiconazole and (f) dimethoatemeasured in leachate from (�) soil- and (�) biomix-filled lysimeters.

However, irrigation was applied to each column to sim-ulate runoff from an area of hard standing. A surveyof local farms carried out prior to the study con-cluded that the preferred location of a biobed wouldbe adjacent to the existing pesticide-mixing area. Onthe farms surveyed, the mixing area was generally con-structed from concrete and as such would generaterun-off in response to both rainfall and cleaning oper-ations. Once covered, the top 10-cm layer dried outto form a cap. Hydrological connectivity was inter-rupted and severely restricted evaporation from thesystem. Minimal water loss resulted in saturated con-ditions below 10 cm within 12 months, agreeing withobservations reported for covered Swedish biobeds.25

Microbial biomass was used to assess levels of bio-logical activity within the biobed. Over a 12-monthperiod, biomass decreased in the 0–10-cm layer. Thiswas probably a function of low moisture content, butthere may also have been inhibition by the high levelsof retained pesticide.27,34 Adequate water is essentialfor microbial activity and thus biodegradation. Exper-iments have generally shown an increase in the rateof pesticide loss with increase in soil moisture statusup to 5 kPa (field capacity).35,36 At moisture con-tents below 75%, microbial activity in biobeds can belimited.25 Although pesticides were effectively retainedin the lined system, residues levels of ≤52% werestill recovered after 12 months. Generally, persistence

increases with increasing concentration,25,37–40 and,at high concentrations, pesticides have been shownto depress microbial biomass and bioactivity; con-sequently, degradation may have been inhibited.39

In many agricultural situations the use of tankmixes and complex spray programmes is commonpractice.34,41,42 There is evidence that the sorp-tion and persistence of a number of pesticides maybe changed when used in combination with otherpesticides.30,34,42–44 On the basis of the results, ittherefore appears that lined biobeds would be unlikelyto cope with the large volumes of waste associatedwith tank and sprayer washings, as they would becomewaterlogged and microbial activity would be reduced.Some form of water management might resolve theseproblems, but this would probably result in increasedcosts and time inputs for the user.

The use of unlined biobeds removed the needto manage water inputs whilst at the same timemaintaining near-optimum conditions for pesticidedegradation, as rain-water is able to enter andsubsequently drain from the system. The studiesdemonstrated that the concentrations of pesticideleaching from the biomix-filled lysimeters weresignificantly lower than from soil lysimeters. Only themost mobile compounds leached to any great extent,and even for these compounds the system appearedto retain or degrade more than 99% of the applied

Pest Manag Sci 60:645–654 (online: 2004) 651

P Fogg et al

0 2 4 6 8

0-5 cm

5-10 cm

10-20 cm

20-30 cm

>30 cm

Dep

th

Recovery %(a)

Soil

Biomix

0 2 4 6 8

0-5 cm

5-10 cm

10-20 cm

20-30 cm

>30 cm

Dep

th

Recovery %(f)

0 5 10 15

0-5 cm

5-10 cm

10-20 cm

20-30 cm

>30 cm

Dep

th

Recovery %(c)

0 10 20 30 40

0-5 cm

5-10 cm

10-20 cm

20-30 cm

>30 cm

Dep

th

Recovery %(e)

0 10 20 30 40 50

0-5 cm

5-10 cm

10-20 cm

20-30 cm

>30 cm

Dep

thRecovery %(d)

0 10 20 30 40

0-5 cm

5-10 cm

10-20 cm

20-30 cm

>30 cm

Dep

th

Recovery %(b)

Figure 6. Amounts (±1 SE) of (a) isoproturon, (b) pendimethalin, (c) chlorpyrifos, (d) chlorothalonil, (e) epoxiconazole and (f) dimethoate expressedas percentage of the applied dose remaining in unlined biobeds 254 days after the first treatment.

Table 3. Mass balance for un-lined topsoil and biomix lysimetersa

Topsoil Biomix

PesticideLeached

(%)Retained

(%)Degraded

(%)Leached

(%)Retained

(%)Degraded

(%)

Isoproturon 1.5 0.7 97.8 0.1 7.0 92.9Pendimethalin 0 37.4 62.6 0 28.8 71.2Chlorpyrifos 0 14.6 85.4 0 13.1 86.9Chlorothalonil 0.2 34.2 65.5 0 25.6 71.4Epoxiconazole 0.3 24.7 75.0 0 30.0 70.0Dimethoate 8.4 4.0 87.5 0 7.3 92.7

a Mass balance calculated 217 days after last application of isoproturon, pendimethalin and chlorpyrifos and 83 days after application of chlorothalonil,epoxiconazole and dimethoate.

dose. Whilst >99% removal was achieved for thesix compounds tested, maximum concentrations ofthe two most mobile compounds, isoproturon anddimethoate, were 127 and 50.4 µg litre−1 respectively.Studies in Denmark using 2-m3 lysimeters lookedat the leaching potential over a 2-year period ofisoproturon and mecoprop in both biomix- and clay-filled lysimeters after receiving two simulated pesticidespills each of 8 g.27 The results showed that total

amounts of isoproturon leached were 1947 mg fromthe soil compared with 32 mg from the biobed; formecoprop 574 and 175 mg leached from the soil andbiobed respectively. Such values may be unacceptableto regulatory authorities. For example, even thoughthe Danish study demonstrated that the biobedsystem was able to retain a significant amount ofthe applied dose, pesticide concentrations in leachatewere unacceptable to the Danish EPA. In addition,

652 Pest Manag Sci 60:645–654 (online: 2004)

Degradation and leaching potential of pesticides in biobeds

the biobed matrix was classified as hazardous waste.In the UK, the Environment Agency has proposedregulating biobed performance against the GroundWater Regulations (1998). These regulations stipulatethat concentrations of pesticide reaching groundwatermust be <0.1 µg litre−1.

Henriksen et al27 proposed that one method ofreducing the concentrations of pesticide leaching outof the biobed would be to cover them during the winterperiod, thus excluding excess rain-water. In additionthey suggested that a closed biobed would removethe issue of pesticides leaching from the system—thework reported here suggests that this may not bea practical solution. The installation of secondarytreatment options (eg activated carbon) at the outletof a biobed, has also been investigated and shown tosignificantly reduce leachate concentrations.28

In order to prevent the biobed matrix beingclassified as hazardous waste, it is essential thatthe pesticide retained by the biobed matrix isdegraded. In Denmark27 and France (Higginbothampers comm) studies have shown that biobeds collect,retain and degrade pesticides. However, the regulatoryauthorities in both countries have classified the matrixas hazardous waste, therefore requiring specialisttreatment for its disposal. Analysis of the biobedmatrix from this study showed that most pesticidewas retained in the top 5 cm of the biobed, anobservation supported by Toller and Flaim,45 andthat after nine months a significant proportion of thenon-leached pesticide had been degraded. With <30%of the most persistent compound (epoxiconazole)remaining after nine months (compared with ≤52%in the lined systems), accumulation from one growingseason should not occur. Laboratory investigationshave compared pesticide behaviour in sterile and non-sterile biomix and concluded that degradation is theprinciple mechanism responsible for the reduction inmeasured concentrations of pesticide.30

5 CONCLUSIONSStudies with lined biobeds demonstrated that pes-ticides with a range of physico-chemical propertieswere effectively retained. However, monitoring of soilmoisture status indicated that lined biobeds neededto be covered in order to exclude rain-water from thesystem. Once covered, the surface layer (0–10 cm)rapidly dried to form a hydrophobic layer, severelyrestricting evaporation and thus moisture loss. Thisresulted in saturated conditions below 10 cm depthwithin 12 months of construction. The drying out ofthe surface layer was also associated with a decreasein microbial biomass in the treated biobed columns.In the untreated biobeds microbial biomass remainedrelatively constant, indicating that the retained pes-ticide residues may have an inhibitory effect on thebiomix microbial community. Whilst all pesticidestested were degraded, the rate of degradation for some

compounds was slow, a function of low moisture con-tent and microbial activity. Studies with lined biobedshave highlighted that water management is crucialand that accumulation of some pesticides may be pos-sible. Unlined biobed columns were uncovered andleachate was allowed to flow out of the bottom ofthe column, thus removing any need to manage waterinputs. Of the six pesticides tested only the two mostmobile (Koc < 100) pesticides leached, and, for these,>99% was retained; a significant proportion of theretained chemical was degraded within nine months.Under the controlled conditions of these experiments,unlined biobeds appear capable of treating the pesti-cide waste and washings that originate from spray fillsites. In order for biobeds to be approved for use it islikely that the performance of the system will have toimprove so that maximum concentrations of pesticidein leachate are close to the 0.1 µg litre−1 limit. Con-centrations of pesticide in leachate will be controlledby a number of factors including (1) the hydraulicload, (2) the depth of the biobed and (3) the lengthof time between application and significant rainfall.Experiments are therefore currently being made toinvestigate the effects of each of these parameters onbiobed performance.

ACKNOWLEDGEMENTSThe authors acknowledge financial support from thefollowing: Department for the Environment Food andRural Affairs, Environment Agency, Crop ProtectionAssociation, Monsanto Agricultural Company.

Opinions expressed within this paper are those of theauthors and do not necessarily reflect the opinion ofthe sponsoring organisations. No comments should betaken as an endorsement or criticism of any compoundor product.

REFERENCES1 Hankinson G and Welland G, The effectiveness of a stewardship

campaign in Severn–Trent water, in Pesticide behaviour in soiland water, BCPC Symposium No 78, BCPC, Farnham, Surrey,UK, pp 325–332 (2001).

2 Hillier DC and White SL, Pesticide trends in raw and treateddrinking water, in Pesticide behaviour in soil and water, BCPCSymposium No 78, BCPC, Farnham, Surrey, UK, pp 307–312(2001).

3 Gatzweiler E, Schmidt B, Feyerabend M, Jones RL andArnold DJ, Isoproturon—field studies on the behaviourin fine textured soils, Proc XI Symp Pestic Chem, 11–15September, Cremona, Italy, pp 305–312 (1999).

4 Reese-Stahler G and Pestmer W, Measurement of selectedpesticides and their impact on surface water in a researchcatchment, Proc XI Symp Pestic Chem, 11–15 September,Cremona, Italy, pp 433–440 (1999).

5 Brown C, Hodgkinson R, Rose D, Syers JK and Wilkinson, S,Movement of pesticides to surface waters from heavy clay soil.Pestic Sci 43:131–140 (1995).

6 Fogg P, Carter AD and Brown CD, A comparison of the useand regulatory interpretation of lysimeters and field studiesto determine the leaching potential of a seed dressing, in ProcBrighton Crop Prot Conf—Pest Dis, BCPC, Farnham, Surrey,UK, pp 1283–1287 (1984).

Pest Manag Sci 60:645–654 (online: 2004) 653

P Fogg et al

7 Yoder DC, Corwin BK, Mueller TC, Hart WE, Mote CR andWills JB, Development of a system to manage pesticide-contaminated wastewater. Trans ASAE 44:877–890 (2001).

8 Kruger J and Nilsson E, Catchment scale risk-mitigationexperiences—key issues for reducing pesticide transport tosurface waters, in Pesticide behaviour in soil and water, BCPCSymposium No 78, BCPC, Farnham, Surrey, UK, pp 319–324(2001).

9 Carter AD, Pesticide contamination of water sources and themonitoring data across the EU, Proc XI Symp Pestic Chem,11–15 September, Cremona, Italy, pp 11–20 (1999).

10 Mason PJ, Foster IDL, Carter AD, Walker A, Higginbotham S,Jones RL and Hardy IAJ, Relative importance of pointsource contamination of surface waters: River Cherwellcatchment monitoring study, Proc XI Symp Pestic Chem,11–15 September, Cremona, Italy, pp 405–412 (1999).

11 Higginbotham S, Jones RL and Gatzweiler E, Point-source pes-ticide contamination: quantification and practical solutions,in Proc Brighton Crop Prot Conf—Weeds, BCPC, Farnham,Surrey, UK, pp 681–686 (1999).

12 Carter A, How do pesticides get to water—The Cherwellproject. Keeping pesticides out of water, Proc CIWEM andPesticide Action Network UK, 7 February, London (2001).

13 Neumann M, Chultz R, Schafer, Muller W, Wilfied M andLiess M, The significance of entry routes and non-pointsources of pesticides in small streams. Water Res 36:835–842(2002).

14 Muller K, Bach M, Hartmann H, Spiteller M and Frede H-G, Point- and non-point-source pesticide contamination inthe Zwester Ohm catchment, Germany. J Environ Qual31:309–318 (2002).

15 Fischer Von P, Hartmann H, Bach M, Burhenne J, Frede H-G and Spiteller M, Pesticide pollution in three watersheds.Gesunde Pflanzen 50(5):142–146 (1998).

16 Kruger J, Pesticides in stream water within an agriculturalcatchment in southern Sweden. Sci Total Envir 216:227–251(1998).

17 Rose SC, Mason PJ, Foster IDL, Walker A and Carter A, Thedesign of a pesticide handling and washdown facility, inPesticide behaviour in soil and water, BCPC Symposium No 78,BCPC, Farnham, Surrey, UK, pp 379–384 (2001).

18 Code of practice for the safe use of pesticides on farms andholdings, Ministry of Agriculture Fisheries and Food, UK,PB3528 (1998).

19 The Ground Water Regulations, Statutory Instrument No 2746,HMSO, UK, ISBN 0 11 079799 X (1998).

20 Wise C, Reducing pesticide contamination of water: a farmingview. Pesticide News 26:14–16 (1994).

21 Balsari P and Airoldi G, A survey to determine the amountor unused product and disposal methods used in pesticideapplication, in Managing pesticide waste and packaging, BCPCSymposium No 70, BCPC, Farnham, Surrey, UK, pp 195–202(1998).

22 Heacox L, Ground water contamination not widespread. FarmChem 153:30–35 (1991).

23 Torstensson L and Castillo MdP, Biobeds minimise environ-mental risks when filling agricultural spraying equipment, Procof COST 66 Workshop, 13–15 May, Stratford-upon-Avon,UK, pp 223–224 (1996).

24 Torstensson L and Castillo MdP, Use of biobeds in Swedento minimise environmental spillages from agricultural sprayequipment. Pestic Outlook 8:24–27 (1997).

25 Torstensson L, Experiences of biobeds in practical use inSweden. Pestic Outlook 11:206–211 (2000).

26 Spliid NH and Helweg A, Fate of pesticides in a fullscale biobed, Internat Symp Non-agricultural Use of Pesti-cides—Environmental Issues and Alternatives, 07–09 May, TheRoyal Veterinary and Agricultural University, Copenhagen,Denmark, pp 57–58 (2003).

27 Henriksen VV, Binder A, Nielsen M, Laurensen B, Spliid NH,Helweg A, Felding G and Hansen LS, Leaching of pesticidesfrom washing-sites and capacity of biobeds to retain pesticides,

Proc 16 Danish Plant Prot Conf, March, DJF-rapport nr 9, pp47–63 (1999).

28 Pussemier L, Goux S, Elsen YV and Mariage, Q, Biofilters foron-farm clean-up of pesticide wastes. Med Fac LandbouwwUniv Gent 63:11–27 (1998).

29 Linde KM, Helweg A and Spliid NH, Release potential ofpesticides from heavily polluted biobed soil, Internat SympNon-agricultural Use of Pesticides—Environmental Issues andAlternatives, 07–09 May, The Royal Veterinary and Agricul-tural University, Copenhagen, Denmark, p 127 (2003).

30 Fogg P, Boxall ABA, Walker A and Jukes A, Pesticide degra-dation in a ‘biobed’ composting substrate. Pest Manag Sci59:527–537 (2003).

31 Garthwaite DG and Thomas MR, Pesticide usage survey report159: Arable farm crop in Great Britain 1998, Pesticide UsageSurvey Group, Central Science Laboratories, Sand Hutton,York, YO41 1LZ, pp 27–28.

32 Parsons RG and Jones RL. A new approach to the design oflysimeter facility, in Proc Brighton Crop Prot Conf—Weeds,BCPC, Farnham, Surrey, UK, pp 855–860 (1993).

33 Mele PM and Carter MR, Estimation of microbial biomass byninhydrin-reactive nitrogen using liquid chloroform. Canad JSoil Sci 76:37–40 (1996).

34 Felsot AS and Dzantor EK, Effect of alachlor concentration andan organic amendment on soil dehydrogenase activity andpesticide degradation rate. Environ Toxicol Chem 14:23–28(1995).

35 Walker A and Allen R, Influence of soil and environmentalfactors on pesticide persistence, in Soils and crop protectionchemicals, BCPC Symposium No 27, BCPC, Farnham, Surrey,UK, pp 89–100 (1984).

36 Baloch RI and Grant, RK, The investigation of degradationand metabolism of chlopyralid in two standard and threeagricultural soils, in Pesticides in soil and water, BCPCSymposium No 47, BCPC, Farnham, Surrey, UK, pp 101–106(1991).

37 Helweg A, Fomsgaard IS, Reffstrup TK and Sørensen H,Degradation of mecoprop and isoproturon in soil: influenceof initial concentration. Internat J Environ Anal Chem70:133–148 (1998).

38 Gan J, Becker RL, Koskinen WC and Buhler DD, Degradationof atrazine in two soils as a function of concentration.J Environ Qual 25:1064–1072 (1996).

39 Gan J, Koskinen WC, Becker RL and Buhler DD, Effect ofconcentration on persistence of alachlor in soil. J EnvironQual 24:1162–1169 (1995).

40 Fogg P, Boxall ABA and Walker A, Degradation of pesticidesin biobeds: the effect of concentration and pesticide mixtures.J Agric Food Chem 51(18):5344–5349 (2003).

41 Singh BK, Walker A and Wright DJ, Degradation of pesticidesin combination and their effect on soil microbial activity, inPesticide behaviour in soil and water, BCPC Symposium No 78,BCPC, Farnham, Surrey, UK, pp 145–150 (2001).

42 Nkedi-Kizza P and Brown KD, Sorption, degradation andmineralization of carbaryl in soils, for single-pesticide andmultiple pesticide systems. J Environ Qual 27:1318–1324(1998).

43 Kaufman DD, Kearney PC, Von Endt DW and Miller DE,Methylcarbamate inhibition of phenylcarbamate metabolismin soil. J Agric Food Chem 18:513–519 (1970).

44 Karanth NGK, Anderson, JPE and Domsch KH, Degradationof the herbicide diclofop-methyl in soil and influence ofpesticide mixtures on its persistence. J Biosci 6:829–837(1984).

45 Toller G and Flaim GM, Fate of pesticide residues in an organicfilter, Proc X Symp Pestic Chem, 30 September–02 October,Piacenza, Italy, pp 405–412 (1996).

46 Wauchope RD, Butler TM, Hornsby AG, Augustijn-BeckersPWM and Burt JP, SCS–ARS–CES Pesticide propertiesdatabase for environmental decision making. Rev EnvironContam Toxicol 123:1–164 (1992).

47 Tomlin CDS (ed), The Pesticide Manual, 12th edn, BCPC,Farnham, Surrey, UK (2000).

654 Pest Manag Sci 60:645–654 (online: 2004)