Chemosphere 84 (2011) 369–375

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Effect of fluctuating soil humidity on in situ bioavailability and degradationof atrazine

Anastasiah Ngigi a,b, Ulrike Dörfler b, Hagen Scherb c, Zachary Getenga a, Hamadi Boga d, Reiner Schroll b,⇑a Department of Physical Sciences, Masinde Muliro University of Science and Technology, Kakamega, Kenyab Helmholtz Zentrum München, German Research Center for Environmental Health (GmbH), Institute of Soil Ecology, 85764 Neuherberg, Germanyc Helmholtz Zentrum München, German Research Center for Environmental Health (GmbH), Institute of Biomathematics and Biometry, 85764 Neuherberg, Germanyd Department of Botany, Jomo Kenyatta University of Agriculture and Technology, Nairobi, Kenya

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

Article history:Received 10 January 2011Received in revised form 30 March 2011Accepted 31 March 2011Available online 4 May 2011

Keywords:Drying–rewetting cyclesIn situ bioavailabilityMetabolismMineralization

0045-6535/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.chemosphere.2011.03.068

⇑ Corresponding author. Tel.: +49 89 3187 3319; faE-mail address: schroll@helmholtz-muenchen.de (

This study elucidates the effect of fluctuating soil moisture on the co-metabolic degradation of atrazine(6-chloro-N2-ethyl-N4-isopropyl-1,3,5-triazine-2,4-diamine) in soil. Degradation experiments with 14C-ring-labelled atrazine were carried out at (i) constant (CH) and (ii) fluctuating soil humidity (FH). Tem-perature was kept constant in all experiments. Experiments under constant soil moisture conditions wereconducted at a water potential of �15 kPa and the sets which were run under fluctuating soil moistureconditions were subjected to eight drying–rewetting cycles where they were dried to a water potential ofaround �200 kPa and rewetted to �15 kPa. Mineralization was monitored continuously over a period of56 d. Every two weeks the pesticide residues in soil pore water (PW), the methanol-extractable pesticideresidues, the non-extractable residues (NER), and the total cell counts were determined. In the soil withFH conditions, mineralization of atrazine as well as the formation of the intermediate product deisopro-pyl-2-hydroxyatrazine was increased compared to the soil with constant humidity. In general, we found asignificant correlation between the formation of this metabolite and atrazine mineralization. The cellcounts were not different in the two experimental variants. These results indicate that the microbialactivity was not a limiting factor but the mineralization of atrazine was essentially controlled by the bio-availability of the parent compound and the degradation product deisopropyl-2-hydroxyatrazine.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Pesticides are favourably degraded by microorganisms in soil(Rüdel et al., 1993). But it has to be considered that for most chem-icals just a certain amount of these compounds is bioavailable andonly this portion can be degraded by microorganisms (Katayamaet al., 2010). The microbial breakdown is not only dominated bythe activity of the microbes but also by the mass transfer of pesti-cides to microorganisms (Bosma et al., 1997) and the prevalentwater regime in soils (Han and New, 1994). On the microscopicscale, pesticides and pesticide degrading microorganisms are dif-ferently distributed in soils, and pesticides must mostly diffuse tothe more or less immobile microbes to be metabolized by them.Hampered mass transfer of chemicals to the degrading microbescould therefore be a limiting factor in biodegradation (Bosmaet al., 1997). Thus, soil moisture is one of the most importantparameters regulating pesticide bioavailability and degradation.

ll rights reserved.

x: +49 89 3187 3376.R. Schroll).

Differences in soil moisture seem to have a more intensive effecton the degradation of chemicals than differences in soil tempera-ture. Degradation half-life of isoproturon increased by a factor of10–15 when soil water potential was reduced from �660 MPa to�56 kPa in soil material of a clay-enriched illuvial horizon,whereas a change in soil temperature from 10 �C to 22 �C was ofminor importance (Alletto et al., 2006).

In laboratory experiments, a soil water potential of �15 kPawas identified for optimal pesticide mineralization (Ilstedt et al.,2000; Schroll et al., 2006). In contrast to laboratory studies, wheresoil humidity in most cases is constant, under outdoor conditionsthe soil is exposed to varying water regimes. Depending on thesoil properties, precipitation events, air temperature, and uptakeby plants, soil moisture will vary to a great extent during thevegetation period. These variations in soil water will have a con-siderable effect on sorption behaviour of pesticides, their bioavail-ability, and finally their transformation and degradation. Diverseand even contrary results of drying–rewetting are reported in lit-erature, varying from favouring, inhibiting, or no effects of soildrying on the sorption and degradation behaviour of pesticidesin soils. García-Valcárcel and Tadeo (1999) found lower concen-trations of hexazinone and simazine in soil solution after several

370 A. Ngigi et al. / Chemosphere 84 (2011) 369–375

drying–rewetting cycles (from 90% to 20% field capacity) in a san-dy loam soil. In five soils ranging from sandy loam to clay, theadsorption of the herbicide imazaquin was increased when thesoils were dried to 50%, 25% or 0% of field capacity with subse-quent returning to field capacity (Goetz et al., 1986). The in-creased imazaquin adsorption was explained by reduction inwater film thickness coating the soil minerals and thus favouringthe access of imazaquin to active surfaces. The effect of dryingand rewetting cycles on the desorption behaviour of diuron andterbuthylazine was studied in detail by Lennartz and Louchart(2007). They performed classical batch experiments and adsorp-tion was followed by one to three drying cycles (<3% water con-tent) before conducting the desorption experiments. Dryingresulted in decreased desorption of the pesticides. The authorssuggest that drying cause structural modifications of the soil or-ganic matter due to shrinking-like processes, which impede diffu-sion and the pesticide molecule is trapped in micropore spaces.Summarizing these findings, the authors found a higher sorptionof the selected pesticides after soil drying. White et al. (1998)found inhibiting as well as favouring as well as no effects of dry-ing–wetting cycles (‘‘dried to constant weight’’ and ‘‘remoistenedto �1.1 bar’’) on the mineralization of phenanthrene and di(2-eth-ylhexyl)phthalate, depending on the incubation and aging periodsfor both compounds.

In contrast to the upper cited literature, the following findingswould suggest a potentially higher bioavailability and degradationof pesticides under FH conditions. Belliveau et al. (2000) postulatedthat the intraparticle diffusion of a dissolved pesticide must be pre-ceded by the intraparticle diffusion of water. In laboratory experi-ments, Belliveau et al. (2000) and Gamble et al. (2000) coulddemonstrate that after drying the mobility of water in an organicrich soil was reduced and only very slow. The mobility of the pes-ticides 2,4-D and atrazine was comparable to the slow water pen-etration (Belliveau et al., 2000).

The aim of the present study was to further elucidate the impactof FH conditions on pesticide bioavailability, pesticide degradationand pesticide mineralization in an agricultural soil. The pesticideselected for this study was atrazine (6-chloro-N2-ethyl-N4-isopro-pyl-1,3,5-triazine-2,4-diamine), which was used globally in thepast to control pre- and post-emergence broadleaf and grassyweeds in major crops. Atrazine can be very persistent in soil(Jablonowski et al., 2009) and it is by far the most frequently foundxenobiotica in ground water and increasing restrictions of its usehave been introduced (Premazzi and Stecchi, 1990).

2. Materials and methods

2.1. Soils

The soil material was a humic cambisol from an agriculturalfield (Kelheim; latitude 48.917�, longitude 11.867�, altitude348 m) in Germany without atrazine history for the past 20 years.Therefore, we could expect to select a soil with co-metabolic andthus non-growth-linked degradation dynamic; using soil materialwith a growth-linked degradation dynamic might have compli-cated the interpretation of the results. Moreover, co-metabolicdegradation dynamic is still the rule but not the exception in agri-cultural soils (Krutz et al., 2010). The relevant soil characteristicsare: clay (<2 lm) 11%, silt (2–63 lm) 19%, sand (63 lm–2 mm)70%, org. C 1.3%, total N 0.1%, pH (CaCl2) 6.9, and water contentof 18.1% at a water potential of �15 kPa and a soil density of1.3 g cm�3. Before the experiments were started, the soil samples(depth 0–10 cm) were sieved (<2 mm) and kept at room tempera-ture (20 ± 1 �C) for 5 d after moistening to a water potential closeto but below �15 kPa.

2.2. Chemicals

14C-ring-labelled atrazine, with a specific radioactivity of351.5 MBq mmol�1 and a radiochemical purity of >98% was pur-chased from Sigma–Aldrich (St. Louis, MO, USA). Non-labelledatrazine and the metabolite standards deethylatrazine (DEA),deisopropylatrazine (DIA), deethyl-deisopropylatrazine (DEDIA),2-hydroxyatrazine (OH-ATR), deethyl-2-hydroxyatrazine (OH-DEA), deisopropyl-2-hydroxyatrazine (OH-DIA), and deethyl-deisopropyl-2-hydroxyatrazine (OH-DEDIA) were obtained fromEhrenstorfer (Augsburg, Germany). All chemicals had a purity of>99%. Scintillation cocktails were obtained from Packard (Dreieich,Germany). All other chemicals and solvents were of analyticalgrade and were purchased from Merck (Darmstadt, Germany).

2.3. Degradation experiments

2.3.1. Pesticide applicationAtrazine degradation experiments were carried out in an aer-

ated closed laboratory system as described previously (Schrolland Kühn, 2004) with 50 g soil (dry weight; d.w.) in 100 mL doublewalled glass incubation vessels. 14C-labelled and non-labelled atra-zine were mixed and dissolved in methanol to give a final specificradioactivity of 66.3 Bq lg�1. A volume of 0.5 mL of this applicationstandard with a radioactivity of 33.1 kBq was applied dropwisewith a Hamilton syringe to an aliquot of 3.5 g of oven-dry, grindedsoil. The aliquot was stirred with a spatula for 2 min until a homog-enous distribution of the herbicide was achieved. After evaporationof methanol the aliquot was mixed for another 2 min with freshsoil (46.5 g dry weight) yielding a total sample amount of 50 gdry soil per experiment with a pesticide concentration of10 mg kg�1. This concentration corresponds to a realistic agro-nomic application rate of 500 g ha�1 when assuming that the dis-tribution of the herbicide might be at 3–4 mm depth shortlyafter the application in a field with a soil density of approximately1.3 g cm�3. The soils were then transferred to the incubation flasks,compacted to a density of 1.3 g cm�3 and water content was ad-justed to a water potential of �15 kPa to obtain maximum pesti-cide mineralization at CH conditions (Schroll et al., 2006).

2.3.2. CH and FH experimental set upFor CH conditions the incubation vessels were placed in the

dark at 20 ± 1 �C and connected to a trapping system. Three timesper week humidified air (1.0 L h�1) was drawn via a pump throughthe system for 1 h. After passing through the flasks, the air wastrapped in a series of four wash bottles filled with 10 mL 0.1 MNaOH to trap 14CO2 from mineralization process. After each aera-tion the trapping solution was collected and the traps were filledwith fresh 0.1 M NaOH solution. The weight of the incubation ves-sels was monitored gravimetrically every week to control for anywater loss.

For experiments with FH conditions, the incubation vesselswere placed in the dark at 20 ± 1 �C and connected to a trappingsystem. For the drying cycle non-humidified air (3.0 L h�1) wascontinuously drawn via a pump through the system for one week.The air, having passed through the flasks, was sampled and trappedas it was described above. During the drying cycles the incubationflasks were weighed daily to monitor the water loss gravimetri-cally. After one week of drying, the rewetting of the soil was doneimmediately after weighing the incubators to adjust the water con-tent to the original value corresponding to a water potential of�15 kPa. After rewetting, the next drying cycle was started. In to-tal, eight drying and rewetting cycles were conducted.

From the collected 0.1 M NaOH solutions of both variants,2 mL aliquots were taken and mixed with 3 mL of scintillationcocktail Ultima Flo AF (Packard, Dreieich, Germany). Radioactivity

A. Ngigi et al. / Chemosphere 84 (2011) 369–375 371

was detected by scintillation counting using a TriCarb 2800 TR(Packard, Dreieich, Germany). The detection limit for 14CO2 was0.05 lg kg�1 soil (d.w.).

2.3.3. Soil samplingAfter every two rewetting cycles, exactly 14 d, four replicates

from the experiments with CH and FH were sampled. Thirty gramsof each soil sample (d.w.) were used for PW extraction to deter-mine the pesticide bioavailability (Folberth et al., 2009). Subse-quently, these soil samples were extracted with methanol todetermine the quality and quantity of extractable residues as wellas NER. Another aliquot (1 g of each soil sample) was taken directlyfrom the incubators and was used for counting the cultivable bac-terial cells.

2.4. Pesticide in situ bioavailability

In situ bioavailability of atrazine was determined using the PWextraction method as described by Folberth et al. (2009). In brief,30 g (d.w.) of soil was transferred to a custom-built centrifuge con-tainer. This container consists of an upper and lower part. Theupper cup contained the soil sample to be extracted. In the lowerpart, PW was collected. Both pieces were connected by a canal. Aglass frit with 120 lm average pore size was placed underneaththe soil sample to prevent the canal from being clogged by soil par-ticles. The top of the upper cup was sealed with aluminum foil andclosed with an aluminum cap. Centrifugation was carried out usinga Beckman J2-21 centrifuge and a Beckman JA-14 rotor (Beckman,Krefeld, Germany) at 21 �C for 90 min with a relative centrifugalforce of 9000g. Radioactivity in the extracted water was measuredby liquid scintillation counting using 4 mL of the scintillation cock-tail Ultima Gold XR (Packard, Dreieich, Germany) and 0.1 mL PWaliquot to quantify the dissolved pesticide concentration in soilsolution. The total dissolved pesticide amount was calculated bymultiplication of the concentration of the aliquot with the totalamount of water in the soil (Folberth et al., 2009). The PW sampleswere stored at �20 �C before HPLC analysis.

2.5. Solvent extraction of soil, clean up and analysis

After centrifugation the same soil aliquots were extracted withmethanol in an accelerated solvent extractor (ASE 200, Dionex,Idstein, Germany) at 90 �C, with a pressure of 10 MPa (Gan et al.,1999). Aliquots of 0.5 mL of each extract were mixed with 4.5 mLUltima Gold XR and measured by liquid scintillation counting. Sub-sequently, extracts were concentrated with a rotary evaporator toa volume of 2–3 mL.

The concentrated methanolic soil extracts were dissolved in250 mL distilled water. These solutions as well as samples of PWwere cleaned up with Isolate Triazine columns (500 mg, Separtis,Grenzach-Wyhlen, Germany). After extraction, the SPE columnswere dried under a gentle nitrogen-stream and eluted with10 mL methanol. The eluate was concentrated to a volume of1 mL with a rotary evaporator and further concentrated to a vol-ume of 0.2 mL under a gentle nitrogen-stream. The samples wereimmediately analyzed by HPLC or stored at �20 �C before analysis.

For residue analysis 20 lL of each soil extract or PW samplewere injected to a HPLC system that was equipped with a L-6200Intelligent Pump (Merck-Hitachi, Darmstadt, Germany), a UV/VISdetector (220 nm, Merck-Hitachi, Darmstadt, Germany), and aradioactivity detector LB 506 C1 (Berthold, Wildbad, Germany);column: LiChrospher 100 RP-18, 2 lm, 4 � 250 mm (Merck-Hitachi, Darmstadt, Germany). The mobile phase consisted of0.003 M KH2PO4, pH 3 (A) and acetonitrile (B) at a flow rate of0.8 mL min�1. The gradient program was: T0 min 20% (A);T10 min 38% (A); T24 min 75% (A); T29 min 75% (A); T33 min

20% (A); T40 min 20% (A). Parent compound and metabolites wereidentified by comparison of their retention times with referencesubstances. The method detection limits – based on radioactivitydetection – were as follows: atrazine: 23.5 lg kg�1 soil (d.w.);DEA: 20.5 lg kg�1; DIA: 18.9 lg kg�1; OH-DIA: 16.9 lg kg�1.

2.6. Quantification of 14C-labelled NER

After ASE, soil material was dried and homogenized intensively.Three aliquots of each soil sample were filled into combustion cupsand mixed with 3–4 drops of saturated aqueous sugar solution toguarantee a complete oxidation of the 14C. The combustion wasconducted with an automatic sample-oxidizer 306 (Packard,Dreieich, Germany). 14CO2 was trapped in Carbo- Sorb E (Packard,Dreieich, Germany) and mixed with Permafluor E (Packard, Drei-eich, Germany) prior to scintillation counting. The detection limitfor NER was: 9.0 lg kg�1 soil (d.w.).

2.7. Bacterial cell counting

For extraction of bacterial cells from soil, the following solutionwas used: 0.1 g NaCl, 0.02 g CaCl�2H2O, 0.2 g MgSO4�7H2O, 5.0 gTween 80. This solution was adjusted to 1 L with Milli Q waterand autoclaved for 20 min at 121 �C. Soil bacteria were extractedfrom the soil by mixing 1 g fresh soil with 99 mL of the extractionsolution in a 200 mL jar. The mixture was shaken vigorously at150 rpm for 1 h. The soil particles were then allowed to sedimentfor 10 min before several dilution steps were conducted. Dilutionsof 10�2 were spread in duplicates on plates with LB medium. TheLB was prepared by mixing 10 g trypton enzymatic digest fromcasein, 5 g yeast extract, 5 g NaCl, 15 g Agar and 0.1 mg cyclohex-imide (per 1 L Milli Q water). The mixture was sterilized for20 min at 121 �C. The number of colony forming units (CFU) wasdetermined after 3 d of incubation at 25 �C.

2.8. Data analysis

For statistical data analyses, the program package SAS 9.1 wasused to set up regression models for trend analyses with procedureREG. Time varying treatment effects were modelled as main ef-fects, eventually adjusted for interaction of independent treatmentvariables with time (days). For ordinary or inverse-varianceweighted mean value comparisons, procedure TTEST was used.(SAS Institute Inc: SAS/STAT User’s Guide, Version 9.1. Cary NC:SAS Institute Inc.; 2003).

3. Results and discussion

3.1. Soil water potential

In the CH experimental set up, the soil water potential was keptconstant at �15 kPa, whereas in the FH experimental set up, thewater potential varied between �15 kPa and around �200 kPa ateach drying and rewetting cycle (Fig. 1). Thus, the water regimein the various experimental set ups was quite different and theselected drying method was appropriate in achieving an efficientand reproducible reduction in soil humidity under laboratoryconditions.

3.2. Mineralization of atrazine

The cumulative mineralization at the end of the investigationperiod of 56 d was relatively low for both, CH and FH (Fig. 2). Underconstant water regime 3.4% and under fluctuating water regime5.1% of the applied pesticide was mineralized. Since the soil had

0

50

100

150

200

250

300

0 10 20 30 40 50 60

Time (d)

soil

wat

er p

oten

tial (

kPa)

Fig. 1. Water tension of the soil during the drying–rewetting cycles. Bars indicatestandard deviation of 16 (beginning) to four (end) samples.

0

1

2

3

4

5

6

0 10 20 30 40 50 60Time (d)

cum

. min

eral

izat

ion

(% o

f app

l. 14C

)

constant soil humidity (CH)

fluctuating soil humidity (FH)

Fig. 2. Cumulative mineralization of 14C-atrazine in a humic cambisol underconstant (CH) and fluctuating soil humidity (FH). Bars indicate standard deviationof 16 (beginning) to four (end) samples, respectively.

372 A. Ngigi et al. / Chemosphere 84 (2011) 369–375

no history of atrazine use for the past 20 years, the low mineraliza-tion is a characteristic of a co-metabolic biodegradation. However,the variant FH showed significantly (p < 0.0001) higher mineraliza-tion compared to the control (CH).

For a further explanation of the mineralization behaviour ofatrazine, every two weeks four replicates were sampled and ana-lyzed for atrazine and its transformation products (see Section2.3.3). At each sampling time, mass balances were established.The recoveries were quite good and ranged between 96.1% and96.6% of the applied 14C in the variants CH and FH, respectively.

3.3. Identification of the factors governing mineralization of atrazine

To elucidate the degradation pathway and to identify the fac-tors that govern the mineralization of atrazine in soil humic camb-isol, the amounts of extractable atrazine, atrazine metabolites,NER, and the mineralized atrazine were compared at the four sam-pling points (days 14, 28, 42 and 56). Several significant correla-tions could be identified.

3.3.1. Relationship between total atrazine and NERAs can be seen from Table 1, NER levels are increasing for both

CH and FH variants from sampling to sampling, while the respec-tive atrazine levels are deceasing in both cases. To test whetherthere is any relationship between total atrazine and the formationof NER, we calculated respective correlations between both values.There exist significant correlations between the totally extractableatrazine (atrazine in methanol extract plus atrazine in PW) and theformation of NER for both variants, CH (p = 0.0124) and FH(p = 0.0114) (Fig. 3). Both regressions were not significantlydifferent from each other as neither the main effect of CH vs. FH

(p = 0.5762) nor the interaction of atrazine with FH (p = 0.6615)were significant, and therefore data from both variants could becalculated in a common regression (y = �0.70x + 5.35; R2 = 0.98),meaning that the processes leading to NER formation are identicalfor both variants. Nevertheless, a remarkable difference betweenboth variants was identified: in the variant FH atrazine was higherand NER were lower than in the control (CH) at each respectivesampling point (Table 1 and Fig. 3). This is reflected by significantlydifferent inverse-variance weighted mean values of atrazine for CH(3.1, SE = 0.2) and for FH (4.1, SE = 0.3) (p = 0.0362). Belliveau et al.(2000) have shown that after drying soil samples the penetrationrate of both water and pesticide into micropores and strong bind-ing sites is reduced, thereby impeding the formation of NER. Thelower NER formation under FH conditions could be explained bythese mechanisms.

Since the NER formation was apparently retarded in variant FH,more atrazine was potentially available for other processes, likemineralization. In fact, the FH variant showed higher mineraliza-tion than the control: in Fig. 2, the increase of cumulative mineral-ization is significantly (p < 0.0001) accelerated under FH conditionscompared to CH conditions. In general, the formation of NER inboth variants increased considerably with time, which is in accor-dance with former results (Clay and Koskinen, 1990).

3.3.2. Relationship between atrazine and its metabolitesAtrazine is stepwise degraded until it is finally mineralized. In

methanol extracts and in PWs we found the degradation productsDIA, DEA, and OH-DIA (Table 1).

Depending on the extraction procedure we found significantcorrelations (i) for total atrazine and total DIA in CH (Fig. 4a,p = 0.0148) and (ii) for methanol extractable atrazine and DIA inFH (Fig. 4b, p = 0.0287). The other two correlations (see Fig. 4aand b) were not significant, and this was most likely caused bythe relative small amount of data pairs. As can be seen fromFig. 4a and b, with decreasing atrazine concentration, the concen-tration of DIA increased. Since the slope of the regression lineswere not statistically different (p > 0.3), it can be concluded thatthe degradation processes and transformation rates – formingDIA from atrazine – are identical for both variants. Under CH con-ditions, DIA was formed to a lower extent than in the soil under FHconditions, which was caused by the lower atrazine concentrationin the soil with CH regime. Thus, the DIA formation rate was lim-ited by the availability of atrazine in the control (CH).

N-dealkylation of s-triazines is an important degradation path-way in many microorganisms, and Behki and Kahn (1986) foundthat the formation of deisopropylatrazine is favoured over the for-mation of deethylatrazine. In our study, we also found much highertotal DIA residues than total DEA residues (Table 1).

3.3.3. Relationship between DIA, OH-DIA and mineralized atrazineThe intermediate product DIA was compared with the amount

of mineralized atrazine. For each 14-d sampling interval, the indi-vidual 14-d-mineralization-rates were calculated for both variants;for this purpose the measured 14CO2 radioactivity was transformedvia the respective molecular weight into atrazine equivalents andcorrelated with the total DIA residues at each sampling point. Sincethe results above showed in both variants identical processes bywhich atrazine is degraded to DIA, the following correlations werecalculated with the common data pool from both variants. Fig. 5ashows a significant positive correlation (p = 0.0045) between DIAand pesticide mineralization.

Most of the above calculated correlations were conducted withtotal residues of atrazine and DIA in methanol extract plus PW:these total residues represent the potentially available atrazineand DIA residues. Since we found a significant correlation betweenDIA and pesticide mineralization, and since microbial degradation

Table 1Concentrations of atrazine and degradation products in methanol (MeOH) extract and pore water (PW), cumulative mineralization and non extractable residues (NER) in soilswith constant (CH) and fluctuating moisture (FH) conditions at 4 different sampling times: after 14 days (T14), 28 days (T28), 42 days (T42), 56 days (T56). n = 4 ± standarddeviation. Pore water: HPLC analysis was conducted with pooled samples from 4 replicates to exceed detection limit, therefore no standard deviation can be given.

Moisture condition T14 T28 T42 T56

Atrazin and degradation products MeOH extract PW MeOH extract PW MeOH extract PW MeOH extract PW

CH Atrazine (lg g�1) 4.53 ± 0.30 0.96 3.80 ± 0.18 0.04 3.05 ± 0.04 0.06 2.61 ± 0.09 0.11Deisopropyl-2-hydroxyatrazine (lg g�1) 0.05 ± 0.04 0.17 0.09 ± 0.05 0.14 0.03 ± 0.01 0.23 0.01 ± 0.03 0.21Deisopropylatrazine (lg g�1) 0.60 ± 0.14 0.11 0.80 ± 0.12 0.06 0.89 ± 0.02 0.06 0.84 ± 0.06 0.13Deethylatrazine (lg g�1) 0.07 ± 0.06 0.37 0.01 ± 0.01 0.52 n.d. 0.75 n.d. 0.55

FH Atrazine (lg g�1) 5.09 ± 0.15 0.86 4.30 ± 0.04 0.08 3.60 ± 0.04 0.12 3.31 ± 0.21 0.05Deisopropyl-2-hydroxyatrazine (lg g�1) 0.09 ± 0.05 0.15 0.07 ± 0.04 0.20 0.05 ± 0.03 0.27 0.05 ± 0.01 0.26Deisopropylatrazine (lg g�1) 0.69 ± 0.15 0.10 0.77 ± 0.10 0.09 0.92 ± 0.04 0.10 1.01 ± 0.05 0.08Deethylatrazine (lg g�1) 0.04 ± 0.03 0.33 n.d. 0.70 n.d. 0.92 n.d. 0.81

Mineralization and NER T14 T28 T42 T56

CH Cum. mineralization (lg g�1) 0.03 ± 0.00 0.08 ± 0.01 0.19 ± 0.01 0.34 ± 0.02NER (lg g�1) 1.6 ± 0.2 2.5 ± 0.1 3.3 ± 0.1 3.6 ± 0.1

FH Cum. mineralization (lg g�1) 0.05 ± 0.00 0.11 ± 0.01 0.26 ± 0.03 0.51 ± 0.05NER (lg g�1) 1.3 ± 0.1 2.1 ± 0.1 2.8 ± 0.0 3.0 ± 0.0

n.d. = not detectable.

y = -0.72x + 5.46R2 = 0.98

y = -0.66x + 5.18R2 = 0.98

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

atrazine (µg g-1)

non

extra

ctab

le re

sidu

es (µ

g g-1

)

CHFH

Fig. 3. Correlations between total atrazine residues and non-extractable residues(calculated in atrazine-equivalents) in a humic cambisol under constant (CH) andfluctuating soil humidity (FH). Bars indicate standard deviation of four samples.

y = -0.11x + 1.42R2 = 0.88

y = -0.10x + 1.24R2 = 1.0

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

atrazine (µg g-1)

deis

opro

pyla

trazi

ne (µ

g g-1

)

CHFH

y = -0.18x + 1.56R2 = 0.96

y = -0.14x + 1.25R2 = 0.77

0.5

0.6

0.7

0.8

0.9

1.0

1.1

0.0 1.0 2.0 3.0 4.0 5.0 6.0

atrazine (µg g-1)

deis

opro

pyla

trazi

ne (µ

g g-1

)

CHFH

(a)

(b)

Fig. 4. Correlations between atrazine and deisopropylatrazine residues in a humiccambisol under constant (CH) and fluctuating soil humidity (FH). Bars indicatestandard deviation of four samples: (a) for total atrazine and total deisopropylart-azine, and (b) for MeOH extractable atrazine and MeOH extractabledeisopropylatrazine.

A. Ngigi et al. / Chemosphere 84 (2011) 369–375 373

occurs preferably in the water phase, we analyzed the relationshipbetween the metabolites in PW and mineralized atrazine. Here wefound a significant correlation (p = 0.0173) between OH-DIA andmineralization (Fig. 5b). Dechlorination of DIA to the respectivehydroxylated compound OH-DIA was demonstrated by Behki andKahn (1986). OH-DIA is an important intermediate product in atra-zine mineralization since ring cleavage apparently occurs onlyafter hydroxylation (Kaufman and Kearney, 1970). This metaboliteis a polar compound and was detected almost exclusively in thePW fraction (Table 1). Therefore, the method of in situ PW extrac-tion gave a valuable insight into the degradation mechanisms ofatrazine, and we could show that higher atrazine mineralizationat FH was caused by higher levels of OH-DIA in comparison tothe control (CH).

From the detected correlations we suggest that the pathwayATRAZINE – DIA – OH-DIA is an important route in the breakdownof atrazine in soil humic cambisol. Although we could not elucidatethe complete pathway of atrazine degradation in this study, wecould identify the main intermediate products, which play animportant role in the mineralization of atrazine in this soil.

3.3.4. Relationship between bacterial cell counts and mineralizationThere were no significant differences in the cell counts for the

constant and fluctuating water regimes (p > 0.10). No direct rela-

tionship could be established between the cell counts and themineralization of atrazine. This shows that the mineralization ofatrazine is limited by the availability of atrazine and itsdegradation products in soil solution rather than by the bacterialactivity.

y = 0.51x - 0.31R2 = 0.76

0.0

0.1

0.2

0.3

0.4

0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1

deisopropylatrazine (µg g-1)

min

eral

ized

atra

zine

(µg

g-1)

y = 1.24x - 0.15R2 = 0.64

0.0

0.1

0.2

0.3

0.30.20.1

deisopropyl-2-hydroxyatrazine (µg g-1)

min

eral

ized

atra

zine

(µg

g-1)

(a)

(b)

Fig. 5. Correlations between deisopropylatrazine (DIA) residues, deisopropyl-2-hydroxyatrazine (OH-DIA) residues and mineralized atrazine in a humic cambisolunder constant and fluctuating soil humidity: (a) correlation between total DIA andmineralized atrazine. Bars indicate standard deviation of four samples respectively,and (b) Correlation between OH-DIA in soil pore water and mineralized atrazine.

374 A. Ngigi et al. / Chemosphere 84 (2011) 369–375

3.4. Influence of FH conditions on the bioavailability of atrazine andmetabolites

In the soil with FH conditions, higher residues of atrazine weredetected, indicating a higher bioavailability of atrazine under suchconditions. At the sampling day 14 (T14) a 10–20-fold higheratrazine concentration was detected in the PW fraction ascompared with the later sampling times (Table 1). This could bea hint that the in situ adsorption process was not at equilibriumafter 14 d and that atrazine adsorption proceeds very slowly undersuch conditions. Gamble et al. (2000) suggest that the sorption ofchemicals can be described by a two-step model: a rapid equilibra-tion at the surface of soil particles followed by a slow diffusion intothe soil particles where they are more strongly bound. Applyingthis theory to our results, this would mean that at least the secondstep was not accomplished after 14 d.

In literature it is reported that drying–rewetting cycles lead tohigher sorption, lower bioavailability, and lower biodegradationof organic chemicals caused by soil structure modifications, whichfavours pesticide adsorption and/or impedes desorption (Goetzet al., 1986; White et al., 1998; García-Valcárcel and Tadeo,1999; Lennartz and Louchart, 2007). These findings and explana-tions could not be confirmed by our results. Most likely one ofthe reasons for these apparent discrepancies between our resultsand results from the literature are the varying humidity conditionsin the soils of the various experiments. Some effects (e.g. shrinkingand swelling) mentioned in the literature were found when soilswere intensively dried, which was not the case in our study. More-over, the direct comparability of the various experiments and re-

sults is hampered because some authors expressed water contentin percentage and others in soil water potential. For a better com-parison we would suggest to use water potential data for describ-ing water availability in soils. This would enable a directcomparability of pesticide in situ bioavailability in different soils.Using an identical soil density will also help to harmonize experi-mental conditions e.g. when in situ sorption, desorption or turn-over of chemicals in soils is studied (Schroll et al., 2006).

In our study soil water potential and thus soil water content wasvaried during the drying–rewetting processes (Fig. 1). Diffusion ofsolutes (e.g. nutrients, pesticides) in soils is directly related to thecross section for flow (Papendick and Campbell, 1981) and thusthe lower water contents during the drying period most likely re-duce the diffusion of atrazine to sorption sites. Thus, moreatrazine is available for microbial processes leading to a higher deg-radation to intermediate products and finally to a higher minerali-zation in the soil with FH regime (Fig. 2). This hypothesis issupported by the studies of Gamble et al. (2000) and Belliveauet al. (2000). In laboratory experiments Gamble et al. (2000) dem-onstrated that water considerably influences the interactions of or-ganic chemicals with both the surfaces and the interiors of soilparticles. By using magnetic resonance imaging Belliveau et al.(2000) could affirm the connection between water uptake and atra-zine uptake in an organic rich soil. After wetting of an air-dried soil,some of the water penetrated only very slowly into the soil to thestrong binding sites. Parallel studies on atrazine adsorption re-vealed a little bit slower mobility of atrazine to the micropores asit was observed for water. This shows that in general drying andrewetting can impede the diffusion of water and pesticide to theadsorption sites resulting in a higher bioavailability of the organicchemicals.

4. Conclusion

The increased co-metabolic mineralization of atrazine in soilhumic cambisol under FH conditions was due to a higher bioavail-ability of atrazine and its metabolites under these conditions. Themethod of in situ sampling of soil PW turned out to be a valuabletool for elucidation of the degradation mechanisms of atrazine.With help of this tool deisopropyl-2-hydroxyatrazine was identi-fied as an important intermediate product in atrazine degradationand mineralization in this humic cambisol. Since slow atrazineadsorption was observed in this soil and since the increased atra-zine mineralization under FH regime was most likely due to thesorption behaviour, in situ long term sorption studies should beconducted for a better understanding of such complex interactionsin soils.

Acknowledgements

We thank the Helmholtz Zentrum München for providing thefacilities and consumables for this study. We are also grateful toDAAD for the scholarship granted to Anastasiah Ngigi that enabledher to be part of the team in this study.

References

Alletto, L., Coquet, Y., Benoit, P., Bergheaud, V., 2006. Effects of temperature andwater content on degradation of isoproturon in three soil profiles. Chemosphere64, 1053–1061.

Behki, R.M., Kahn, S.U., 1986. Degradation of atrazine by Pseudomonas: N-dealkylation and dehalogenation of atrazine and its metabolites. J. Agric. FoodChem. 34, 746–749.

Belliveau, S.M., Henselwood, T.L., Langford, C.H., 2000. Soil wetting processesstudied by magnetic resonance imaging: correlated study of contaminantuptake. Environ. Sci. Technol. 34, 2439–2445.

A. Ngigi et al. / Chemosphere 84 (2011) 369–375 375

Bosma, T.N.P., Middeldorp, P.J.M., Schraa, G., Zehnder, A.J.B., 1997. Mass transferlimitation of biotransformation: quantifying bioavailability. Environ. Sci.Technol. 31, 248–252.

Clay, S.A., Koskinen, W.C., 1990. Characterization of alachlor and atrazinedesorption from soils. Weed Sci. 38, 74–80.

Folberth, C., Scherb, H., Suhadolc, M., Munch, J.C., Schroll, R., 2009. In situ massdistribution quotient (iMDQ) – a new factor to compare bioavailability ofchemicals in soils? Chemosphere 75, 707–713.

Gamble, D.S., Bruccoleri, A.G., Lindsay, E., Langford, C.H., 2000. Chlorothalonil in aquartz sand soil: wetting effects on sorption capacity and bound residue.Environ. Sci. Technol. 34, 125–129.

Gan, J., Papiernik, S.K., Koskinen, W.C., Yates, S.R., 1999. Evaluation of acceleratedsolvent extraction (ASE) for analysis of pesticide residues in soil. Environ. Sci.Technol. 33, 3249–3253.

García-Valcárcel, A.I., Tadeo, J.L., 1999. Influence of soil moisture on sorption anddegradation of hexazinone and simazine in soil. J. Agric. Food Chem. 47, 3895–3900.

Goetz, A.J., Wehtje, G., Walker, R.H., Hajek, B., 1986. Soil solution and mobilitycharacterization of imazaquin. Weed Sci. 34, 788–793.

Han, O.S., New, P.B., 1994. Effect of water availability on degradation of 2,4-dichlorophenoxyacetic acid (2,4-D) by soil microorganisms. Soil Biol. Biochem.26, 1689–1697.

Ilstedt, U., Nordgren, A., Malmer, A., 2000. Optimum soil water for respiration beforeand after amendment with glucose in humid tropical acrisols and a boreal morlayer. Soil Biol. Biochem. 32, 1591–1599.

Jablonowski, N.D., Köppchen, S., Hofmann, D., Schäffer, A., Burauel, P., 2009.Persistence of 14C-labeled atrazine and its residues in a field lysimeter soil.Environ. Pollut. 157, 2126–2131.

Katayama, A., Bhula, R., Burns, G.R., Carazo, E., Felsot, A., Hamilton, D., Harris, C.,Kim, Y.-H., Kleter, G., Koedel, W., Linders, J., Peijnenburg, J.G.M.W., Sabljic, A.,

Stephenson, Racke, D.K., Rubin, B., Tanaka, K., Unsworth, J., Wauchope, R.D.,2010. Bioavailability of xenobiotics in the soil environment. Rev. Environ.Contam. T 203, 1–86.

Kaufman, D.D., Kearney, P.C., 1970. Microbial degradation of s-triazine herbicides.Residue Rev. 32, 235–265.

Krutz, L.J., Shaner, D.L., Weaver, M.A., Webb, R.M.T., Zablotowicz, R.M., Reddy, K.N.,Huang, Y., Thomson, S.J., 2010. Agronomic and environmental implications ofenhanced s-triazine degradation. Pest Manage. Sci. 66, 461–481.

Lennartz, B., Louchart, X., 2007. Effect of drying on the desorption of diuron andterbuthylazine from natural soils. Environ. Pollut. 146, 180–187.

Papendick, R.I., Campbell, G.S., 1981. Theory and measurement of water potential.In: Parr, J.F., Gardner, W.R., Elliot, L.F. (Eds.), Water Potential Relations in SoilMicrobiology. SSSA Special publication number 9. Soil Science Society ofAmerica, Madison, Wisconsin, pp. 1–22.

Premazzi, G., Stecchi, R., 1990. Evaluation of the impact of atrazine on the aquaticenvironment. Commission of the European Communities, Report EUR CD-NA-12569-EN-C. ECSC Brussels, Luxembourg. pp. 1–95.

Rüdel, H., Schmidt, S., Kördel, W., Klein, W., 1993. Degradation of pesticides in soil:comparison of laboratory experiments in a biometer system and outdoorlysimeter experiments. Sci. Total Environ. 132, 181–200.

Schroll, R., Kühn, S., 2004. Test system to establish mass balances for 14C-labeledsubstances in soil–plant–atmosphere systems under field conditions. Environ.Sci. Technol. 38, 1437–1544.

Schroll, R., Becher, H., Dörfler, U., Gayler, S., Grundmann, S., Hartmann, H.P., Ruoss, J.,2006. Quantifying the effect of soil moisture on the aerobic microbialmineralization of selected pesticides in different soils. Environ. Sci. Technol.40, 3305–3312.

White, J.C., Quiñones-Riviers, A., Alexander, M., 1998. Effect of wetting and dryingon the bioavailability of organic compounds sequestered in soil. Environ.Toxicol. Chem. 17, 2378–2382.