Arch. Environ. Contam. Toxicol. 40, 295–302 (2001) DOI: 10.1007/s002440010175

A R C H I V E S O F

Environmental Contamination a n d Toxicology © 2001 Springer-Verlag New York Inc.

Transformation Pathways of

14

C-Chlorothalonil in Tropical Soils

J. B. Regitano, V. L. Tornisielo, A. Lavorenti, R. S. Pacovsky Departamento de Ecotoxicologia, Centro de Energia Nuclear na Agricultura, Universidade de Sa˜o Paulo, Caixa Postal 96, CEP: 13400-970, Piracicaba, Sa˜o Paulo, Brazil

Received: 29 March 2000 /Accepted: 1 October 2000

Abstract. Chlorothalonil (CTN) is a chlorinated wide-spectrum fungicide, heavily and widely applied throughout the world. This study was undertaken to directly evaluate the rates and forms of 14C-labeled CTN dissipation in three acid Brazilian soils (Typic Humaquept [GH], Typic Quartzipsamment [AQ], and Typic Hapludox [LE]). Mineralization was not the major metabolic pathway of CTN-degrading microorganisms. However, CTN dissipation was fast in all soils and was mainly due to biodegradation (responsible for 50%, 54%, and 73% of 14 C-CTN dissipation in the GH, LE, and AQ soils, respectively), as well as to formation of soil-bound 14C residues (responsible for 46%, 34%, and 18% of 14C-CTN dissipation in the GH, LE, and AQ soils, respectively). Most soil-bound 14C residues were formed in the first day, but aging also contributed to the formation of less reversible forms of CTN-soil complexes. In these acid soils, the most abundant metabolite formed from CTN degradation was 3-carbamyl-2,4,5-trichlorobenzoic acid. A significant fraction of the CTN that had been assumed to be rapidly degradable in soils in previous reports has turned out to be soil-bound residues. Although bioavailability of any compound is reduced when soil complexes are formed, further research is needed to evaluate accumulation and availability of CTN soil-bound residues over long-term applications, and the consequent detrimental effects on the environment and on soil quality and fertility.

Chlorothalonil (tetrachloroisophthalonitrile, CTN) is a foliar, nonsystemic, and broad-spectrum chlorinated fungicide, highly efficient against pathogens that infect mainly vegetables, fruits, and ornamentals. The CTN reaction with sulfhydryl groups and gluthatione present in proteins or in cofactors in fungi has been proposed as the mechanism of fungicidal activity (Roberts and Hudson 1999). After 35 years of commercial use, CTN is still heavily and widely applied throughout the world. It is the second most widely used agricultural fungicide in the U.S.A., with 5 million kg applied annually (Cox 1997). Chlorinated pesticides are considered potential pollutants

Correspondence to: J. B. Regitano; email: [email protected]

due to their high application rate, their persistence, and their toxicity to humans and other species. CTN is not as persistent as the other organochlorine pesticides; therefore, manufacturers dislike when CTN is treated as an organochlorine pesticide. However, CTN is contaminated with the carcinogen hexachlorobenzene and has been classified as a “probable human carcinogen” by the U.S. Environmental Protection Agency (Cox 1997). CTN is extremely toxic to fish in acute lethal doses (Davies and White 1985), although it undergoes rapid metabolism via the glutathione pathway to give polar, water-soluble metabolites (Davies 1988; Davies and White 1985). CTN residues have been commonly found in a large variety of vegetables (Andersson and Bergh 1991; Luke et al. 1988; ValverdeGarcia et al. 1993); it can leach through sandy soils, reaching groundwater concentrations as high as 272 ␮g L⫺1 (Krawchuk et al. 1987). CTN has been found in groundwater of four U.S. states (Cox 1997). Unfortunately, water monitoring studies have not been widely performed in Brazil. In addition, the compound 4-hydroxy-2,5,6-trichloroisophthalonitrile (Mt-1, probably the primary metabolite resulting from CTN breakdown) is often found in soils, plants, and animals (Figure 1). It is 30 times more acutely toxic than CTN itself, and it is more persistent and mobile in soil (Cox 1997). The residual toxicity of this compound may be responsible for the suppression of CTN degradation in soils after repeated applications (Motonaga et al. 1998). Most reports in the literature have shown that CTN is rapidly degraded by soil microorganisms and does not accumulate in ecosystems. The half-life of CTN degradation in soil after its first application ranges from 5 to 36 days (Katayama et al. 1991b; Motonaga et al. 1998; Sato and Tanaka 1987; Sun et al. 1985; Takagi et al. 1991; Walker et al. 1988; van der Pas et al. 1999). However, these reports considered CTN dissipation to correspond with degradation, which was calculated as the difference between the amount applied and that extracted by a relative mild extractor. The use of the term degradation in such way is too broad because it includes other forms of dissipation, such as formation of soil-bound residues, volatilization, uptake, etc. Therefore, for the purpose of our research, the term degradation is used to refer only to the chemical and/or biological transformation of the extractable (“available”) CTN. Although mineralization, the complete breakdown of CTN to CO2, is part of the degradation process, it will be discussed independently.

296

J. B. Regitano et al.

pathways of CTN in three Brazilian soils that have a wide variety of soil physical-chemical properties. Tropical soils are usually more acidic than temperate soils, and it may affect the metabolic pathway of CTN. This study was performed using radiolabeled 14C-CTN, which permitted us to directly account for the different processes of transformation, such as mineralization, degradation, and formation of soil-bound residues. In addition, CTN sorption coefficients and soil microbial activities were supplied to support our findings.

Materials and Methods The experiment used a randomized 3 ⫻ 7 factorial design, involving 3 soils and 7 periods of incubation. Three replicates were used for each of the 21 treatments.

Chemicals Uniformly ring-labeled 14C-CTN (specific activity ⫽ 0.20 MBq mmol⫺1 and purity ⬎ 98%), analytical-grade CTN (purity ⬎ 95%), and major metabolites 4-hydroxy-2,5,6-trichloroisophthalonitrile (Mt-1, Figure 1), 3-cyano-2,4, 5,6-tetrachlorobenzamide (Mt-2, Figure 1), and 3-carbamyl-2,4,5-trichlorobenzoic acid (Mt-3, Figure 1) were provided by ISK Biosciences (Mentor Ohio, part of Zeneca Inc.). Analytical-grade solvents (toluene, methanol, and acetone; from Mallinckrodt, St. Louis, MO) suitable for pesticide residue analysis were employed, as well as S¢int-A XF (Packard, Meriden, CT) liquid scintillation reagent.

Soils Fig. 1. Transformation pathways of chlorothalonil (CTN) in soils. Scheme adapted from Roberts and Hutson (1999)

Moreover, the term dissipation is used here to address other transformation pathways besides degradation. The previous reports, cited above, have not addressed formation of soil-bound residues. However, data about reduced pesticide mobility or biocidal activity due to their binding to humic substances have been commonly found in the literature (Calderbank 1989; Kloskowski and Fu¨hr 1987). Moreover, the presence of a pesticide in a reversible or slowly reversible form bound to a soil fraction may raise its persistence by hindering transport and reducing availability (Chung and Alexander 1998). Therefore, formation of soil-bound residues is a very important piece of information. Rouchaud et al. (1988), using harsh extraction, found that 37% or 26% of the applied CTN were bound to a loamy sand soil under either broccoli or Chinese cabbage crops, respectively, 2 months after application. Gamble et al. (2000), using online HPLC microextraction, found that 30.5% of the applied CTN was bound to a quartz sand soil 18 days after application. None of these authors used 14 C-labeled material, the best technique to directly measure formation of soil-bound residues. Information about pesticide dissipation with time is essential in monitoring environmental risks associated with pesticide use. Therefore, our purpose was to evaluate the transformation

Samples of three soils (Typic Humaquept [GH], Typic Quartzipsamment [AQ], and Typic Hapludox ([LE]) were collected from the plough layer (0 –10 cm) of farming areas in Sa˜o Paulo State, Brazil. They were selected to present a wide variety of soil physical-chemical properties (Table 1).

Sorption Aliquots (5 ml) of 14C-CTN solutions at five different concentrations (methanol-suspended 14C-CTN stock solution was used to obtain CTN solutions at concentrations of 0.05, 0.09, 0.20, 0.38, and 0.76 ␮g ml⫺1, diluted in 0.005 M CaCl2 with 0.2% of the solvent carrier) were added to 1 g of air-dried soils passed through a 2-mm sieve. Soil-slurries were shaken (8 h, 25 ⫾ 2°C), centrifuged (12,100 g, 10 min), and aliquots of 0.5 ml of the supernatants were removed to measure the solution concentration of CTN at equilibrium (Ce). Equilibration time and soil to solution ratio were established based on a pretest performed at the highest 14C-CTN concentration (0.76 ␮g ml⫺1). The amount of CTN sorbed (S) was calculated by the difference between its initial concentration (Ci) and equilibrium solution concentration (Ce). The Freundlich sorption coefficients (Kf) were estimated by fitting the results to the equation (1): S ⫽ K f C eN

(1)

where N is the exponential fit. The partition coefficients (Kd) were estimated by assuming that equation (1) is linear (N ⫽ 1), whereas Koc values were estimated by normalizing Kd to the soil organic carbon content (OC ⫽ OM/1.72).

14

C-Chlorothalonil in Tropical Soils

297

Table 1. Data about soils’ physical-chemical properties, sorption, and microbial activity Soils Parameters

GHa

LE

AQ

b

Soil physical-chemical properties Clay (g kg⫺1) 550 57 Organic matter (g kg⫺1) 4.8 pH-H2Oa pH-CaCl2 (0.01 mol L⫺1) 4.0 227 CEC (mmolc kg⫺1) Sorptionc Kf 55.2 N 1.08 74.7 Kd (L kg⫺1) 2,254 Koc (L kg⫺1) Microbial activityd ␮a (␮mol g⫺1 h⫺1) 0.016

830 28 4.5 3.5 147

20 13 4.8 3.9 25

28.4 0.97 25.4 1,560

1.8 0.99 1.9 251

0.009

0.003

GH ⫽ Typic Humaquept, LE ⫽ Typic Hapludox, AQ ⫽ Typic Quartzipsamment. b Soil properties determined according van Raij et al. (1987) and EMBRAPA (1997), CEC ⫽ cation exchange capacity. c Kf ⫽ sorption coefficient, N ⫽ exponential fit, Koc ⫽ Kd normalized to OC, OC ⫽ OM/1.72. d ␮a ⫽ soil microbial activity. a

Microbial Activity At time zero, samples (1 g, dry weight) of freshly moistened soil were placed into biometer flasks and were treated with 0.3 ml of a 2 mM 14 C-glucose solution. Soils were then allowed to incubate for 60 min. The 14C-CO2 evolved from glucose degradation was trapped in 0.25 ml of monoethanolamine, which was further analyzed by liquid scintillation (LS) counter (Packard TR-1600, Meriden, CT). The amount of 14 C-glucose degraded, in terms of ␮mol g⫺1 h⫺1, was regarded as directly proportional to soil microbial activity (Freitas et al. 1979).

Extraction and Formation of Soil-Bound Residues At 0, 1, 7, 14, 28, 56, and 90 days of incubation, extraction of soils containing the original CTN and its metabolites was performed. Soil samples (10 g) were extracted twice with 40 ml acetone/1.0 M HCl solution (4:1 v/v) for 15 min on a rotatory shaker. Afterward, soil slurries were centrifuged at 12,100 g for 10 min, and 0.5-ml aliquots of the liquid supernatants were taken for LS analyses (ISK Biosciences, unpublished methods). After extraction, soil-bound residues (14C) were measured by combusting 0.4 g (dry weight) of the extracted soils in a Biological Oxidizer (Harvey Instruments, OX500).

Degradation The remaining soil extract was vacuum evaporated (35°C) to remove acetone, and the pH was further adjusted with 3 ml of 5 M H2SO4. To partition CTN and its metabolites into the organic phase, 30% NaCl and 50 ml diethyl ether were added together in a separatory funnel, which was manually shaken for 2 min. The settled aqueous phase was drained and partitioned into flesh diethyl ether again. The ether phases were then combined and evaporated to dryness. Residues were suspended in 3 ml of acetone, and a 0.1 ml aliquot was taken for LS analysis. The efficiency of the partition method was evaluated by checking the radioactivity present in the aqueous phase. Then, a 0.5 ml aliquot of the aqueous phase was mixed with 1 ml distilled water and was added to 15 ml of scintillation solution for LS analysis (ISK Biosciences, unpublished methods). For metabolite identification, a 0.1 ml aliquot of the acetone-suspended extract was applied to silica gel plates (60 F254, Merck, Steinheim, Germany) and eluted in toluene/methanol/acetone (2:1:1 v/v/v). The Rf values for 14C-residues were evaluated in an Automatic TLC-Linear Analyzer (Berthold GmbH & Co., Berlin, Germany). Putative identifications were made based on the relative mobility of the pure metabolite standards. It is important to point out that the TLC technique does not provide a definitive identification of the metabolites, but with compounds undergoing oxidative degradation, the Rf values are very indicative of the presumed metabolite.

Results and Discussion Mineralization Sorption After collection, soil samples were passed through a 2 mm sieve and stored moistened at laboratory conditions (semi-dark room, 25 ⫾ 2°C) for about 1 month to permit degradation of any fresh crop residues in the soil. In order to perform incubation, an aliquot of 1 ml of nonlabeled 1.53 mg ml⫺1 CTN solution in methanol was added to 10 g subsamples of each soil, which were previously dried and pulverized in a porcelain grinder. Further, another aliquot of 1 ml of 0.63 mg ml⫺1 14 C-CTN solution (0.47 MBq ml⫺1, in methanol) was added to the same soil subsamples. After homogenization and solvent evaporation, these soil subsamples were mixed (using a domestic hand-mixer) with original soils to reach a dry weight of 1 kg, which would correspond to an application rate of 2.16 mg kg⫺1 CTN. Following mixing, 50 g (dry weight) of each soil were placed into 250 ml biometer flasks and were moistened to 75% field capacity. Flasks were incubated for up to 90 days in a semi-dark room with constant temperature (25 ⫾ 2°C). The amount of 14C-CO2 evolved from CTN mineralization was trapped in 10 ml of 0.2 N NaOH. At preestablished time intervals (0, 7, 14, 21, 28, 42, 56, 74, and 90 days), 1 ml aliquots of the NaOH solution were withdrawn, mixed with scintillation solution (15 ml), and analyzed by LS.

Sorption varied from moderate to high among the studied soils (Kf, Table 1). The GH soil, having higher soil organic carbon content (OC), presented very high sorption, whereas the AQ soil, having lower OC and clay contents, presented moderate sorption (Table 1). Motanaga et al. (1998) also observed very high sorption (Kd ⫽ 79.4 L kg⫺1) for a soil with high organic matter (OM) content (57 g kg⫺1). There was a strong relationship between OC and Kf values for data from this work and also from Kawamoto and Urano (1989) and Motonaga et al. (1998) (Figure 2). For all data sets, the greater the OC, the higher the CTN sorption. Kawamoto and Urano (1989) concluded that OM was the main soil fraction affecting sorption of organochlorine pesticides in their soils. Kawamoto and Urano (1989) found that it was possible to roughly predict Koc values for organochlorine pesticides knowing their octanol-water partition coefficients (Pow), using the following equation: log Koc ⫽ 0.638 log Pow ⫹ 1.137. For CTN (Pow ⫽ 437), the predicted Koc was equal to 663 L kg⫺1. This

298

Fig. 2. Chlorothalonil (CTN) sorption plotted as a function of OC contents. Data from Kawamoto and Urano (1989), and Motanaga et al. (1998) were used

value differed somewhat from ours (Table 1). The higher Koc values for the GH and LE soils suggested that the clay content played a secondary but important role in total CTN sorbed.

Mineralization The rate of CTN mineralization was slow in all studied soils. After 90 days, only 13.8%, 5.7% and 2.8% of the CTN applied was evolved as 14C-CO2 in the GH, LE, and AQ soils, respectively (Figure 3). In agreement, Rouchaud et al. (1988) found that only 10% to 12% of the soil applied CTN disappeared, most likely as CO2, 2-months after treatment. Katayama et al. (1997) suggested no mineralization of CTN occurred in pure cultures using 37 strains of bacteria. These results indicated that complete mineralization is not the major metabolic pathway of CTN-degrading microorganisms. If CTN were extensively mineralized, then about 4 mol of chloride anion should be released per 1 mol of CTN. However, between 0.75 to 1 mol of chloride anion was actually released after complete degradation of 1 mol of CTN (Motonaga et al. 1996, 1998). Based on these results, the authors suggested that one possible degradation pathway of CTN in soil is the nucleophilic substitution of 4-chlorine atom with hydroxide anion, forming Mt 1 (Figure 1). However, CTN mineralization was faster in the GH soil with higher OC and higher microbial activity and slower in the AQ soil with lower OC and lower microbial activity (Table 1). The greater CTN mineralization in the GH soil was likely due to the greater activity of CTN-degrading microorganisms in those soils with higher OM content. Soil OM provides a source of nutrients and/or surfaces for bacterial attachment (Takagi et al. 1991; Walker et al. 1988), and therefore usually sustains a greater number of microorganisms. The literature has strongly suggested that the major mechanism of CTN degradation in soil is biological and that aerobic conditions are the most suitable for degradation to take place

J. B. Regitano et al.

Fig. 3. Chlorothalonil (CTN) mineralization in Brazilian soils (GH, LE, and AQ) plotted as a function of the incubation time

(Katayama et al. 1991a, 1997; Mori et al. 1996; Motonaga et al. 1996; Sato and Tanaka 1987; Walker et al. 1988). It is known that most bacterial strains isolated from soil were capable of degrading CTN and that they usually did not degrade CTN in the absence of other carbon sources (Katayama et al. 1991a; Mori et al. 1996; Sato and Tanaka 1987). In addition, CTN degradation was not linked to microbial population growth (Mori et al. 1996; Motonaga et al. 1996). All these studies showed evidence that CTN was degraded mainly by cometabolism.

Extraction and Formation of Soil-Bound Residues The percentage of soil-extracted 14C residues tended to decrease with time (Figure 4), whereas the percentage of soilbound 14C residues tended to increase with time (Figure 5). The changes in either soil-extracted or soil-bound 14C residues were most abrupt in the first day, but there were still significant changes up to 14 days of incubation (Figures 4 and 5). Afterward, the curves reached a plateau, which corresponded to equilibrium, where about 46%, 57%, or 71% of the 14C applied had been extracted and at about 46%, 34%, or 18% of 14C applied had been bound to the GH, LE, or AQ soils, respectively. The high OM content appeared to be controlling the lower extractability of 14C residues and the higher formation of soil-bound 14C residues in the GH soil, with the opposite trend for the AQ soil. Mineralization, therefore, accounted for (on average) no more than 10%. Gamble et al. (2000) found that 30.5% of the applied CTN was bound to a quartz sand soil similar to our AQ soil 18 days after application. This value was significantly higher than ours, and it was probably because they have used a milder extraction procedure. Recent studies by Alexander and co-workers have shown that mild extraction procedures reflect better the real bioavailability of organic molecules (Chung and Alexander 1998; Tang and Alexander 1999). Therefore, further research should be carried out to clear CTN bioavailability in soils.

14

C-Chlorothalonil in Tropical Soils

Fig. 4. Soil-extracted 14C residues resulting from chlorothalonil (CTN) application in Brazilian soils (GH, LE, and AQ) plotted as a function of the incubation time

Fig. 5. Soil-bound 14C residues resulting from chlorothalonil (CTN) application in Brazilian soils (GH, LE, and AQ) plotted as a function of the incubation time

It is likely that soil-bound 14C residues resulting from CTN application were formed in two steps: a fast phase (approximately 24 h) followed by a slow phase (Figure 5). A diagnostic test, based on Crank’s solution for Fick’s Law differential equation (ÀD ⫽ A tZ, where ÀD ⫽ refers to the amount of pesticide diffused into particle interiors, t ⫽ refers to time), can be used to determine whether the bound residue measurements are consistent with diffusion (Gamble et al. 2000). If so, Z should have the value of 1/2 and ÀD should be linearly related to the square root of incubation time (t1/2) (Khan 1973; Kookana et al. 1992; Leenher and Ahlrichs 1971). The initial nonlinearity (Figure 6) suggested that only at longer times was the reaction process limited by diffusion (Sparks 1989). However, those graphs indicated that the critical period for obser-

299

Fig. 6. Soil-bound 14C residues resulting from chlorothalonil (CTN) application in Brazilian soils (GH, LE, and AQ) plotted as a function of the square root of incubation time

vation was in the first 7 days. Unfortunately, our observations were not performed within the first week, which will require a more rigorous study, including interpretation of the diffusion process. If we assume linearity between ÀD and t1/2 only in order to speculate, we would have slope values that could indicate the binding rate of CTN during the slow phase, while the yintercept values could indicate the amount of CTN bound during the fast phase. For the GH soil, ÀD ⫽ 39.9 ⫹ 0.75 t1/2 (R2 ⫽ 0.61*), while for the LE soil, ÀD ⫽ 25.0 ⫹ 1.38 t1/2 (R2 ⫽ 0.74**), and for the AQ soil, ÀD ⫽ 6.4 ⫹ 1.71 t1/2 (R2 ⫽ 0.75**). Therefore, approximately 40%, 25%, and 6% of 14C applied could have been bound rapidly to the GH, LE, and AQ soils, respectively. On the other hand, the higher slope (1.71) for the AQ soil might indicate that aging effects were most pronounced in this soil. Consequently, higher amounts of soilbound residues might have been formed in AQ soil during a slow-phase reaction than in either LE (slope ⫽ 1.38) or GH (slope ⫽ 0.75) soils. It is possible that fewer retention sites were available for rapid binding of the residues during the fast-phase reaction in soils with low OM contents. The formation of soil-bound residues has shown time dependence and has often been limited by diffusion through interstitial micropores within soil aggregates and/or through the threedimensional matrix of soil OM (Brusseau and Rao 1989; Pignatello 1990a, 1990b; Pignatello and Huang 1991; Steinberg et al. 1987). Nam and Alexander (1998) showed that sequestration and reduced bioavailability of nonpolar organic molecules occured when hydrophobic compounds entered into nanopores having hydrophobic surfaces. In addition, Gamble et al. (2000) showed theoretical evidence that the kinetics of CTN

**p ⬍ 0.01 *p ⬍ 0.05

300

J. B. Regitano et al.

mass transfer among three different states was consistent with intraparticle diffusion. On the other side, the release of chloride ions constituted direct evidence that there was formation of covalent bonds between chlorophenols and humic acids (Dec and Bollag 1994). According to these authors, the formation of covalent bonds involved oxidative coupling and was mediated by microbial enzymes. The facts that CTN did not dissipate in autoclaved soils (Katayama et al. 1991b, 1997; Mori et al. 1996; Motonaga et al. 1996; Sato and Tanaka 1987; Walker et al. 1988) and that common enzymes ubiquitous in bacteria were probably involved in CTN degradation (Katayama et al. 1997), reinforce the point that soil microbial activity was required to form CTN soil-bound residues and that covalent bounds should be involved.

Degradation On average, total recovery represented 96.4 ⫾ 4.6% of the applied 14C-CTN. At time zero, extractable recovery ranged from 95% to 98%. About 100% of this radioactivity corresponded to the original CTN in the LE and AQ soils, but metabolite separation using our TLC technique did not succeed for the GH soil. Therefore, metabolites resulting from CTN breakdown in this soil were not further evaluated. In the other two soils, the percentage of Mt-3 increased, whereas the percentage of soil-extracted CTN continuously decreased during the overall incubation (Figure 7). Mt-3 was the most abundant product formed from CTN breakdown in both soils. After 56 days, Mt-3 corresponded to 18% and 25% of the applied CTN in the LE and AQ soils, respectively. Mt-1 and Mt-2 (Figure 1) were also found in significant amounts up to 14 days, but they represented less than 10% of the total applied CTN. Mt-2 was the first metabolite to accumulate, and its presence was observed immediately after the first day of incubation (Figure 7). Katayama et al. (1997) demonstrated that bacteria of various taxonomic groups were capable of degrading CTN in pure culture. Although removal of chlorine atoms from aromatic rings has often been carried out by specific enzymes or specific microorganisms limited in diversity in an environment, Katayama et al. (1997) suggested that common enzymes ubiquitous in bacteria were involved in the degradation of CTN by the substitution of chlorine atoms with hydroxyl or methylthio groups. Roberts and Hutson (1999) reported two major metabolic pathways of CTN breakdown. One involved the displacement of one chlorine atom by a hydroxyl group to generate Mt-1 or via reductive dechlorination (Figure 1). The second major metabolic pathway involved the oxidation/hydration of one cyano group to a corresponding amide and organic acid to generate Mt-2 and Mt-3, subsequently (Figure 1). In this case, our data supported the second mechanism as the preferential metabolic pathway for CTN degradation in acid soils, such as these. It is important to point out that we do not believe that reductive dechlorination is important metabolic pathway for CTN degradation. The hydrolytic nature of CTN degradation was already demonstrated by Rouchaud et al. (1998). In spite of this, Sato and Tanaka (1987) observed that the CTN dissipation rate increased with an increase in water content of up to 60% water holding capacity (WHC). However, at 100% WHC, where anaerobic conditions dominate, the dissipation was very slow (Sato and Tanaka 1987).

Fig. 7. Chlorothalonil (CTN) degradation and its resulting metabolites in Brazilian soils (A ⫽ GH soil, B ⫽ LE soil, and C ⫽ AQ soil) plotted as a function of the incubation time

Results similar to ours were obtained by the Pesticide Safety Directorate (1994). However, Rouchaud et al. (1988) were the first ones to report the formation of 1,3-dicarbomoyl-2,4,5,6-tetrachlorobenzene, the intermediate compound between Mt-2 and Mt-3 (Figure 1), as the major ‘available‘ compound resulting from soil CTN hydrolysis. Two months after soil treatment, this intermediate compound represented 22% of the applied CTN, whereas Mt-1 represented 37% (of which, approximately 80% was soil-bound). Therefore, a significant portion of Mt-1 may not have been identified in our research because it may have been soil-bound. Motonaga et al. (1996, 1998) noticed that 10 –15% of the applied CTN corresponded to Mt-1 4 weeks after application and that Mt-1 was the major metabolite resulting from CTN breakdown. Sato and Tanaka (1987) also suggested Mt-1 as the major metabolite resulting from CTN breakdown. However, the analytical procedure adopted by Sato and Tanaka (extraction with acetone) and Motonaga and co-workers (extraction with aqueous acetone, 87%) may not have been adequate to extract polar compounds, such as the amides and acidic degradation products, that had been converted from the CN groups. These extraction procedures may obscure the formation of such com-

14

C-Chlorothalonil in Tropical Soils

301

Table 2. Relative contribution of microbial degradation (MD) and formation of soil-bound residues (BR) to CTN dissipation Soils GHa Incubation (days) 0 1 7 14 28 56 90 a

LE

MD (% of 0 49 48 46 47 48 52

BR 14

MD (% of

C-applied) 4 39 42 44 46 47 45

AQ BR 14

(% of

C-applied)

0 23 45 48 50 56 54

MD

4 24 29 32 36 33 38

0 24 38 57 76 72 78

BR 14

C-applied) 1 5 14 14 14 24 20

GH ⫽ Typic Humaquept, LE ⫽ Typic Hapludox, AQ ⫽ Typic Quartzipsamment.

pounds and may lead to erroneous interpretations of the results. On the other side, their soils were less acidic than ours (pH-H2O ⬎ 5.5), and it favors substitution of a chlorine atom by an hydroxyl group (Rouchad et al. 1988). Mori et al. (1996) observed that neutral pH favors CTN-degradation, but unfortunately they did not determine which pathway was stimulated by raising the pH. Mt-1 is of great environmental concern, as it has higher persistence, availability, and toxicity than CTN itself (Cox 1997). van der Pas (1999) observed that Mt-1 was detected in upper groundwater, at an average concentration of 0.1 to 0.2 ␮g dm⫺3. Moreover, Mt-1 tended to accumulate in soil after repeated application of high rates of CTN (40 mg kg⫺1), and it may suppress CTN degradation (Motonaga et al. 1998).

Dissipation CTN dissipation was very fast in all studied soils (Figure 7). After 24 h of incubation, approximately 88%, 47%, or 29% of the applied CTN was already dissipated in the GH, LE, or AQ soils, respectively. CTN was still dissipating at considerable rates up to 14 days of incubation in the LE and AQ soils. Using the half-life (t1/2) values calculated (assuming first-order kinetics), it was observed that CTN dissipation was faster in the GH soil (t1/2 ⫽ 0.4 days, R2 ⫽ 0.96**) than in the LE soil (t1/2 ⫽ 7.1 days, R2 ⫽ 0.84**), and the slowest dissipation occurred in the AQ soil (t1/2 ⫽ 13.1 days, R2 ⫽ 0.97**). Except for the GH soil, similar values of half-life for CTN were found in the literature (Katayama et al. 1991b; Motonaga et al. 1998; Sato and Tanaka 1987; Sun et al. 1985; Takagi et al. 1991; Walker et al. 1988; van der Pas et al. 1999). Neither chemical hydrolysis, nor photolysis, nor volatilization are important pathways for CTN dissipation in soils (Roberts and Hutson 1999). Therefore, based on what was discussed before, the overall fast dissipation of CTN in these soils would be attributed mainly to microbial degradation and to formation of less reversible soil-bound residues. Table 2 shows the relative contribution of these both factors to the total dissipation of CTN. The much faster dissipation rate of CTN in the GH soil than in the LE or AQ soil was primarily due to its much higher OC, which contributed to the

**p ⬍ 0.01

higher sorption, to the faster formation of soil-bound residues, and to the greater microbial degradation initially (Table 2). However, over the long term, organic compounds are usually less available for microbial degradation in the sorbed state, and therefore they are more slowly biodegraded (Kawamoto and Urano 1989; Steen et al. 1978). It may explain the relatively constant contribution of microbial degradation to CTN dissipation in the GH soil and its increased contribution to CTN dissipation in the AQ soil with time (Table 2). Katayama et al. (1995) observed accelerated dissipation of CTN in a soil amended with high rates of farmyard manure, mainly due to formation of soil-bound residues with the humus fraction. Later, in a more detailed study, Mori et al. (1998) found that the enhancement of CTN dissipation in soils that received farmyard manure was also due to the increase in fungal degrading capacity, although the major CTN degraders in soil were bacteria. Therefore, a great part of the CTN that was assumed to be rapidly degraded in soils in previous studies turned out to be soil-bound residues. In general, the formation of soil-bound residues reduces the compound’s bioavailability significantly, and therefore toxicological effects caused by CTN soil-bound residues are not expected. However, more research regarding CTN soilbound residues is needed to evaluate its accumulation and bioavailability over long-term applications and consequently the detrimental effects on soil organisms and on soil quality and fertility. It would be interesting to evaluate if the extraction procedure adopted here would reflect CTN bioavailability.

Conclusions 1. CTN dissipation was fast in all soils and was mainly due to soil binding and microbial degradation. 2. Most of the soil-bound residues resulting from CTN application were rapidly formed in the first day of incubation, but aging also contributed to the formation of soil-bound residues, mainly in the coarse-textured AQ soil. 3. The 3-carbamyl-2,4,5-trichlorobenzoic acid was the most abundant metabolite formed from CTN breakdown in Brazilian soils.

Acknowledgments. The authors express their gratitude to Conselho Nacional de Desenvolvimento Cientifico e Tecnológico for the partial support (CNPq, Processo 301160/93-7) and to Marianne Bischoff for her assistance with the literature.

302

References Andersson A, Bergh T (1991) Pesticide residues in fresh fruit and vegetables on the Swedish market, January 1985–December 1989. Fres J Anal Chem 339:387–389 Brusseau ML, Rao PSC (1989) The influence of sorbate-organic matter interactions on sorption nonequilibrium Chemosphere 18: 1691–1706 Calderbank A (1989) The occurrence and significance of bound residues in soil. Rev Environ Contain Toxicol 108:72–103 Chung N, Alexander M (1998) Differences in sequestration and bioavailability of organic compounds aged in dissimilar soils. Environ Sci Technol 32:855– 860 Cox C (1997) Chlorothalonil. J Pest Reform 17:14 –20 Davies PE (1988) Disappearance rates of chlorothalonil (TCIN) in the aquatic environment. Bull Environ Contam Toxicol 40:405– 409 Davies PE, White RWG (1985) The toxicology and metabolism of chlorothalonil in fish. I. Lethal levels for Salmo gairdneri, Galaxias maculatus, G. truttaceous and G. auratus and the fate of 14 C-TCIN in S. gairdneri. Aquatic Toxicol 7:93–105 Dec J, Bollag JM (1994) Dehalogenation of chlorinated phenols during oxidative coupling. Environ Sci Technol 28:484 – 490 Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA) (1997) Manual de métodos de análise de solo, 2nd ed. Embrapa, Rio de Janeiro, CNPS, 212 p Freitas JR, Vose PB, Nascimento Filho V, Ruschel AP (1979) Estimativa da atividade da microflora heterotrófica em solo Terra Roxa Estruturada usando respirometria com glicose-14C. Energ Nucl Agric 1:123–130 Gamble DS, Bruccoleri AG, Lindsay E, Langford CH, Leys GA (2000) Chlorothalonil in a quartz sand soil: speciation and kinetics. Environ Sci Technol 34:120 –124 Katayama A, Isemura H, Kuwatsuka S (1991a) Population change and characteristics of chlorothlonil degrading bacteria in soil. J Pestic Sci 16:239 –245 Katayama A, Isemura H, Kuwatsuka S (1991b) Suppression of chlorothalonil dissipation in soil by repeated applications. J Pest Sci 16:233–238 Katayama A, Mori T, Kuwatsuka S (1995) Abiotic dissipation of chlorothalonil in soil accelerated by amendment with high applications of farmyard manure. Soil Biol Biochem 27: 147– 151 Katayama A, Itou T, Ukai T (1997) Ubiquitous capability to substitute chlorine atoms of chlorothalonil in bacteria. J Pest Sci 22:12–16 Kawamoto K, Urano K (1989) Parameters for predicting fate of organochlorine pesticides in the environment: (II) Adsorption constant to soil. Chemosphere 19:1223–1231 Khan SU (1973) Equilibrium and kinetic studies of the adsorption of 2,4-D and picloram on humic acid. Can J Soil Sci 53:429 – 433 Kloskowski R, Fu¨hr F (1987) Aged and bound herbicide residues in soil and their bioavailabiliyy. Part 2: Uptake of aged and nonextractable (bound) [carbonyl-14C]methabenzthiazuron residues by maize. J Environ Sci Health B22:623– 642 Kookana RS, Aylmore LAG, Gerritse RG (1992) Time dependent sorption of pesticides during transport in soils. Soil Sci 154:214 –225 Krawchuk BP, Webster GR, Barrie GR (1987) Movement of pesticides to groundwater in an irrigated soil. Water Pollut Res J Canada 22:129 –145 Leenher JA, Ahlrichs JL (1971) A kinetic and equilibrium study of the adsorption of carbaryl and parathion upon soil organic matter surfaces. Soil Sci Soc Am Proc 35:700 –704 Luke MA, Masumoto HT, Cairms T, Hundley HK (1988) Levels and incidences of pesticide residues in various foods and animal feeds analyzed by the Luke multiresidue methodology for fiscal years 1982–1986. J Assoc Off Anal Chem 71:415– 433 Mori T, Fujie K, Kuwatsuka S, Katayama A (1996) Accelerated

J. B. Regitano et al.

microbial degradation of chlorothalonil in soils amended with farmyard manure. Soil Sci Plant Nutr 42:315–322 Mori. T, Fujie K, Katayama A (1998) Bacterial and fungal contributions to chlorothalonil degradation in soil. Soil Sci Plant Nutr 44:297–304 Motonaga K, Takagi K, Matumoto S (1996) Biodegradation of chlorothalonil in soil after suppression of degradation. Biol Fert Soils 23:340 –345 Motonaga K, Takagi K, Matumoto S (1998) Suppression of chlorothalonil degradation in soil after repeated application. Environ Tox Chem 17:1469 –1472 Nam K, Alexander M (1998) Role of nanoporosity and hydrophobicity in sequestration and bioavailability: tests with model solids. Environ Sci Technol 32:71–74 Pignatello JJ (1990a) Slowly reversible sorption of aliphatic halocarbons in soils. I. Formation of residual fractions. Environ Tox Chem 9:1107–1115 Pignatello JJ (1990b) Slowly reversible sorption of aliphatic halocarbons in soils. II. Mechanistic aspects. Environ Tox Chem 9:1116 – 1126 Pignatello JJ, Huang LQ (1991) Sorptive reversibility of atrazine and methalachlor residues in field soil samples. J Environ Qual 20: 222–228 PSD (1994) Pesticide safety directorate (UK MAFF), evaluation no. 112 Roberts T, Hutson D (1999) Metabolic pathways of agrochemicals. Part 2: Insecticides and fungicides. The Royal Society of Chemistry, Cambridge, United Kingdom, pp 1380 –1384 Rouchaud J, Roucourt P, Vanachter A (1988) Hydrolytic biodegradation of chlorothalonil in the soil and in cabbage crops. Toxicol Environ Chem 17:59 – 68 Sato K, Tanaka H (1987) Degradation and metabolism of a fungicide, 2,4,5,6-tetrachloro-isophthalonitrile (TPN) in soil. Biol Fertil Soils 3:205–209 Sparks DL (1989) Kinetics of soil chemical processes. Academic Press, London, pp 128 –145 Steen WC, Paris DF, Baughman GL (1978) Partitioning of selected polychlorinated biphenyls to natural sediments. Water Res 12: 655– 657 Steinberg SM, Pignatello JJ, Sawhney BL (1987) Persistence of 1,2dibromoethane in soils: Entrepment in intraparticle micropores. Environ Sci Tecnol 21:1201–1208 Sun T, Hashimoto T, Wada H, Takai Y (1985) Effects of organic amendments on soil microbial ecosystem modified by long-term application of pesticides. Nippon Dojohiryogaku Zasshi 56:31–36 (in Japanese) Takagi K, Wada H, Yamazaki S (1991) Effect of a long term application of a fungicide, chlorothalonil (TPN) on upland ecosystem. Soil Sci Plant Nutr 37:583–590 Tang J, Alexander M (1999) Mild extractability and bioavailability of polycyclic aromatic hydrocarbons in soil. Environ Tox Chem 18:2711–2714 Valverde-Garcia A, Gonzalez-Pradas E, Aguile-Del Real A, UreñaAmate MD (1993) Determination and degradation study of chlorothalonil residues in cucumbers, peppers, and cherry tomatoes. Anal Chim Acta 276:15–23 van der Pas LJT, Matser AM, Boesten JJTI, Leistra M (1999) Behavior of metamitron and hydroxy-chlorothalonil in low-humic sandy soils. Pestic Sci 55:923–934 van Raij B, Quaggio JA, Cantarella, H, Ferreira ME, Lopes AS, Bataglia OC (1987) Análise química do solo para fins de fertilidade. Fundaçào Cargill, Campinas, 170 pp Walker WW, Cripe CR, Pritchard PH, Bourquin AW (1988) Biological and abiotic degradation of xenobiotic compounds in in vitro estuarine water and sediment/water systems. Chemosphere 17: 2255–2270