Behavior of nursery-box-applied fipronil and its sulfone metabolite in rice paddy fields

Agriculture, Ecosystems and Environment 179 (2013) 69–77

Contents lists available at ScienceDirect

Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee

Behavior of nursery-box-applied fipronil and its sulfone metabolite in rice paddy fields Dang Quoc Thuyet, Hirozumi Watanabe ∗ , Takashi Motobayashi, Junghun Ok Tokyo University of Agriculture and Technology, 3-5-8, Saiwaicho, Fuchu, Tokyo 183-8509, Japan

a r t i c l e

i n f o

Article history: Received 20 November 2012 Received in revised form 25 July 2013 Accepted 26 July 2013 Available online 24 August 2013 Keywords: Nursery box Granular insecticide Fipronil Fipronil sulfone Metabolite Paddy field

a b s t r a c t The granular insecticide fipronil has been widely applied in rice nursery boxes, both before transplanting (BT) and during at-sowing (AS) treatments to control insect pests at the early stages of rice cultivation in Japan. Although a potential effect of fipronil on paddy ecosystems and downstream aquatic environments has been observed, the environmental effect of this substance in paddy fields remains unsought. Here we investigate the environmental behavior of nursery-box-applied granular fipronil and its sulfone metabolite in paddy water and paddy soils during BT and AS treatments performed in a paddy field in Japan. Although the fipronil concentrations in the paddy water in the AS treatment were significantly lower than those measured in the BT treatment, no significant differences were observed in the paddy soil between the two treatments. Fipronil was mainly found in the 0- to 5-cm surface soil layer of the rice-root zone, where its concentrations were approximately ten times higher than those in the soil of the inter-row zone. The insecticide concentration in the 0- to 1-cm layer of the inter-row zone in the surface soil was approximately 2.5 times higher than that in the 0- to 5-cm layer. The maximum concentrations of fipronil in the 0- to 1-cm surface soil layer ranged from 65.8 to 92.1 ␮g/kg on the first day after rice transplanting (DAT), and the corresponding values in the paddy water ranged from 0.9 to 2.5 ␮g/L. The dissipation of fipronil from the paddy water and paddy soil was described by first-order kinetics. The compound’s half-life (DT50 ) was 0.9–3.1 days in paddy water and 12.3–26.4 days in paddy soil. Compared to the BT treatment, the AS treatment may pose a smaller risk to the paddy water and the adjacent environment. Fipronil sulfone was found in every water and soil sample, with the maximum concentrations ranging from 0.4 to 0.9 ␮g/L in the paddy water and from 9.7 to 59.2 ␮g/kg in the paddy soil on the third DAT. These values gradually decreased over time. Ecotoxicological risk assessments of fipronil products in rice paddies should not only consider the toxicity of fipronil itself but also that of fipronil sulfone because of its relatively high concentrations in paddy water and paddy soil © 2013 Elsevier B.V. All rights reserved.

1. Introduction The application of pesticides in rice nursery boxes has become a popular method to control pests during the early periods of rice cultivation in Asia. Fipronil, 5-amino-1-[2,6dichloro-4-(trifluoromethyl)phenyl]-4-(trifluoromethylsulfinyl)1H-pyrazole-3-carbonitrile, is a systemic insecticide (Aajoud et al., 2003), which is available as granular formulations and has been popularly used in rice nursery-box applications in Japan to control rice-leaf folders (Cnaphalocrocis medinalis) and brown plant hoppers (Nilaparvata lugens) (Syobu et al., 2002; Gyoutoku et al., 2007). The total application of granular fipronil has annually increased, and the average application of fipronil granules (1% a.i.) during 1998–2009 was 721 ton/year (JPPA, 2000–2010). In

∗ Corresponding author. Tel.: +81 42 367 5889; fax: +81 42 367 5889. E-mail address: [email protected] (H. Watanabe). 0167-8809/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.agee.2013.07.012

conventional rice cultivation, pesticides are directly applied to nursery boxes (30 cm × 60 cm) containing rice seedlings before these rice seedlings are transplanted into the rice paddy by rice transplanters (Kurogochi, 2003). Depending on the farmer’s agricultural practice, fipronil granules can be applied at different periods, such as before transplanting (BT) or at the time of sowing (AS) (Thuyet et al., 2011a). Flooded paddies are considered as a safe environment for many aquatic species (Tourenq et al., 2001; Bambaradeniya et al., 2004; Wilby et al., 2006). However, extensive use of fipronil may have unfavorable effects on nontarget aquatic organisms, and it has been reported to have a significant ecological impact on aquatic communities in paddy ecosystems (Hayasaka et al., 2012; Sánchez-Bayo et al., 2013). Fipronil and its metabolites have also been detected in aquatic environments, such as in the Sakura River, which flows through a region comprising many paddy fields, during the pesticide application period in May 2008 and 2009 in the Ibaraki prefecture, Japan (Iwafune et al., 2010, 2011). Many studies have

70

D.Q. Thuyet et al. / Agriculture, Ecosystems and Environment 179 (2013) 69–77

shown that fipronil sulfone, a fipronil metabolite, has a similar or even greater toxicity to aquatic invertebrates than the parent fipronil (Schlenk et al., 2001; Aajoud et al., 2003; Gunasekara et al., 2007). Iwafune et al. (2011) studied Japanese aquatic organisms and found that the 48-h half maximal effective concentration (EC50 ) values of fipronil and fipronil sulfone for the caddisfly Cheumatopsyche brevilineata (Iwata; Trichoptera: Hydropsychidae) were 0.133 and 0.066 ␮g/L, respectively, and those for Daphnia magna (Straus; Cladocera: Daphniidae) were 42.9 and 5.17 ␮g/L, respectively. The studies of the environmental impact of fipronil on paddy fields suggest that the dragonfly (Sympetrum species) population has rapidly decreased since 1989, and that the observed reduction was positively correlated with the increased use of nursery-box-applied fipronil and imidacloprid (Jinguji et al., 2009; Ueda, 2011). After only nine days of fipronil application in a micro paddy lysimeter, the survival of Sympetrum infuscatum decreased to 0% (Jinguji et al., 2012). The use of fipronil has been restricted in China since 2009 owing to its excessive application in fields and its harmful effects on shellfish, bees, and the river environment (Zhao et al., 2011). Although a potential effect of fipronil on paddy ecosystems has been reported, the environmental fate of nursery-box-applied fipronil granules and its metabolites has not been thoroughly studied to date in rice paddies. The failure to account for the contribution of fipronil sulfone to the total toxicity of fipronil products can lead to a significant underestimation of the toxicity of fipronil containing product to the environment. A better understanding of the behavior of fipronil and its metabolite, fipronil sulfone, in rice paddies is required to minimize the effects of these substances on the environment and nontarget species, as well as to improve the risk assessment in aquatic ecosystems. Therefore, the aim of this study was to investigate the environmental fate of nursery-boxapplied granular fipronil and its sulfone metabolite in a Japanese rice paddy under BT and AS treatments.

2. Materials and methods 2.1. Field experiments Experiments were conducted in two paddy plots of 26 m × 7 m (1 BT and 1 AS) from May 15 to June 20, 2008 and in six paddy plots of 8 m × 6 m (3 BT and 3 AS) from May 15 to June 20, 2009 at the experimental paddy field of the Tokyo University of Agriculture and Technology (TUAT) in Fuchu, Tokyo, Japan. Prior to the experiments, a 15-cm-thick surface soil layer was collected from the experimental field to determine the physical and chemical properties of the soil. The soil’s pH was measured by the glass electrode method (JSSSPN, 1997), and the organic carbon content and cationexchange capacity of the samples were determined following the methods developed by Allison (1965) and Cantino (1944), respectively. The total carbon and total nitrogen contents were measured by the dry combustion method (JSSSPN, 1997). Measurements of the particle density and soil particle distribution were performed following the standard soil particle density test (JSA, 1990) and the pipette method (JSSSPN, 2004), respectively. The soil presented in the experimental plots was a light clay soil with an organic carbon content of 3% (Thuyet et al., 2011a). Further physicochemical properties of this soil are shown in Table 1. The field setup and water management of the paddy plots were similar to those reported in our recent study (Thuyet et al., 2011a). Briefly, the paddy plots were irrigated daily, up to 3 cm during the first week and up to 4 cm during the following weeks if the water level was <3 cm and <4 cm, respectively. Water was held in the paddy plots using plastic borders, and no runoff occurred during the monitoring period, except following large rainfall events. The water-balance components, such as rainfall, evapotranspiration, water level, and surface runoff,

Table 1 Physicochemical properties of paddy water and paddy soil (0–15 cm) in the experimental plots. Water

Value

pHa Eh (mV)a EC (␮S/cm)a

8.4 ± 0.5 528.6 ± 61.0 206.1 ± 73.8

Soil

Value

Eh at 3 cm (mV)a Eh at 1 cm (mV)a pH (H2 O)b Organic carbon content (%)b Total carbon content (%)b Total nitrogen content (%)b Cation exchange capacity (cmolc /kg) b Particle density (g/cm) b Sand (%)b Silt (%)b Clay (%)b Soil texture (ISSS)b

−119.9 ± 144.2 333.4 ± 157.5 6.3 3.0 5.2 0.35 20.1 2.58 40.7 32.9 26.4 Light clay (LiC)

ISSS, International Society of Soil Science. a Average values of daily monitored data at 4:00 PM during the experimental period. b Measurement values prior to the experimental period.

are shown in Fig. 1 and Table 2. Rainfall data was collected from the meteorological station at TUAT, evapotranspiration was measured using a lysimeter (30 cm × 50 cm × 30 cm) placed next to the experimental plots, and the water level and surface runoff were recorded using water-level sensors (LSP-100, UIJIN Co. Ltd., Tokyo). Percolation was calculated from the other monitored water-balance components using the water-balance equation (Watanabe et al., 2007). The redox potential (Eh), pH, and electrical conductivity (EC) of the paddy water and soil were also daily monitored during the experimental period (Thuyet et al., 2011a), and the results are shown in Table 1. The granular insecticide Prince® (1% fipronil; BASF Agro Ltd., Tokyo, Japan) was applied to nursery boxes at the recommended rate of 10 kg/ha. For the AS treatment, the nursery boxes were filled with approximately 2 cm of bed soil, followed by germinated rice seeds, the prescribed amount of granular fipronil, and approximately 1 cm of soil on top at sowing, approximately 14 days before transplanting. Prince® was applied on the surface of the BTtreatment nursery boxes immediately before transplanting. At the time of transplanting, the rice seedlings in the nursery box were 14-days old, approximately 10- to 15-cm tall, and at a leaf stage of around 2–2.5. The rice seedlings for both AS and BT treatments were transplanted to the surface soil with a spacing of 16 cm × 30 cm using a transplanting machine. The average transplanted depth was 4.4 cm from the surface soil in 2009 but was not measured in 2008. 2.2. Sampling and chemical analysis Five soil samples were collected from each nursery box immediately before transplanting by driving a stainless core (2.5 cm in diameter) from the surface soil down to the bottom of the nursery box to examine the initial fipronil concentrations in the nurserybox soils for the BT and AS treatments performed in 2008 and 2009. Paddy water and a 0- to 1-cm surface paddy soil layer in the inter-row zone were collected on the 1st, 3rd, 7th, 14th, 21st, 28th, and 35th days after rice transplanting (DAT) in both 2008 and 2009, following the methods described by Thuyet et al. (2011a). Briefly, five 125-ml surface water samples were taken from five spots within each paddy plot (i.e., from a central spot and from the four corners) and were then combined into one representative water sample for each sampling day (Thuyet et al., 2011a). Similarly, a composite paddy-soil sample containing five 50-g surface

D.Q. Thuyet et al. / Agriculture, Ecosystems and Environment 179 (2013) 69–77

71

Fig. 1. The daily water balance in the before-transplanting (BT) (a, c) and at-sowing (AS)-treated paddy plots (b, d) in 2008 and 2009.

soil samples (0–1 cm) in the inter-row zone was also collected on each sampling day. In addition, 0- to 5-cm surface paddy soil samples in the inter-row and root zones were collected in 2009 by using the same sampling intervals. Initially, a plastic ring (15 cm in diameter) was installed at the sampling location, and the paddy water was removed with a manual pump. Subsequently, the rice plants in the root zone were manually cut and removed, and a stainless core with a diameter of 5 cm was driven from the surface of the paddy soil down to a soil depth of 5 cm to collect a soil sample. The samples were collected in triplicate and kept frozen at −20 ◦ C until chemical analysis was performed. Fipronil standard, fipronil sulfone standard (≥99% purity), analytical-grade solvents (Wako, Osaka, Japan), and a Milli-Q Water

Purification System (Millipore, Billerica, MA, USA) were used for the chemical analysis. Water and soil samples were thawed at ambient temperature before extraction. The samples were then passed through a 1.2-␮m glass-fiber filter (GF/C; Whatman, Maidstone, UK). Extraction of fipronil and fipronil sulfone from the water samples was performed on a solid-phase ENVI C18 Superclean cartridge (500 mg/6 ml, Supelco; Sigma–Aldrich, St. Louis, MI, USA) followed by elution with 10-ml acetonitrile (Jinguji et al., 2012). The eluate was blown down to 1 ml under a gentle stream of nitrogen gas. Fipronil and fipronil sulfone were recovered from the interrow-zone and root-zone soil samples by the extraction method described by Thuyet et al. (2011a). Briefly, the rice seedlings were manually removed from the nursery-box soil before the extraction.

Table 2 Water balance in the before-transplanting (BT) and at-sowing (AS)-treated paddy plots in 2008 and 2009. 2008

2009

BT

Input (cm) Irrigation Precipitation Total Output (cm) Runoff Percolation ET Total

AS

BT

AS

TWD cm

DAWD cm/day

TWD cm

DAWD cm/day

TWD cm

DAWD cm/day

TWD cm

DAWD cm/day

22.7 31.2 53.9

0.6 0.9 1.5

16.2 31.2 47.4

0.5 0.9 1.4

20.5 ± 1.3 21.8 42.3 ± 1.3

0.6 ± 0.0 0.6 1.2 ± 0.0

25.3 ± 1.9 21.8 47.0 ± 1.9

0.7 ± 0.0 0.6 1.3 ± 0.1

12.6 26.3 15.4 54.3

0.4 0.8 0.4 1.6

6.4 25.1 15.4 47.0

0.2 0.7 0.4 1.3

3.9 ± 0.2 24.9 ± 1.4 12.7 41.5 ± 1.4

0.1 ± 0.0 0.7 ± 0.0 0.4 1.2 ± 0.0

8.0 ± 0.4 26.2 ± 1.8 12.7 46.8 ± 1.8

0.2 ± 0.0 0.8 ± 0.1 0.4 1.3 ± 0.0

TWD, total water depth (cm); DAWD, daily average water depth (cm/day).

72

D.Q. Thuyet et al. / Agriculture, Ecosystems and Environment 179 (2013) 69–77

The nursery-box soil (5 g) was finely ground and then sonicated in 20-ml acetonitrile for 20 min. The mixture was shaken for 2 h and then the liquid was decanted, whereas the residue was shaken again with 30-ml acetonitrile for 30 min. The liquid and the mixture were passed through a 1.2-␮m glass fiber (GF/C; Whatman, Maidstone, UK) to separate the soil residue and the extraction solvent. The remaining residues were then washed with 30-ml acetonitrile. All the extracts were combined and filled with acetonitrile up to 100 ml. The paddy soil sample was first centrifuged to remove the soil water and then pulverized. Twenty grams of the dried soil were sonicated for 20 min and shaken for 2 h in 60-ml acetonitrile. The mixture was then passed through a 1.2-␮m glass-fiber filter (GF/C; Whatman, Maidstone, UK) to separate the liquid phase from the soil. The liquid phase was then evaporated under vacuum and the residue was dissolved in 2-ml acetonitrile. All the samples were passed through a 0.2-␮m filter (Whatman, Maidstone, UK) and kept in a high-performance liquid chromatography (HPLC) vial at 4 ◦ C prior to HPLC analysis. Analyses were performed on a Shimadzu HPLC system consisting of an LC20AD Separation Module and an SPD-M20A photodiode array detector. The HPLC conditions have been reported in a previous publication (Jinguji et al., 2012). Briefly, the mobile phase, consisting of acetonitrile/water (60:40, v/v) was passed through a Shimadzu C-18 column (150 mm × 4.6 mm × 4.6 ␮m, Shimadzu Corporation, Kyoto, Japan) in isocratic mode at a flow rate of 1 ml/min and a constant temperature of 40 ◦ C. The detector was set at a wavelength of 280 nm. The retention times of fipronil and fipronil sulfone were 7.8 and 11.2 min, respectively, and the detection limits for the two compounds were 0.05 and 0.08 ␮g/L, respectively, in the water samples, and 0.1 and 0.2 ␮g/kg, respectively, in the paddy soil samples. The recoveries (n = 3) of fipronil and fipronil sulfone were 93.0 ± 4.6% and 89.3 ± 3.1%, respectively, in the water samples, and 85.5 ± 7.3% and 87.5 ± 4.7%, respectively, in the paddy soil samples. 2.3. Statistical analysis A paired two-tailed Student’s t-test was performed in this study to compare the pesticide concentrations related to the AS and BT treatments. The significant difference was determined at the 95% confidence level. The degradation rate of the pesticide was calculated from the following first-order kinetics equation (Eq. (1)): Ct = C0 e−kt

(1)

Or from the logarithm-transformed equation Ln(Ct ) = −kt + Ln(C0 )

(2)

where Ct represents the concentration of pesticide at time t (in ␮g/L), C0 represents the initial concentration of pesticide (in ␮g/L), k is the degradation rate (in 1/days), and t is the time after pesticide application (in days). The half-life (DT50 ) of the pesticide was calculated from the following equation: DT50 =

ln 2 k

(3)

The DT50 calculation started with transplanting and the DT50 of fipronil in paddy water and paddy soil were determined from the substance’s concentrations in the corresponding samples during the first 7 and 35 days after transplanting. Eq. (2) was established by using the linear least-squares regression method. Upper and lower 95% confidence intervals were also calculated for each equation. All the calculations were performed with Microsoft Excel 2010.

3. Results and discussion 3.1. Water-balance monitoring Six major precipitation events (≥2 cm/day) were observed during the monitoring period in 2008. The largest rainfall event was 9.7 cm and occurred on the fourth DAT, which resulted in 5.0- and 4.9-cm runoffs on this day from the BT and AS plots, respectively (Fig. 1a and b). Another runoff event, which occurred on the 13th DAT (Fig. 1a and b), resulted in 2.2- and 1.0-cm runoffs from the BT and AS plots, respectively. An unexpected 4.2-cm runoff occurred on the 16th DAT in the BT plot because of a discharge from a damaged border after consecutive rainfall events on the 13th to 15th DATs (Fig. 1a). The average percolation rates in 2008 for the BT and AS plots were 0.8 and 0.7 cm/day, respectively (Table 2). The rainfall pattern during the monitoring period in 2009 is shown in Fig. 1c and d. The rainfall intensities and total rainfall amount during this period were lower than those in 2008 (Fig. 1c and d). The average total rainfall in the experimental plots obtained between May 15 and June 20 during the last 15 years, from 1998 to 2012, was 20.5 ± 7.1 cm, whereas the values measured in 2008 and 2009 were 31.2 and 21.8 cm, respectively (Table 2). This indicates a normal rainfall volume in 2009 but a fairly high rainfall volume in 2008. In 2009, most of the rainfall events occurred during the second and third weeks after transplanting (Fig. 1c and d). The percolation rates for the BT and AS plots obtained in 2009 were 0.7 and 0.8 cm/day, respectively (Table 2). 3.2. Initial fipronil conditions in the nursery boxes before transplanting The granular pesticide was applied to the AS-treated nursery boxes 14 days before transplanting; thus, a part of the initial mass might have been lost by degradation and runoff from the bottom of the nursery boxes owing to over irrigation during incubation. However, the initial concentrations of fipronil in the nursery-box soil samples measured before transplanting were 69.6 ± 11.8 mg/kg (n = 5) and 63.7 ± 10.3 mg/kg (n = 5) for the BT and AS treatments, respectively; hence, the difference was not significant (t-test, p = 0.42). In other words, the initial concentrations of fipronil in the BT and AS treatment nursery boxes can be considered to be the same at the time of transplanting. These results agreed with those obtained for granular imidacloprid applied to rice nursery boxes (Thuyet et al., 2011a). Although fipronil sulfone was found both in the BT and AS treatment nursery boxes, the concentration was relatively low, accounting for 0.6% and 1.7% of the applied mass of fipronil in the nursery boxes for the BT and AS treatments, respectively, at the time of transplanting. The insignificant pesticide loss observed during the 14 days of incubation in the nursery box for the AS treatment could be explained by the fact that the box was left unsubmerged during this time so that most of the fipronil remained in the granular formulation, because the relatively low water content prevented its mobility. The degradation of fipronil is known to depend on the soil’s moisture content; for example, Ying and Kookana (2002) reported that the half-lives of this substance were 68 and 198 days in the same soil at 60% and 15% moisture, respectively. Any released fipronil in the nursery-box soil should therefore degrade more slowly because the soil moisture in the nursery box is lower than that in the paddy soil. 3.3. Fipronil behavior in paddy water The granular pesticide with the nursery-box soil, and the rice seedlings were driven into the paddy-field soil with a transplanting machine. Although most of the granular pesticide was placed in the rice-seedling root zone, a small portion of the granules drifted

D.Q. Thuyet et al. / Agriculture, Ecosystems and Environment 179 (2013) 69–77

(a) 2008

Concen ntration (µg/L L)

3.0 25

The decrease in pesticide concentration in the paddy water observed over time can be generally explained by dilution upon irrigation and losses through percolation and runoff, in addition to photodegradation, hydrolysis, and biochemical degradation (Thuyet et al., 2011b). The decrease could also result from pesticide sorption from the paddy water to the paddy soil. In this study, photodegradation was considered to be the major fipronil degradation process in paddy water, because Thuyet et al. (2011b) reported that fipronil is very sensitive to solar radiation, and that its photodegradation half-life is 36.7 h in paddy water. The concentrations of the insecticide in the paddy water sharply declined to values below the detection limit on the seventh DAT in 2008 (Fig. 2a) because of precipitation (approximately 9.7 cm) on the fourth DAT (Fig. 1a and b). However, no rainfall occurred during the first week in 2009 (Fig. 1c and d), and therefore, fipronil could still be detected on the seventh DAT (Fig. 2b). The slow decrease in fipronil concentration observed in the paddy water after the seventh DAT in 2009 (Fig. 2b) could be explained by a release of adsorbed fipronil from the paddy soil into the paddy water.

Firponil - BT Firponil Fipronil sulfone - BT Firponil - AS Fipronil sulfone - AS

2.0 1.5 10 1.0 0.5 0.0 0

40

(b) 2009

3.0 Concentration (µg/L)

10 20 30 Day after transplanting

2.5

Fipronil - BT Fipronil sulfone - BT Fi Fipronil il - AS Fipronil sulfone - AS

2.0 15 1.5 1.0

3.4. Fipronil behavior in paddy surface soil

0.5 00 0.0 0

10 20 30 Day after transplanting

73

40

Fig. 2. Concentrations of fipronil and fipronil sulfone in paddy water in 2008 (a) and 2009 (b).

around the root zone of the surface soil in the paddy field. Some granules were present on the surface soil and exposed to paddy water after transplanting, whereas the rest was buried in the surface paddy soil. The pesticide was assumed to desorb from the buried granules to the surface soil (Thuyet et al., 2011a, 2012b). Nursery-box-applied fipronil is involved in two major processes in paddy water, namely in the dissolution of the granular pesticide that is directly exposed to the paddy water and in the desorption of buried pesticide from the surface of the paddy soil. Fig. 2 shows the fipronil concentrations in paddy water for both BT and AS treatments in 2008 and 2009. The fipronil concentration reached a maximum value on the first DAT [ranging from 0.9 ± 0.1 to 2.5 ± 0.3 ␮g/L, which corresponds to 0.2–2.4% of the initial applied mass] and then rapidly decreased until the third DAT. Hayasaka et al. (2012) also observed that the fipronil concentration was below 1.0 ␮g/L in paddy plots shortly after transplanting. The fipronil concentration was quite low after the seventh DAT in 2009 and then slowly decreased, whereas it was below the detection limit after the seventh DAT in 2008. The fipronil concentration on the 35th DAT varied from not detectable to 0.2 ± 0.1 ␮g/L in 2009. The pesticide concentrations on the first DAT in both AS and BT plots were higher in 2008 than in 2009, probably owing to the initial field conditions and the transplanting practice, because different transplanting machines were used in 2008 and 2009. The type of transplanting machine and its settings affect the transplanting depth and pesticide drift conditions. In addition, local differences in water depth, resulting from incomplete field leveling, affected the exposure of the pesticide in the paddy water, while soil conditions, such as the elasticity resulting from soil paddling during soil preparation, may have affected the degree of pesticide buried in the soil.

Fipronil has low-to-moderate sorption characteristics in paddy soil, with the sorption coefficient (Koc ) ranging from 542 to 1176 depending on the soil type (Ying and Kookana, 2001; Gunasekara et al., 2007), as well as a low-to-moderate water solubility (2.4 mg/L, pH = 9) (Tomlin, 2006). The estimated soil–water partitioning coefficient (Kd ) of fipronil, determined using the average KOC value reported above, was 25.6 L/kg in this experiment, which suggests that the pesticide should be found at relatively higher concentrations in the paddy soil compared to the paddy water. The fipronil concentrations determined in the 0- to 1-cm surface soil layer of the inter-row zone in 2008 and 2009 are presented in Fig. 3. The average maximum concentration ranged from 65.8 ± 4.9 to 92.1 ± 6.7 ␮g/kg on the first DAT, which accounts for 7.2–8.3% of the applied fipronil mass, and the concentration gradually decreased over time. At the end of the monitoring period on the 35th DAT, the fipronil concentration was found to be between 10.6 ± 0.9 and 18.4 ± 7.0 ␮g/kg. Pesticide concentrations in the paddy soil may vary depending on the transplanting practice and soil characteristics (Thuyet et al., 2011a). Hayasaka et al. (2012) observed significant variations in the fipronil concentrations of surface soils during their monitoring period. The distribution of fipronil in the root zone was compared with that in the inter-row zone in 2009. The differences in concentration observed in Fig. 4a and b indicate that fipronil mainly persisted in the rice-root zone, where the concentration was approximately ten times greater than that in the inter-row zone. However, the recovered fipronil mass in the root-zone core represented approximately 10% of the total applied mass. It should be noted that in this study, the mass of pesticide in the root zone was considered to be the mass of pesticide available inside the sampling cylinder core, which had an internal diameter of 5 cm and a height of 5.1 cm, corresponding to a total sampling volume of 100 ml. The root-zone sample was collected at the top 0- to 5-cm surface soil layer surrounding the rice-root area, and therefore, the root zone may not have covered the entire plume of pesticide surrounding this zone. Neither the distribution of fipronil surrounding the riceroot zone nor the fast transport of fipronil at depths below 5 cm, plant uptake or volatilization, were measured in this study. The volatilization of fipronil is negligible because of its low vapor pressure of 3.7 × 10−4 mPa (Gunasekara et al., 2007). During machine transplanting, it was observed that some granules on the rice nursery plant were lost on the paddy soil and others were floating on the paddy water. Therefore, it was not possible to make a complete mass balance for fipronil in the monitored rice paddy. The

D.Q. Thuyet et al. / Agriculture, Ecosystems and Environment 179 (2013) 69–77

(a) 2008

100 80 60

(a) Inter-row zone

Firponil - BT Fipronil sulfone - BT Firponil - AS Fipronil sulfone - AS

50 Conc centration (µg g/kg)

Concentration (µg/kg g)

74

40 20 0 10 20 Day after transplanting

(b) 2009

Concentrattion (µg/kg)

100 80 60

30

40

20 0 10 20 Day after transplanting

30

30 20 10 0

10 20 Day after transplanting

(b) Root zone

Fipronilil - BT Fi Fipronil sulfone - BT Fipronil - AS Fipronil sulfone - AS

40

0

40

0

Concentratio on (µg/kg)

0

Fipronil - BT Fipronil sulfone - BT Fi Fipronil il - AS Fipronil sulfone - AS

40

Fig. 3. Concentrations of fipronil and fipronil sulfone in the 0- to 1-cm soil surface layer in paddy soil in the inter-row zone in 2008 (a) and 2009 (b).

concentrations of fipronil in the top 0- to 5-cm surface soil layer on the first DAT were 343.8–394.3 ␮g/kg in the root zone and 29.6–31.8 ␮g/kg in the inter-row zone. The values on the 35th DAT were 140.9–157.0 ␮g/kg in the root zone and from below the detection limit to 9.4 ␮g/kg in the inter-row zone (Fig. 4). On the first DAT, approximately 80% of the fipronil in the top 0–5 cm surface soil of the inter-row zone was concentrated in the top 0–1 cm (Figs. 3b and 4a). Large variations in the fipronil concentrations were observed among triplicates in the 0- to 5-cm surface soil layer (Fig. 4a), compared to the 0- to 1-cm surface soil layer (Fig. 3), in the inter-row zone. The downward movement of fipronil in the paddysoil profile is controlled by the vertical percolation rate. The large fipronil variations observed in the top 0- to 5-cm surface soil layer possibly result from a heterogeneous spatial distribution of the percolation in this layer. A large variation in the percolation rate in a paddy plot was also reported by Watanabe et al. (2007). The fipronil concentrations in the surface paddy soil gradually decreased during the monitoring period because of biodegradation and pesticide loss through percolation. 3.5. Dissipation kinetics of fipronil in paddy water and paddy soil In 2008 and 2009, the dissipation of fipronil in paddy water and paddy soil was found to fit first-order kinetics (r2 ≥ 0.6), with the dissipation rate in paddy water being faster than that in paddy soil. The DT50 values of fipronil are presented in Table 3, where it can be seen that in 2008 and 2009, DT50 was 0.9–3.1 days in the paddy water and 12.3–26.4 days in the paddy soil in the inter-row zones in both BT and AS treatments (Table 3). The dissipation of fipronil from the paddy water and the 0- to 1-cm surface paddy soil was faster in 2008 than in 2009 (Table 3). The faster dissipation of fipronil in

500 400 300

30

40

Fipronilil - BT Fi Fipronil sulfone - BT Fipronil - AS Fipronil sulfone - AS

200 100 0 0

10 20 30 Day after transplanting

40

Fig. 4. Concentrations of fipronil in the 0- to 5-cm soil surface layer in paddy soil in the inter-row (a) and root zones (b) in 2009.

2008 is probably due to the large amount of rainfall that occurred during the first week after transplanting (Fig. 1a and b), whereas there was almost no rainfall during that period in 2009 (Fig. 1c and d). Fipronil slowly degraded in the root zone, and its DT50 was >19 days. The DT50 values of fipronil in paddy water were reported to be 5.4 and 1.1 days during the rice cultivation periods in 2010 and 2011, respectively (Hayasaka et al., 2012). Fipronil dissipation from the paddy water was very fast, similar to what has been reported for other pesticides, such as imidacloprid (2.0–2.4 days) (Thuyet et al., 2011a), whereas the half-lives of the fungicide tricyclazole were reported to range from 11.4 to 12.1 days (Phong et al., 2009) and those of herbicides from 2 to 6 days (Watanabe et al., 2007). For several paddy soils in China, Tan et al. (2008) reported that the DT50 of fipronil was 21–34 days. Hayasaka et al. (2012) also estimated DT50 values of 25.7 and 28.5 for this substance during the 2010 and 2011 rice-cultivation periods, respectively. Fipronil tends to degrade faster in paddy water than in arable soil, DT50 = 65 days (PPDB, 2013). A higher dissipation rate in paddy soils could result from a higher moisture content, hydrolysis, and biodegradation of fipronil in the paddy soil. A laboratory study demonstrated that the moisture content of a soil can greatly affect the dissipation of fipronil, and that the fipronil dissipation half-live increases with soil moisture content (Ying and Kookana, 2002). In this study, the paddy soil was submerged during the monitoring period, thus its water content could be saturated and much higher than that of an unsaturated arable soil. The higher water content could also lead to a higher hydrolysis rate of fipronil. Moreover, the paddy soil was considered to be a favorable environment for many aquatic microorganisms (Tourenq et al., 2001; Bambaradeniya et al., 2004; Wilby et al., 2006; Zhang et al., 2008), which means that the

D.Q. Thuyet et al. / Agriculture, Ecosystems and Environment 179 (2013) 69–77

75

Table 3 Dissipation kinetics of fipronil in paddy water and paddy soil the before-transplanting (BT) and at-sowing (AS)-treated paddy plots in 2008 and 2009. 2008 2

r DT50 e (day)

BT-Wa

BT-S-1cmb

AS-Wa

AS-S-1cmb

1.0 0.9 (0.9, 1.0)

0.9 12.5 (11.1, 14.4)

1.0 1.1 (0.9, 1.2)

0.9 12.3 (10.3, 15.1)

BT-Wa

BT-S-1cmb

BT-S-5cmc

BT-RS-5cmd

r DT50 e (day)

0.9 2.9 (2.1, 4.4)

0.6 15.5 (11.3, 24.3)

0.7 26.4 (20.1, 38.4)

0.9 19.4 (16.8, 23.1)

2009

AS-Wa

AS-S-1cmb

AS-S-5cmc

AS-RS-5cmd

0.9 3.1 (2.3, 4.6)

0.8 16.8 (13.3, 22.9)

0.7 19.2 (13.5, 33.0)

0.9 26.2 (23.1, 30.2)

2009 2

2

r DT50 e (day)

r2 is the square of the correlation coefficient of natural logarithm of concentration of fipronil and the time after transplanting. DT50 of fipronil was calculated for the first 7 days in paddy water and the 35 days in paddy soil after transplanting. a The values are corresponding to paddy water. b The values are corresponding to 0–1 cm surface paddy soil in the inter-row zone. c The values are corresponding to 0–5 cm surface paddy soil in the inter-row zone. d The values are corresponding to 0–5 cm surface paddy soil in the root zone. e Lower and upper 95% confidence intervals are provided in parentheses.

microbial activity and consequent biodegradation of fipronil in paddy soils could be higher than those in arable soils. 3.6. Effects of the BT and AS treatments on the fate of fipronil Thuyet et al. (2011a) reported that an AS treatment with a controlled-release granular formulation clearly leads to lower concentrations of imidacloprid both in the paddy water and the 0- to 1-cm surface soil compared to a BT treatment performed using a conventional-release granular formulation. However, the study of Thuyet et al. (2011a) could not identify whether the difference resulted from the treatment method or the granular formulation. In the present study, fipronil was formulated in the same conventional granular formulation and was applied to a paddy field following the BT and AS treatment methods during two consecutive years. The results showed that the BT treatment led to a higher fipronil concentration in the paddy water compared to the AS treatment. Significant differences were observed in the peak fipronil concentrations in paddy water in 2008 and 2009, with the maximum values being 2.5 ± 0.3 and 1.2 ± 0.2 ␮g/L (t-test, p = 0.003) in 2008 and 1.2 ± 0.1 and 0.9 ± 0.1 ␮g/L (t-test, p = 0.01) in 2009 for the BT and AS treatments on the first DAT, respectively. The difference was still clear in 2008, during the later period, but not in 2009 (Fig. 2). The fipronil concentrations in the 0- to 1-cm surface soil layer were 92.1 ± 6.7 and 90.8 ± 6.9 ␮g/kg (t-test, p = 0.83) in 2008 and 75.1 ± 6.7 and 65.8 ± 4.9 ␮g/kg (t-test, p = 0.12) in 2009 for the BT and AS treatments, respectively, on the first DAT. The differences in fipronil concentrations were not significant between the BT and AS treatments neither in the top 0- to 1-cm surface soil (Fig. 3a and b) nor in the top 0- to 5-cm surface soil (t-test, p = 0.69; on the first DAT) for both the inter-row and root zones (t-test, p = 0.09; on the first DAT) (Fig. 4), indicating that the fipronil concentration in the surface soil is not affected by the treatment method. The characteristics of the BT and AS treatments significantly affect the environmental fate of fipronil in the paddy water. After transplanting, some fipronil granules were transferred onto the surface soil and directly dissolved into the surface water, resulting in peak concentrations on the first DAT. The remaining fipronil granules were covered by soil and slowly diffused into the surrounding soil and water. Fipronil was then partitioned into the surface soil and paddy water in accordance with its soil/water partitioning coefficient. As granules were applied over the nursery-box soil in the BT treatment, most of the granules stayed in the surface paddy soil after transplanting. Because the granules were directly exposed to the paddy water, they were more likely to dissolve into

the paddy water. In contrast, the granules in the AS treatment were covered by the nursery-box soil; thus, they had less chance to come in contact and dissolve into the paddy water as compared to the BT treatment (Thuyet et al., 2011a). This difference probably resulted in a higher fipronil concentration in the paddy water for the BT treatment compared to the AS treatment. Although the differences in fipronil concentrations were not significant between the BT and AS treatments in the paddy soil, they were significant in the paddy water. The ratio of fipronil concentration in the surface soil to that in the paddy water on the first DAT was 36.9 and 68.5 in 2008 and 63.4 and 74.9 in 2009 for the BT and AS treatments, respectively. The AS treatment led to a lower fipronil concentration in the paddy water as compared to the BT treatment. Therefore, with proper management, the AS treatment may pose a smaller risk to the paddy water and the adjacent aquatic environment compared to the BT treatment. 3.7. Fipronil sulfone in paddy water and paddy soil Fipronil sulfone is a major metabolite formed during the oxidation of fipronil in water and soil. Its toxicity to many species is usually equal or higher than that of fipronil (Gunasekara et al., 2007). In this study, fipronil sulfone was found in most of the water and soil samples (Figs. 2–4), and the maximum concentrations in paddy water were 0.9 and 0.5 ␮g/L for the BT and AS treatments in 2008 and 0.4 ␮g/L for both BT and AS treatments in 2009 on the third DAT (Fig. 2). The fipronil sulfone concentration also peaked on the third DAT in the paddy soil and remained stable until the end of the monitoring period. Large variations in the fipronil sulfone concentrations were observed among replicates (Fig. 4), which is probably a result of the heterogeneous soil conditions. The formation of fipronil sulfone depends on the chemical properties of the medium, particularly the Eh of the paddy water and paddy soil. Oxygen is supplied to the paddy water through dissolved air and the photosynthesis of algae and other aquatic plants. Rice-field soils are divided into different zones according to the availability of O2 and oxidized ions, such as nitrate (NO3 − ), sulfate (SO4 2− ), manganese (Mn4+ ), ferric (Fe3+ ) ions, and microbial respiration. The top 0- to 1-cm surface soil layer is considered to be an oxidized layer because of O2 diffusion and transport from the paddy water and percolation (Tsukano, 1986; Kohno et al., 1995; Takagi et al., 1998). Beneath the oxidized layer, plow soil gradually becomes anaerobic after flooding owing to O2 depletion by insufficient diffusion of O2 and decomposition of organic substances, coupled with the reduction of NO3 − , SO4 2− , Mn4+ , and Fe3+

76

D.Q. Thuyet et al. / Agriculture, Ecosystems and Environment 179 (2013) 69–77

(Tsukano, 1986; Stumm and Morgan, 1995). Average Eh values in the paddy water and paddy soil during the monitoring period are presented in Table 1. The Eh values in the paddy water and at 1-cm soil depth are under oxidizing conditions, whereas those at a 3-cm soil depth vary from the slightly oxidative to the reductive state (Table 1). These oxidative conditions could promote the formation of fipronil sulfone via fipronil oxidation. The peak fipronil sulfone concentration on the third DAT in both paddy water and paddy soil (Figs. 2–4) can be explained by the high oxygen concentration during the early days after transplanting. Consequently, fipronil sulfone was detected in most of the water and soil samples. There were no obvious differences between the fipronil sulfone concentrations in the AS-treated and BT-treated plots in both 2008 and 2009. However, the concentrations of fipronil sulfone both in paddy water and paddy soil were relatively high during the entire monitoring period, and the dissipation rates in both media were lower than those of fipronil. The relatively high concentrations of fipronil sulfone, measured in paddy water and paddy soil for both AS and BT treatments, indicate the significant contribution of this compound to the total toxicity of fipronil products on aquatic communities in paddy ecosystems. The presence and persistence of fipronil sulfone resulted in the increase and persistence of the ecological pesticide risk in paddy ecosystems. Jinguji et al. (2012) observed the extinction of the dragonflies Sympetrum infuscatum and Sympetrum frequens in the paddy water of a micro paddy lysimeter within nine days after fipronil application to the system, although the fipronil concentration was relatively low (0.4–1.3 ␮g/L). The presence of fipronil sulfone in the paddy water, as observed in this study, could have partially contributed to this extinction. Fipronil sulfone was reported to account for 33.3% to 52.8% of the total toxicity of fipronil products on the bluegill sunfish Lepomis macrochirus in a water sample (Thuyet et al., 2012a). The effects of fipronil sulfone on the aquatic community are not only limited to paddy ecosystems but they also extend to downstream river ecosystems. Iwafune et al. (2011) found that the risk quotients of fipronil and fipronil sulfone on caddisflies Cheumatopsyche brevilineata in a downstream river flowing through a paddy-field area were 0.391 and 0.191, respectively. Therefore, a comprehensive monitoring and precise ecotoxicological risk assessment, including the parent compound and its metabolite, are very important to accurately assess the ecotoxicological impact of pesticide products and protect paddy ecosystems.

4. Conclusion The behaviors of nursery-box-applied fipronil and fipronil sulfone in a paddy field were affected by the treatment methods. While the AS treatment resulted in a lower fipronil concentration in the paddy water (compared to the BT treatment), the concentrations of fipronil and fipronil sulfone were not significant in the paddy soil for both treatments. Fipronil was mainly found in the 0- to 5-cm surface soil layer in the rice-root zone and in the top 0- to 1-cm surface soil layer. The dissipation of fipronil from the paddy water and surface soil was described by a first-order kinetics. The DT50 values of fipronil were 0.9–3.1 days in paddy water and 12.3–26.4 days in paddy soil. The AS treatment may pose a lower risk to the paddy water and the adjacent aquatic environment compared to the BT treatment. Fipronil sulfone was found in every water and soil sample, reaching a maximum value on the third DAT. Because of its relatively high concentration in paddy water and surface soil and its high toxicity to most species, the contribution of fipronil sulfone to the overall toxicity of fipronil products (in addition to fipronil alone) should be considered when ecotoxicological risk assessments of fipronil are conducted in rice paddy fields.

Acknowledgements This work was partially supported by the Heiwa Nakajima Scholarship Foundation and a Grant-in-Aid for Biological Studies on Wildlife under the Framework ExTEND2005 in Japan. Special thanks to Mr. Santosh Kumar, at Tokyo University of Agriculture and Technology, for his assistant in field monitoring in 2008. References Aajoud, A., Ravanel, P., Tissut, M., 2003. Fipronil metabolism and dissipation in a simplified aquatic ecosystem. Journal of Agricultural and Food Chemistry 51, 1347–1352. Allison, L.E., 1965. Organic carbon. In: Black, C.A. (Ed.), Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties. American Society of Agronomy, pp. 1367–1378. Bambaradeniya, C.N.B., Edirisinghe, J.P., De Silva, D.N., Gunatilleke, C.V.S., Ranawana, K.B., Wijekoon, S., 2004. Biodiversity associated with an irrigated rice agroecosystem in Sri Lanka. Biodiversity and Conservation 13, 1715–1753. Cantino, E.C., 1944. Semimicrodetermination of the exchange capacity of soils. Soil Science 57, 399–404. Gunasekara, A.S., Truong, T., Goh, K.S., Spurlock, F., Tjeerdema, R.S., 2007. Environmental fate and toxicology of fipronil. Journal of Pesticide Science 32, 189–199. Gyoutoku, Y., Sone, S., Gunjima, K., Komaki, K., 2007. Residual activities of imidacloprid and fipronil by nursery box application to brown planthoppers collected in 2005. Kyushu Plant Protection Research 53, 24–28 (in Japanese). Hayasaka, D., Korenaga, T., Suzuki, K., Saito, F., Sánchez-Bayo, F., Goka, K., 2012. Cumulative ecological impacts of two successive annual treatments of imidacloprid and fipronil on aquatic communities of paddy mesocosms. Ecotoxicology and Environmental Safety 80, 355–362. Iwafune, T., Inao, K., Horio, T., Iwasaki, N., Yokoyama, A., Nagai, T., 2010. Behavior of paddy pesticides and major metabolites in the Sakura River, Ibaraki, Japan. Journal of Pesticide Science 35, 114–123. Iwafune, T., Yokoyama, A., Nagai, T., Horio, T., 2011. Evaluation of the risk of mixtures of paddy insecticides and their transformation products to aquatic organisms in the Sakura river, Japan. Environmental Toxicology and Chemistry 30, 1834–1842. Jinguji, H., Thuyet, D.Q., Uéda, T., Watanabe, H., 2012. Effect of imidacloprid and fipronil pesticide application on Sympetrum infuscatum (Libellulidae: Odonata) larvae and adults. Paddy and Water Environment 11, 277–284. Jinguji, H., Uéda, T., Goka, K., Hidaka, K., Matsura, T., 2009. Effects of imidacroprid and fipronil insecticide application on the larvae and adults of Sympetrum frequens. Transactions of JSIDRE 259, 355–360 (in Japanese with English abstract). JPPA, 2000–2010. Japan Plant Protection Association, Pesticide Handbook. Japan Plant Protection Association, Tokyo, Japan (in Japanese). JSA, 1990. Test method for density of soil particles, JIS A 1202-1990. Japanese Standards Association (JSA JIS), pp. 7. JSSSPN, 1997. Soil Environment Analysis Method. Japanese Society of Soil Science and Plant Nutrition (JSSSPN), Hakuyusha Co. Ltd., Tokyo (in Japanese). JSSSPN, 2004. Standard soil analysis and measurement method. Japanese Society of Soil Science and Plant Nutrition (JSSSPN), Hakuyusha Co. Ltd., Tokyo (in Japanese). Kohno, E., Ogawa, Y., Iwata, S., 1995. Productivity of paddy fields. In: Tabuchi, T., Hasegawa, S. (Eds.), Paddy Fields in the World. JSIDRE, Tokyo, pp. 341–349. Kurogochi, S., 2003. Translocation and metabolism of pesticides in rice Plants after nursery box application. In: Coats, J.R., Yamamoto, H. (Eds.), Environmental Fate and Effects of Pesticides. American Chemical Society, pp. 139–156. Phong, T.K., Nhung, D.T.T., Motobayashi, T., Thuyet, D.Q., Watanabe, H., 2009. Fate and transport of nursery-box-applied tricyclazole and imidacloprid in paddy fields. Water, Air, & Soil Pollution 202, 3–12. PPDB, 2013. The Pesticide Properties DataBase (PPDB). http://sitem.herts.ac.uk/aeru/iupac/ Sánchez-Bayo, F., Tennekes, H.A., Goka, K., 2013. Impact of systemic insecticides on organisms and ecosystems. In: Trdan, S. (Ed.), Insecticides—Development of Safer and More Effective Technologies. InTech, pp. 367–416. Schlenk, D., Huggett, D.B., Allgood, J., Bennett, E., Rimoldi, J., Beeler, A.B., Block, D., Holder, A.W., Hovinga, R., Bedient, P., 2001. Toxicity of fipronil and its degradation products to Procambarus sp.: field and laboratory studies. Archives of Environmental Contamination and Toxicology 41, 325–332. Stumm, W., Morgan, J.J., 1995. Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters. Wiley-Interscience, New York. Syobu, S., Mikuriya, H., Yamaguchi, J., 2002. Effect of nursery box application of fipronil on the rice leaf-folder, Cnaphalocrocis medinalis Guenee. Kyushu Plant Protection Research 48, 60–63 (in Japanese). Takagi, K., Fajardo, F.F., Inao, K., Kitamura, Y., 1998. Predicting pesticide behavior in a lowland environment using computer simulation. Reviews in Toxicology 2, 269–286. Tan, H., Cao, Y., Tang, T., Qian, K., Chen, W.L., Li, J., 2008. Biodegradation and chiral stability of fipronil in aerobic and flooded paddy soils. Science of the Total Environment 407, 428–437. Thuyet, D.Q., Jorgenson, B.C., Wissel-Tyson, C., Watanabe, H., Young, T.M., 2012a. Wash off of imidacloprid and fipronil from turf and concrete surfaces using simulated rainfall. Science of the Total Environment 414, 515–524.

D.Q. Thuyet et al. / Agriculture, Ecosystems and Environment 179 (2013) 69–77 Thuyet, D.Q., Watanabe, H., Motobayashi, T., 2011a. Effect of formulations and treatment methods of nursery boxes applied with insecticide on the behavior of imidacloprid in rice paddy fields. Journal of Pest Science 36, 9–15. Thuyet, D.Q., Watanabe, H., Takagi, K., Yamazaki, K., Nhung, D.T.T., 2012b. Behavior of nursery-box-applied imidacloprid in micro paddy lysimeter. Journal of Pest Science 37, 20–27. Thuyet, D.Q., Watanabe, H., Yamazaki, K., Takagi, K., 2011b. Photodegradation of imidacloprid and fipronil in rice–paddy water. Bulletin of Environmental Contamination and Toxicology 86, 548–553. Tomlin, C., 2006. The e-Pesticide Manual, 13th ed. The British Crop Protection Council Publication, Farnham, UK. Tourenq, C., Bennetts, R.E., Kowalski, H., Vialet, E., Lucchesi, J.L., Kayser, Y., Isenmann, P., 2001. Are ricefields a good alternative to natural marshes for waterbird communities in the Camargue, southern France. Biological Conservation 100, 335–343. Tsukano, Y., 1986. Transformations of selected pesticides in flooded rice-field soil—a review. Journal of Contaminant Hydrology 1, 47–63. Ueda, T., 2011. EXTEND 2010 biological research on the wildlife, fundamental research, summary of research findings and feasibility study. In: EXTEND 2010., pp. 1–40 (in Japanese).

77

Watanabe, H., Nguyen, M.H.T., Souphasay, K., Vu, S.H., Phong, T.K., Tournebize, J., Ishihara, S., 2007. Effect of water management practice on pesticide behavior in paddy water. Agricultural Water Management 88, 132–140. Wilby, A., Lan, L.P., Heong, K.L., Huyen, N.P.D., Quang, N.H., Minh, N.V., Thomas, M.B., 2006. Arthropod diversity and community structure in relation to land use in the Mekong delta, Vietnam. Ecosystems 9, 538–549. Ying, G.-G., Kookana, R., 2002. Laboratory and field studies on the degradation of fipronil in a soil. Soil Research 40, 1095–1102. Ying, G.-G., Kookana, R.S., 2001. Sorption of fipronil and its metabolites on soils from South Australia. Journal of Environmental Science and Health, Part B 36, 545–558. Zhang, J., Dong, N., Yu, H., Zhou, Y., Lu, Y., Geng, R., Yu, L., 2008. Bacteria diversity in paddy field soil by 16S rDNA-RFLP analysis in Ningxia. Biodiversity Science 16, 586–592. Zhao, X.H., Ning, Z.P., He, Y.P., Shen, J.L., Su, J.Y., Gao, C.F., Zhu, Y.C., 2011. Differential resistance and cross-resistance to three phenylpyrazole insecticides in the planthopper Nilaparvata lugens (Hemiptera: Delphacidae). Journal of Economic Entomology 104, 1364–1368.