Bioresource Technology 99 (2008) 7988–7994

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/ locate/bior tech

Dissipation of pesticides during composting and anaerobic digestionof source-separated organic waste at full-scale plants

Thomas Kupper a,*, Thomas D. Bucheli b, Rahel C. Brändli b, Didier Ortelli c, Patrick Edder c

a Swiss College of Agriculture, SCA, Laenggasse 85, CH-3052 Zollikofen, Switzerlandb Agroscope Reckenholz-Tänikon Research Station ART, CH-8046 Zürich, Switzerlandc Food Authority Control of Geneva, CH-1211 Geneva, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history:Received 14 January 2008Received in revised form 19 March 2008Accepted 22 March 2008Available online 1 May 2008

Keywords:Organic pollutantsFungicidesTriazolesFateOrganic residues

0960-8524/$ - see front matter � 2008 Elsevier Ltd. Adoi:10.1016/j.biortech.2008.03.052

* Corresponding author. Tel.: +41 31 910 21 17; faxE-mail address: thomas.kupper@shl.bfh.ch (T. Kup

In the present study, concentration levels and dissipation of modern pesticides during composting anddigestion at full-scale plants were followed. Of the 271 pesticides analyzed, 28 were detected. Withinthe three windrows studied, total concentrations were between 36 and 101 lg per kg of dry matter(d.m.) in input materials and between 8 and 20 lg kg d.m.�1 in composts after 112 days of treatment.Fungicides and among them triazoles clearly dominated over other pesticides. More than two-thirds ofall pesticides detected in the input materials showed dissipation rates higher than 50% during compo-sting, whilst levels of most triazoles decreased slightly or remained unchanged. The investigation onsemi-dry thermophilic anaerobic digestion suggests that pesticides preferentially end up in presswaterafter solid–liquid separation.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Composting and digestion (i.e. aerobic and anaerobic treatmentof organic wastes) are important waste management strategies inEurope. The end products, compost and digestate, can be used assoil conditioner and fertilizer, thereby recycling nutrients back toagriculture and horticulture. However, compost and digestate con-tain a wide range of organic pollutants (Brändli et al., 2005,2007a,b). They enter compost via aerial deposition and splash orspray induced by road traffic on green waste, accidental input(mostly incorrect separation of input materials, e.g. plastic debris)and deliberate input (e.g. pesticides application on fruits, vegeta-bles, ornamental plants and lawn). Brändli et al. (2007b) and Buy-uksonmez et al. (2000) showed that pesticides widely occur incompost and digestate. Although concentrations of organic pollu-tants in compost and digestate were below the contents that im-pair terrestrial ecosystems, and clear negative impacts ofcompost and digestate on soil organisms were not found (Tørsløvet al., 1997; Pohl et al., in preparation) the occurrence of contami-nants raises some concern since recycling products of high qualityare expected to be free of hazardous compounds.

Composting has been widely adopted as a strategy in biodegra-dation/bioremediation of organic pollutants in recent years (Sem-ple et al., 2001). It has been shown that pesticides mightdissipate during composting of organic residues (Buyuksonmez

ll rights reserved.

: +41 21 910 22 99.per).

et al., 1999; Fogg et al., 2003; Vischetti et al., 2004; Kawata et al.,2006). Information on the fate of organic pollutants during compo-sting and digestion under full-scale conditions of green waste(originating from private gardens and public green areas) and or-ganic kitchen waste (raw organic leftovers from vegetable produc-tion and from private kitchens), i.e. ‘‘source-separated’’ organicwaste, is, however, still sparse. Brändli et al. (2007c) showed thatextrapolation of available laboratory-derived findings to real worldsystems was difficult for polychlorinated biphenyls (PCBs) andpolycyclic aromatic hydrocarbons (PAHs). Therefore, the basis forassessing the potential of composting and digestion in order to re-duce the contamination level of input materials is insufficient.Moreover, knowledge of the fate of currently used pesticides suchas triazole fungicides which are widely used commercially (Bromi-low et al., 1999) and commonly occur in compost and digestate(Brändli et al., 2007b) is not available. Therefore, the present studyaims to characterize the dissipation of pesticides at full-scaleplants. It provides a basis to better assess the potential of compo-sting and anaerobic digestion for production of compost and dige-state with a low contamination level.

2. Methods

2.1. Composting and digestion plants investigated

This study focuses on open windrow composting and semi-drythermophilic anaerobic digestion with subsequent aerobic

T. Kupper et al. / Bioresource Technology 99 (2008) 7988–7994 7989

treatment, which are widely used in the recycling of source-sepa-rated organic waste. The prevailing conditions relevant for the fateof pesticides in these systems can be considered to be representa-tive for aerobic treatment and thermophilic anaerobic digestion.Additionally, open windrow composting allows for feasible sam-pling. Dissipation of pesticides during composting was character-ized by the investigation of three different windrows on twocommercial composting plants. One windrow consisted of mostlygreen waste (windrow ‘‘CG”), the second of a mixture of greenand organic kitchen waste (windrow ‘‘CK”) and the third com-prised output material from a thermophilic anaerobic digestionplant (windrow ‘‘CDK”; Table 1). In order to prepare the windrowsto be investigated, waste materials were shredded, thoroughlymixed and heaped up to windrows with a front loader. All threewindrows were covered with air-permeable fabric, irrigated withfresh water if necessary and turned regularly during composting.

The study of pesticides dissipation during thermophilic anaero-bic digestion took place at a plant that employed the Kompogasprocess (for details on this process: see Brändli et al., 2007c). Theoutput of the fermenter undergoes solid–liquid separation result-ing in digestate (solid phase with a dry matter (d.m.) content ofabout 35%) and presswater (liquid phase with a d.m. content ofabout 12%). The digestate was submitted to subsequent aerobictreatment (open windrow composting: CDK; see above).

Details of the composting process parameters, such as temper-ature development, turning and irrigation frequencies, content ofwater, organic matter, nutrients and heavy metals, are providedin Brändli et al. (2007c).

2.2. Sampling and processing of the samples

Sampling periods at the composting plants reflected the threephases of the process: thermophilic, cooling and maturation stage(Semple et al., 2001). Accordingly, sampling took place on day 0(input material) and on day 14, 56 and 112 of composting. Beforeeach sampling, the windrows were turned at least twice. The mostsuitable sampling techniques for obtaining composite sampleswere used according to the structure of the material: digging atransect or collection of drilling cores on both sides of the windrowover the entire transect about every 3 m of the windrow. The vol-ume of aliquots was large (>>1 l) and the number thereof plentiful(>100). The composite sample was thoroughly mixed and reducedto 600 l for input material and to 60 l of material collected on day14, 56 and 112 for transport to the laboratory.

At the digestion plant input material was sampled on day 0from every shovel of the front loader filling the feeding tank (vol-ume: 54 m3) which continuously loads the fermenter over 24 h.Mixing and volume reduction of the gathered composite samplewas performed as for compost samples (see above). Material forinoculation of the fermenter (i.e. the output of the fermenter beforesolid–liquid separation) could not be sampled directly because thefermenter’s output pipe was not accessible. Instead, the digestatewas sampled continuously over half a day and the corresponding

Table 1Input material composition of the windrows investigated (further information is given in

Acronym of thewindrows

Input material

CG Green waste: 35% trimmings from trees and shrubs, 20% lawn4% soil, 3–4% residues from processing of cereals, Copazym�

CK Mixture of green and organic kitchen waste: 50% green wasmature compost

CDK Output material from a thermophilic anaerobic digestion planof a mixture of green waste, kitchen waste and approx. 10%

* Parts of the input material has been chopped and stored on a large heap over several

input presswater (50 l) was simultaneously collected from its stor-age tank. After 12 days, the output material was sampled. Press-water was again taken from the storage tank. Digestate wassampled directly from the windrow at the composting plant wherethe material underwent aerobic treatment (windrow CDK).

In the laboratory, samples collected on day 0 and 14 were cut bya shredder (Althaus, Ersigen, Switzerland, Model 300 K1) andmixed thoroughly in a concrete mixer before sub-samples(1000 g) for analysis were taken. The drying procedure was per-formed as gently as possible (at 40 �C until the weight remainedconstant, i.e. up to seven days). This low drying temperature pre-vents evaporation of the mostly non-volatile pesticides. A furtherdegradation during sample treatment and storage seems unlikelysince a significant reduction of the water content is achieved with-in the first day of drying. The samples were stored at room temper-ature in the dark. Before analysis, the samples were milled in a ballmill at a particle size of <2 mm. Further details on the sampling andprocessing of the samples are given in Brändli et al. (2007c).

2.3. Analytical methods

Within the analytical program, 271 common pesticides wereanalyzed including the following compound classes (representa-tives are given in brackets): amides (alachlor, metolachlor),anilinopyrimidines (cyprodinil), benzimidazoles (thiabendazole),carbamates (aldicarb), dicarboximides (captan, folpet), imidazoles(imazalil), organochorines (endosulfane, DDT), organophosphorus(chlorpyrifos), sulfonylureas (foramsulfuron), triazines (atrazine),triazoles (difenoconazole) and ureas (isoproturon). The exhaustivelist of analyzed compounds is given in Brändli et al. (2007b).

The analysis of the samples was performed according to Ortelliet al. (2004, 2007). To an aliquot of 20.0 g of compost or digestate10 ml of water was added and pH was adjusted to between 6 and 7.Samples were extracted by shaking 20 min with 40 ml of ethyl ace-tate followed by centrifugation. Five millilitres of the organicsupernatant was collected and evaporated to dryness on a rotaryevaporator. The dried extracts were then re-dissolved in 1 ml ofmethanol and filtered on 0.45 lm nylon membrane before doubleanalysis by liquid-chromatography coupled to tandem mass spec-trometry (LC–MS/MS). Multiple reaction monitoring (MRM) modewas used for specific detection of 160 fungicides and insecticidesfor the first injection and 80 herbicides for the second injection.Limits of detection ranged between 1 and 10 lg kg d.m.�1 for themajority of the substances and up to 50 lg kg d.m.�1 for less sensi-tive compounds.

A second extraction was carried out using an aliquot of 50 g ofsample with 100 ml of acetonitrile in a 500 ml bottle. After shakingfor 30 min, the liquid phase was collected by filtration and put in a1000 ml separatory funnel containing 500 ml of water, 10 ml ofsaturated sodium chloride solution and 50 ml of hexane. For orga-nochlorine pesticides analysis by gas chromatography with elec-tron capture detection (GC–ECD), 1 ml of the organic supernatantwas collected and diluted to 10 ml with hexane before injection.

Brändli et al., 2007c)

clippings, approx. 38% green waste from municipal source-separated collection, 3–(auxiliary agent consisting of enzymes for improvement of compost quality)*

te, 25% horse manure, 5% residues from processing of vegetables, 10% soil and 10%

t which employed the Kompogas process. The input material for digestion consistedfood residues*

days leading to anaerobic conditions in the heap.

7990 T. Kupper et al. / Bioresource Technology 99 (2008) 7988–7994

For organophosphorous pesticides analysis by gas chromatographywith nitrogen phosphorous detection (GC–NPD), 10 ml of organicsupernatant was evaporated to dryness on a rotary evaporatorand then dissolved in 2 ml of hexane before analysis on GC–NPD.A confirmation of pesticides identity was carried out by gas chro-matography with ion trap mass spectrometry detection (GC–MS)in case of positive samples using at least three characteristic ionsfor identification. Limits of detection ranged between 5 and50 lg kg d.m.�1.

2.4. Uncertainties of sampling, processing and analysis

Considering the heterogeneity of input material and compost, itis clear that artifacts are difficult to avoid in such field studies evenif optimal methods for sampling, sample processing and analysisare used. Experiments on sampling errors influencing analytical re-sults of organic compounds were conducted earlier by Breuer et al.(1997). They found total errors, i.e. combined sampling and analyt-ical errors, resulting from a compost field study of about 30% forPCBs and PAHs. Analytical errors determined by Brändli et al.(2006) for single PCBs and PAHs were in the same range. The over-all precision for the pesticide data reported here was assumed tobe similar. For the calculation of sum concentrations (Tables 2and 3), a worst case scenario was selected, i.e. detected but not

Table 2Concentrations (lg kg d.m.�1) of pesticides at two composting plants in windrows consistimaterial from a thermophilic anaerobic digestion plant (CDK) on day 0 and after 14, 56 a

Plant Composting plant 1 Comp

Substrates in windrows Green waste (CG) Green(CK)

Day 0 14 56 112 0

Concentrations in lg kg d.m.�1

Pesticides Sum 36 14 14 8 43

Fungicides Sum 30 11 11 7 29Triazole Cyproconazole 1 <1 1 1 2

Difenoconazole 1 1 2 2 1Fenbuconazole <1 <1 <1 <1 1Flusilazole <1 <1 <1 <1 <1Myclobutanil 1 1 2 1 <1Propiconazole 1 <1 3 2 2Tebuconazole 1 1 2 1 3Triadimefon <1 <1 <1 <1 <1Triadimenol <1 <1 <1 <1 <1

Morpholine Dodemorph 2 3 1 <1 2Fenpropidin <1 <1 <1 <1 <1Fenpropimorph <1 <1 <1 <1 <1

Benzimidazole Carbendazim 20 2 <1 n.a.a 4Thiabendazole <1 <1 <1 <1 2

Carbamate Propamocarb <5 <5 <5 <5 <5Imidazole Imazalil <1 <1 <1 <1 5Pyridine Pyrifenox 1 2 <1 <1 1Spiroketalamine Spiroxamine 2 1 <1 <1 2Strobilurine Azoxystrobin <1 <1 <1 <1 4

Herbicides Sum 2 1 1 – 1Triazine Terbuthylazine-2-hydroxy <1 <1 <1 <1 <1

Terbutryn 1 1 1 n.a.b <1Arylalkanoic acid Mecoprop <10 <10 <10 <10 <10Carbamate Propham <1 <1 <1 <1 <1Oxadiazole Oxadiazon <1 <1 <1 <1 1Urea Diuron 1 <1 <1 <1 <1

Insecticides Sum 2 – – – 13Carbamate Carbofuran <1 <1 <1 <1 10

Primicarb 2 <1 <1 <1 3Triazolec Paclobutrazol 2 2 2 1 <1

Numbers marked with ‘‘<” are below the limit of detection given.a Number not available, i.e. the measured concentration of 54 lg kg d.m.�1 was considb Number not available, i.e. the measured concentration of 7 lg kg d.m.�1 was considec Growth regulator.

quantifiable compounds (3 < signal to noise ratio < 10) were setequal to method quantification limits. Similarly, to follow pesticidedissipation more quantitatively, i.e. to include pesticides that wereoriginally detected (3 < signal to noise ratio < 10) but disappearedduring further treatment, all data with a signal to noise ratio >3were considered quantifiable.

2.5. Calculation of dissipation rates

To follow dissipation of organic pollutants during organic mat-ter degradation, normalization of the measured concentrations to aconservative tracer is necessary. Crude ash turned out to be themost reliable reference parameter as compared to other potentiallysuitable candidates such as heavy metals (Brändli et al., 2007c).The concentrations measured were converted to lg kg�1 crudeash as follows:

ci;norm ¼ci;meas

cash� 1000

where ci,norm denotes the normalized concentration of the com-pound i (lg kg crude ash�1), ci,meas is the quantified concentrationof the compound i (lg kg d.m.�1) and cash is the crude ash content(g kg d.m.�1) in the same sample. A dissipation rate (in %) was usedto quantify the concentration change of a compound in the sub-strate. It was calculated as follows:

ng of green waste (CG), a mixture of green and organic kitchen waste (CK) and outputnd 112 days of composting, respectively

osting plant 2

waste and kitchen waste Output material from a thermophilic anaerobicdigestion plant (CDK)

14 56 112 0 14 56 112

28 10 14 101 44 19 20

27 10 13 94 41 19 192 1 1 3 2 2 33 2 3 4 2 2 2<1 <1 <1 2 2 <1 <1<1 <1 <1 2 2 2 1<1 <1 <1 1 1 1 16 4 4 3 3 2 27 3 5 5 4 2 2<1 <1 <1 2 1 <1 <13 <1 <1 10 8 4 52 <1 <1 15 4 2 1<1 <1 <1 1 <1 <1 <1<1 <1 <1 1 <1 <1 <1<1 <1 <1 1 <1 <1 <11 <1 <1 8 3 <1 <1<5 <5 <5 <5 <5 <5 <52 <1 <1 27 7 2 2<1 <1 <1 1 1 <1 <11 <1 <1 8 1 <1 <1<1 <1 <1 <1 <1 <1 <1

1 – 1 2 2 – 1<1 <1 <1 1 <1 <1 1<1 <1 <1 <1 <1 <1 <1<10 <10 <10 <10 <10 <10 <101 <1 <1 <1 <1 <1 <1<1 <1 1 <1 1 <1 <1<1 <1 <1 1 1 <1 <1

– – – 1 – – –<1 <1 <1 <1 <1 <1 <1<1 <1 <1 1 <1 <1 <1<1 <1 <1 4 1 <1 <1

ered as an outlier and removed from the dataset.red as an outlier and removed from the dataset.

Table 3Concentrations (lg kg d.m.�1) of pesticides in a thermophilic anaerobic digestion plant

Substrates Input material (mostly kitchenwaste)

Digestate used forinoculation

Presswater used forinoculation

Presswater (sampled after 12days)

Concentrations in lg kg d.m.�1

Pesticides Sum 92 36 182 155

Fungicides Sum 70 13 173 143Triazole Cyproconazole 2 <1 5 5

Difenoconazole 2 1 7 8Fenbuconazole 1 <1 3 3Flusilazole 1 <1 4 4Myclobutanil 1 1 2 2Propiconazole 1 <1 6 3Tebuconazole 3 1 9 8Triadimefon <1 <1 2 1Triadimenol <1 <1 16 16

Morpholine Dodemorph 8 3 7 9Fenpropidin <1 <1 1 1Fenpropimorph <1 <1 <1 <1

Benzimidazole Carbendazim 21 4 <1 <1Thiabendazole 4 <1 17 14

Carbamate Propamocarb <5 <5 35 7Imidazole Imazalil 4 <1 51 53Pyridine Pyrifenox 5 2 3 3Spiroketalamine Spiroxamine 4 1 5 6Strobilurine Azoxystrobin 13 <1 <1 <1

Herbicides Sum 1 20 1 3Triazine Terbuthylazine-2-

hydroxy<1 <1 <1 1

Terbutryn <1 6 <1 <1Arylalkanoic

acidMecoprop <10 13 <10 <10

Carbamate Propham <1 <1 <1 <1Oxadiazole Oxadiazon <1 <1 <1 <1Urea Diuron 1 1 1 2

Insecticides Sum 21 1 2 2Carbamate Carbofuran 15 <1 <1 <1

Primicarb 6 1 2 2Triazolea Paclobutrazol <1 2 6 7

Numbers marked with ‘‘<” are below the limit of detection given.a Growth regulator.

Fig. 1. Number of pesticides during composting of green waste (CG), a mixture ofgreen and organic kitchen waste (CK) and output material from a thermophilicanaerobic digestion plant (CDK) on day 0, 14, 56 and 112.

T. Kupper et al. / Bioresource Technology 99 (2008) 7988–7994 7991

Dissipation rate ¼ ðci;norm;day0 � ci;norm;day112Þci;norm;day0

� 100

where ci,norm,day112 denotes the normalized concentration of thecompound i (lg kg crude ash�1) on day 112 and ci,norm,day0 is thenormalized concentration of the compound i (lg kg crude ash�1)on day 0 (i.e. in the input material).

Since anaerobic digestion is a continuous process where outputsubstrates are reintroduced to the input material to inoculate andcondition the substrate, determination of dissipation rates as car-ried out for composting is not feasible. Instead, a mass balance ofthe fermenter was calculated and dissipation rates were derivedfrom this (a description of the method is given in Brändli et al.,2007c).

Dissipation rates of 6100% indicate almost complete or exhaus-tive disappearance, whereas a rate of 0% points to the persistenceof a compound. Taking into account uncertainties of sampling, pro-cessing and analysis (Section 2.4), calculated dissipation rates of>50% include pesticides that can be removed during compostingor digestion to a relevant extent. Compounds with dissipation ratesof 650% are considered to be recalcitrant or slightly removable.

3. Results and discussion

3.1. Pesticide prevalence and concentrations

Out of the 271 pesticides analyzed, 28 were detected. The threewindrows CG, CK and CDK contained 13, 15 and 21 different com-

pounds, respectively, on day 0 and 8, 5 and 10 pesticides, respec-tively, on day 112 (Table 2, Fig. 1). These numbers of occurrencesin finished products were at the lower end of those reported byBrändli et al. (2007b) who found between 7 and 28 compoundsin 18 samples of compost and digestate. As in Brändli et al.(2007b), fungicides dominated in terms of numbers of detectedcompounds and triazoles were the most important fungicides.Compounds such as atrazine or organochlorines that formerly oc-curred in composts (Buyuksonmez et al., 2000) were not found inthe present study at contents higher than 10 lg kg d.m.�1.

7992 T. Kupper et al. / Bioresource Technology 99 (2008) 7988–7994

Total pesticide concentrations were between 36 and 101 lgkg d.m.�1 in input materials of the windrows CG, CK and CDK,and in the range of 8–20 lg kg d.m.�1 in composts on day 112 (Ta-ble 2). Such numbers are lower by a factor of 3–7 than the averageconcentration reported by Brändli et al. (2007b). Like Brändli et al.(2007b) total pesticide concentrations in compost containingkitchen waste (windrows CK, CDK) were higher than in greenwaste compost (windrow CG).

Fungicides also clearly dominated over other pesticides in termsof concentration. Total concentrations ranged from 29 to94 lg kg d.m.�1 for input materials and from 7 to 19 lg kg d.m.�1

for composts on day 112 (Table 2). Again, these numbers areroughly 2–7 times lower than the average concentration of42 lg kg d.m.�1 reported by Brändli et al. (2007b). In addition tothe dominating triazoles accounting for up to 100% of the totalamount of pesticides determined (windrows CG, CK on day 112),the post-harvest fungicides imazalil and thiabendazole were otherimportant compounds. Together, they constituted up to 37% of to-tal pesticides concentrations in input material and up to 11% incompost on day 112 for windrow CDK. It should be noted that onlyone of the 28 detected compounds belongs to the 24 most widelyused pesticides in Switzerland (SGCI, 2006), namely mecoprop,which was detected only once and which ranks 20th in the 2006list. Moreover, the most prevalent triazoles play a minor role inSwitzerland, accounting for 4% at most of all fungicides sold. A rel-evant fraction of the pesticides found in composts (mainly imazalil,thiabendazole and probably triazoles) are usually used in tropicalfruits and vegetables. This indicates that they may stem fromabroad.

The sum of pesticides’ concentration in the input material at thedigestion plant was 92 lg kg d.m.�1 (Table 3). The concentrationsof the digestate used for inoculation and submitted to subsequentaerobic treatment (windrow CDK on day 0; Table 2) were 36 and101 lg kg d.m.�1, respectively. This coincides with the numbersof Brändli et al. (2007b) who found concentrations for digestateranging from 30 to 257 lg kg d.m.�1. There were higher concentra-tions of presswater, ranging up to 182 lg kg d.m.�1. Due to thelarge portion of kitchen waste in the input material, fungicides rep-resented the most significant fraction of pesticides in digestate andpresswater (between 76% and 95% of the total, respectively). Incontrast, herbicides accounted for 56% of the sum of pesticides indigestate used for inoculation with mecoprop as the dominantcompound. This compound is commonly used on lawns (Gereckeet al., 2002) and, therefore, it can be assumed that this digestatecontained a high portion of lawn trimmings or other materialstreated with mecoprop.

This study provides for the first time information on the dis-tribution of pesticides between digestate and presswater. Thesetwo output products from thermophilic anaerobic digestion arematrices with different properties. Presswater contains a largeamount of fine particles. Moller et al. (2002) found that morethan 80% of the particles were in the fine fraction (2–25 lmin diameter) for separated liquids from a decanting centrifugeand a screw press of digested and non-digested pig manure. Itcan be assumed that the portion of fine particles is higher inpresswater compared to the corresponding digestate and conse-quently, exhibits a larger number of sorption sites. This wouldexplain the d.m. concentration of pesticides in presswater whichaccounts on average more than double the content in digestate(Table 3). Accordingly, higher concentrations in presswater havebeen observed for other organic compounds (e.g. PAHs) and alsofor elements such as phosphorous or copper, which preferen-tially partition onto solids (Brändli et al., 2007c). The physical–chemical properties of pesticides such as water solubility oroctanol–water partition coefficient (KOW) did not correlate withthe distribution of the pesticides between presswater and dige-

state and, therefore, seem to be of minor importance for pesti-cide levels in the output products of thermophilic anaerobicdigestion.

3.2. Dissipation of pesticides

3.2.1. Determining processesPesticides are subject to a series of physical–chemical and bio-

logical processes during composting or anaerobic digestion such assorption, leaching, volatilization, biotic and abiotic transformationor mineralization, which determine the extent of their dissipation.Sorption can lead to the formation of bound residues. A part ofthese is non-extractable and thus cannot be captured by chemicalanalysis. This is a well-known process governing the fate of pesti-cides in soil (Gevao et al., 2000) and during composting (Hartliebet al., 2003; Michel et al., 1995, 1997). For propiconazole, it hasbeen suggested that the formation of bound residues contributesto the persistence of this fungicide in soil (Kim et al., 2003). Michelet al. (1995) have found that almost 50% of 14C-ring-labeled 2,4-Dwas complexed with humic acids or the humin fraction of compostderived from yard trimmings. The dissipation rates found in thisstudy are probably driven by a combination of the processes listedabove but the experimental design did not allow for assigning therelevant processes. Knowledge obtained from former studies sug-gests, however, that biotic transformation or mineralization playsan important role, but the formation of bound residues should alsobe considered (Hartlieb and Klein, 2001; Michel et al., 1995, 1997).Volatilisation and/or leaching were found to be of minor impor-tance for certain pesticides (Ertunc et al., 2002; Michel et al.,1995) and other organic compounds (Hund et al., 1999; Sempleet al., 2001).

3.2.2. Dissipation rates during compostingTwenty five compounds of the windrows CG, CK and CDK were

detected in the input material and, therefore, enabled the calcula-tion of dissipation rates as presented in Table 4. Sixteen com-pounds exhibited dissipation rates of 96% to 6100% in at leastone windrow. Fifteen of them were not detectable after 56 daysof composting and in some cases even after 14 days (Table 2). Thisapplied to carbendazim, pyrifenox, spiroxamine and primicarb forall three windrows. On the other hand, difenoconazole, propico-nazole and tebuconazole persisted in one or two windrows. Insome cases, negative dissipation rates (i.e. higher normalized con-centrations in the compost on day 112 compared to the inputmaterial) were observed, probably due to sampling uncertainties(see Section 2.4). For comparison, levels of PCBs and high-molecu-lar-weight PAHs analyzed in the same samples remained stablewhereas less recalcitrant low-molecular-weight PAHs declined(Brändli et al., 2007c).

Out of the seven recalcitrant compounds (i.e. dissipation ratesof 650%), triazoles dominated with five representatives. This is inline with known half lives of several triazoles in soil of up to 200days or more (Bromilow et al., 1999; Gardner et al., 2000; Buergeet al., 2006). Triadimenol was described as being very persistentwhilst propiconazole dissipated to a higher extent (Bromilowet al., 1999). This conflicts with the present study, where triadime-nol was completely removed and propiconazole appeared to per-sist. Triadimenol was found more often than triadimefon andwhen occurring concomitantly, the former exhibited higher con-centrations. The same pattern was observed by Brändli et al.(2007b). This might be due to the reduction of triadimefon to tria-dimenol (Bromilow et al., 1999). Recalcitrant compounds duringcomposting such as triazoles tended to exhibit considerably higherhalf lives in soils than those of readily dissipating compounds(Tomlin, 1997). However, correlations between dissipation ratesand available soil half lives were not found.

Table 4Dissipation rates of pesticides during composting of green waste (CG), a mixture of green and organic kitchen waste (CK) and output material from a thermophilic anaerobicdigestion plant (CDK) after 112 days of composting

Chemical class Compound CG CK CDK

Fungicides Triazole Cyproconazole 1–50% 51–95% 1–50%Difenoconazole 0% 0% 51–95%Fenbuconazole n.a. 96% to 6100% 96% to 6100%Flusilazole n.a. n.a. 51–95%Myclobutanil 1–50% n.a. 1–50%Propiconazole 0% 0% 51–95%Tebuconazole 1–50% 0% 51–95%Triadimefon n.a. n.a. 96% to 6100%Triadimenol n.a. n.a. 51–95%

Morpholine Dodemorph 96% to 6100% 96% to 6100% 51–95%Fenpropidin n.a. n.a. 96% to 6100%Fenpropimorph n.a. n.a. 96% to 6100%

Benzimidazole Carbendazim 96% to 6100% 96% to 6100% 96% to 6100%Thiabendazole n.a. 96% to 6100% 96% to 6100%

Imidazole Imazalil n.a. 96% to 6100% 51–95%Pyridine Pyrifenox 96% to 6100% 96% to 6100% 96% to 6100%Spiroketalamine Spiroxamine 96% to 6100% 96% to 6100% 96% to 6100%Strobilurine Azoxystrobin n.a. 96% to 6100% n.a.

Herbicides Triazine Terbuthylazine-2-h. n.a. n.a. 1–50%Terbutryn 96% to 6100% n.a. n.a.

Oxadiazole Oxadiazon n.a. 1–50% n.a.Urea Diuron 96% to 6100% n.a. 96% to 6100%

Insecticides Carbamate Carbofuran n.a. 96% to 6100% n.a.Primicarb 96% to 6100% 96% to 6100% 96% to 6100%

Growth regulator Triazole Paclobutrazol 51–95% n.a. 96% to 6100%

n.a., not available.

T. Kupper et al. / Bioresource Technology 99 (2008) 7988–7994 7993

Dissipation rates varied between the three windrows (Fig. 2). Inwindrow CG and CDK, about half of the initially detected pesticidesexhibited dissipation rates of 96% to 6100% whilst the portion ofpesticides which was not detectable in CK until day 112 was at67%. About one-third of the compounds were reduced by 51–95%in CDK. The same extent of reduction was achieved by less than10% of the substances in CG and CK. The portion of pesticidesexhibiting a dissipation rate of 650% was higher for CG and CKthan for CDK. The highest dissipation rates were observed in CDKfor compounds detected in all three windrows. This was most pro-nounced for some of the recalcitrant compounds (mainly triazoles).It seems therefore that processes driving dissipation of pesticides

Fig. 2. Percentage of total numbers of pesticides exhibiting different dissipationrates (i.e. 96% to 6100%, 51–95%, 1–50% and 0%) during composting of green waste(CG), a mixture of green and organic kitchen waste (CK) and output material from athermophilic anaerobic digestion plant (CDK) after 112 days.

were enhanced in CDK, compared to CG and CK. Since parameterscharacterizing the composting process (e.g. content of water, or-ganic matter or nutrients and temperature) did not differ signifi-cantly between the windrows, it can be hypothesized that thehigher dissipation of CDK was due to the different properties ofthe input materials. The preceding anaerobic treatment of CDK in-put material might have resulted in a higher degradation orsequestration potential during subsequent composting. A possibleexplanation is the better accessibility of pesticides for micro-organisms after anaerobic treatment. It might also be hypothesizedthat anaerobically digested material provides a lower portion ofeasily degradable substrates compared to fresh organic waste,thereby enhancing co-metabolism of organic compounds such aspesticides. In contrast to the present study, Vorkamp et al. (2003)did not observe any degradation of dodemorph during subsequentaerobic treatment of anaerobically digested substrate on labora-tory scale over eight days. This discrepancy might be due to a high-er efficiency of microbial processes at a full-scale plant, the shortperiod for aerobic treatment and/or the composition of digestedmaterial used by Vorkamp et al. (2003).

3.2.3. Dissipation rates during anaerobic digestionThe mass balance resulted in negative dissipation rates for 20

out of 24 compounds detected in the input material, and 15 ofthem were out of the range of usual uncertainties due to sampling,processing and analysis of about 30%. This points to methodologi-cal difficulties leading to biased results. It seems likely that the res-idence time in the fermenter was not appropriately determined.This is supported by the calculated mass loss during digestionbeing lower (9.6% based on d.m.) than the normal degradation rateof 13% (Leisner, M., Kogas, A.G., personal communication; for fur-ther information see Brändli et al., 2007c). Similar problems oc-curred for PCB and PAH in identical samples (Brändli et al.,2007c). Additionally, it should be noted that anaerobic digestionof organic waste is a continuous process with internal flows. Thecharacterization of the fate of organic compounds in similar sys-

7994 T. Kupper et al. / Bioresource Technology 99 (2008) 7988–7994

tems such as wastewater treatment requires repeated samplingcampaigns. Future studies should take this into consideration.

4. Conclusions

In the present study, dissipation of currently used pesticidesduring composting at full-scale plants was observed. More thantwo-thirds of all pesticides detected in the input materials showeddissipation rates higher than 50%. In contrast, levels of triazoles re-mained largely unchanged. This gives strong evidence of persis-tence during composting of these widely used fungicides.

The investigation on semi-dry thermophilic anaerobic digestionsuggests that pesticides preferentially end up in presswater com-pared to digestate. However, further research is needed to assessthoroughly the fate of pesticides during anaerobic digestion usingan appropriate experimental design.

Acknowledgements

The Federal Office for the Environment and the Swiss FederalOffice of Energy are acknowledged for the financial support, andthe personnel from the plants at Oensingen, Fehraltorf and Oetwilam See are thanked for their good collaboration. We are grateful toT. Poiger Agroscope Changins-Wädenswil Research Station ACWfor revision of the manuscript.

References

Brändli, R.C., Bucheli, T.D., Kupper, T., Furrer, G., Stadelmann, F.X., Tarradellas, J.,2005. Persistent organic pollutants in source-separated compost and itsfeedstock materials – a review of field studies. J. Environ. Qual. 34, 735–760.

Brändli, R.C., Bucheli, T.D., Kupper, T., Stadelmann, F.X., Tarradellas, J., 2006.Optimized accelerated solvent extraction of PCBs and PAHs from compost.Intern. J. Environ. Anal. Chem. 86, 505–525.

Brändli, R.C., Bucheli, T.D., Kupper, T., Furrer, R., Stahel, W., Stadelmann, F.X.,Tarradellas, J., 2007a. Organic pollutants in Swiss compost and digestate. 1.Polychlorinated biphenyls, polycyclic aromatic hydrocarbons and molecularmarkers, determinant processes, and source apportionment. J. Environ. Monit.9, 456–464.

Brändli, R.C., Kupper, T., Bucheli, T.D., Zennegg, M., Huber, S., Ortelli, D., Müller, J.,Schaffner, C., Iozza, S., Schmid, P., Berger, U., Edder, P., Oehme, M., Stadelmann,F.X., Tarradellas, J., 2007b. Organic pollutants in Swiss compost and digestate; 2.Polychlorinated dibenzo-p-dioxins, and -furans, dioxin-like polychlorinatedbiphenyls, brominated flame retardants, perfluorinated alkyl substances,pesticides, and other compounds. J. Environ. Monit. 9, 465–472.

Brändli, R.C., Bucheli, T.D., Kupper, T., Mayer, J., Stadelmann, F.X., Tarradellas, J.,2007c. Fate of PCBs, PAHs and their source characteristic ratios duringcomposting and digestion of source-separated organic waste in full-scaleplants. Environ. Pollut. 148, 520–528.

Breuer, J., Drescher, G., Schenkel, H., Schwadorf, K., 1997. Hohe Kompostqualität istmöglich, Räumliche und zeitliche Variabilität von Kompostinhaltsstoffen,Begleituntersuchungen zum Kompostierungserlass des Landes Baden-Württemberg. Ministerium für Umwelt und Verkehr. Universität Hohenheim,Landesanstalt für landwirtschaftliche Chemie.

Bromilow, R.H., Evans, A.A., Nicholls, P.H., 1999. Factors affecting degradation ratesof five triazole fungicides in two soil types: 1. Laboratory incubations. Pestic.Sci. 55, 1129–1134.

Buerge, I.J., Poiger, T., Muller, M.D., Buser, H.R., 2006. Influence of pH on thestereoselective degradation of the fungicides epoxiconazole and cyproconazolein soils. Environ. Sci. Technol. 40, 5443–5450.

Buyuksonmez, F., Rynk, R., Hess, T.F., Bechinski, E., 1999. Occurrence, degradationand fate of pesticides during composting – Part I. Composting, pesticides, andpesticide degradation. Compost Sci. Util. 7, 66–82.

Buyuksonmez, F., Rynk, R., Hess, T.F., Bechinski, E., 2000. Occurrence, degradationand fate of pesticides during composting. Part II. Occurrence and fate ofpesticides in compost and composting systems. Compost Sci. Util. 8, 61–81.

Ertunc, T., Hartlieb, N., Berns, A., Kliens, W., Schaeffer, A., 2002. Investigations on thebinding mechanism of the herbicide simazine to dissolved organic matter inleachates of compost. Chemosphere 49, 597–604.

Fogg, P., Boxall, A.B.A., Walker, A., 2003. Degradation of pesticides in biobeds: theeffect of concentration and pesticide mixtures. J. Agric. Food Chem. 51, 5344–5349.

Gardner, D.S., Branham, B.E., Lickfeldt, D.W., 2000. Effect of turfgrass on soilmobility and dissipation of cyproconazole. Crop Sci. 40, 1333–1339.

Gerecke, A.C., Scharer, M., Singer, H.P., Muller, S.R., Schwarzenbach, R.P., Sagesser,M., Ochsenbein, U., Popow, G., 2002. Sources of pesticides in surface waters inSwitzerland: pesticide load through waste water treatment plants-currentsituation and reduction potential. Chemosphere 48, 307–315.

Gevao, B., Semple, K.T., Jones, K.C., 2000. Bound pesticide residues in soils: a review.Environ. Pollut. 108, 3–14.

Hartlieb, N., Klein, W., 2001. Fate and behaviour of organic contaminants duringcomposting of municipal biowaste. In: Rees, R.M., Ball, B.C., Campbell, C.D.,Watson, C.A. (Eds.), Sustainable Management of Soil Organic Matter. CABIPublishing CAB International, Oxon, UK, pp. 150–156.

Hartlieb, N., Ertunc, T., Schaeffer, A., Klein, W., 2003. Mineralization metabolism andformation of non-extractable residues of 14C-labelled organic contaminantsduring pilot-scale composting of municipal biowaste. Environ. Pollut. 126, 83–91.

Hund, K., Kurth, H.H., Wahle, U., 1999. Entwicklung einer Untersuchungs- undBewertungsstrategie zur Ableitung von Qualitätskriterien für Komposte.Fraunhofer-Institut für Umweltchemie und Ökotoxikologie, Schmallenberg,Germany.

Kawata, K., Nissato, K., Shiota, N., Hori, T., Asada, T., Oikawa, K., 2006. Variation inpesticide concentrations during composting of food waste and fowl droppings.Bull. Environ. Contam. Toxicol. 77, 391–398.

Kim, I.S., Shim, J.H., Suh, Y.T., 2003. Laboratory studies on formation of boundresidues and degradation of propiconazole in soils. Pest Manag. Sci. 59, 324–330.

Michel, F.C., Reddy, C.A., Forney, L.J., 1995. Microbial-degradation and humificationof the lawn care pesticide 2,4-dichlorophenoxyacetic acid during thecomposting of yard trimmings. Appl. Environ. Microbiol. 61, 2566–2571.

Michel, F.C., Reddy, C.A., Forney, L.J., 1997. Fate of carbon-14 diazinon during thecomposting of yard trimmings. J. Environ. Qual. 26, 200–205.

Moller, H.B., Sommer, S.G., Ahring, B.K., 2002. Separation efficiency and particle sizedistribution in relation to manure type and storage conditions. Bioresour.Technol. 85, 189–196.

Ortelli, D., Edder, P., Cognard, E., 2007. Recent advances in pesticides residueanalyses in food and environmental samples. Mitt. Lebensm. Hyg. 97, 275–287.

Ortelli, D., Edder, P., Corvi, C., 2004. Multiresidue analysis of 74 pesticides in fruitsand vegetables by liquid chromatography–electrospray-tandem massspectrometry. Anal. Chim. Acta 520, 33–45.

Pohl, M., Staempfli, C., Niang, F., Kupper, T., Stadelmann, F.X., Tarradellas, J., Becker-van Slooten, K., in preparation. Ecotoxicological tests applied to compost:observation of inhibiting and stimulating effects. Appl. Soil Ecol.

Semple, K.T., Reid, B.J., Fermor, T.R., 2001. Impact of composting strategies on thetreatment of soils contaminated with organic pollutants. Environ. Pollut. 112,269–283.

SGCI, 2006. Schweizerische Gesellschaft für chemische Industrie.Tomlin, C.D.S., 1997. The Pesticide Manual, 11th ed. The British Crop Protection

Council.Tørsløv, J., Samsøe-Petersen, L., Rasmussen, J.O., Kristensen, P., 1997. Use of waste

products in agriculture. Contamination level, environmental risk assessmentand recommendations for quality criteria. Miljøprojekt nr. 366. Ministry ofEnvironment and Energy, Denmark.

Vischetti, C., Capri, E., Trevisan, M., Casucci, C., Perucci, P., 2004. Biomassbed: abiological system to reduce pesticide point contamination at farm level.Chemosphere 55, 823–828.

Vorkamp, K., Taube, J., Dilling, J., Kellner, E., Herrmann, R., 2003. Fate of thefungicide dodemorph during anaerobic digestion of biological waste.Chemosphere 53, 505–514.