Pesticide Residue Behavior and Risk Assessment in Celery after Se Nanoparticles Application

This study investigates pesticide levels in celery, and compares their degradation, dissipation, distribution, and dietary risk after spraying with selenium (Se) nanoparticles. Abamectin, imidacloprid, acetamiprid, thiamethoxam, and lambda-cyhalothrin were sprayed at 1.6, 6.8, 2.0, 1.0, and 0.7 g a.i. ha−1 followed by a 2 g·ha−1 Se nanoparticle application during the growing period. Thiamethoxam, abamectin, imidacloprid, lambda-cyhalothrin, and acetamiprid in celery degraded following a first order kinetic model after 2 g·ha−1 Se nanoparticles application. With the exception of acetamiprid, the half-lives of thiamethoxam, abamectin, imidacloprid, and lambda-cyhalothrin were reduced from 2.4, 0.5, 1.2, 4.2 days without Se nanoparticles application to 1.4, 0.2, 0.9, 3.7 days with the addition of Se nanoparticles (2 g·ha−1), respectively. The chronic dietary exposure risk probability (RQc) and the acute dietary exposure risk probability (RQa) of celery after Se nanoparticles application were within acceptable limits for consumption except for abamectin.


Introduction
Celery (Apium graveolens L.) is a plant in the apiaceae family, which is rich in phenolic and antioxidant compounds [1,2]. The characteristic odor and flavor of celery consists mainly of a number of volatile compounds (terpenes, phthalates, and aldehydes) [3]. Flavonoids (apigenin, chrysin, luteolin) are subgroups of flavonoids that have been shown to have antioxidant, anti-inflammatory, anti-allergic, neuroprotective, and cardioprotective effects [4]. This biennial herb, originating from Europe, is now cultivated and consumed all over the world. The stalks (including leaves and petioles) are the main edible part of celery [5]. In China, the cultivation area of celery is around 550,000 hectares, ranking highest in the world [6].
Pesticides are widely used to control pests and diseases in soils and field crops [7,8] such as celery, although more recently, wider public concern has been aroused by their human health harm and environmental risk. Neonicotinoid pesticides are now the most widely used insecticides in China, and their continued global use can lead to accumulation in soil, posing potential ecological risks given their long half-life [9]. Recently, nanoparticles have been shown to effectively reduce and mitigate residues in crops by accelerating their degradation [10].
Nanomaterials are very small particles with surface properties that may bond and transport toxic chemical contaminants [11]. Se nanoparticles are synthesized utilizing various chemical, physical, and biological methods [12] and are less toxic and more biocompatible than selenates and selenites [13]. Se nanoparticles have a beneficial impact by increasing the photosynthetic ability of plant pigments and antioxidative metabolism,

Field Trials and Sample Collection
Celery (Ventura variety, supplied by Beijing Genqing Seed Co., Ltd., Beijing, China) was grown in the Xinjiang test facility greenhouse in 2020 with a temperature at around 10 • C. The soil type was ash desert soil, with a pH of 7.9, and organic matter content was 3.6%. The field experiments were designed to investigate the residue persistence of clothianidin, imidacloprid, abamectin and chlorantraniliprole in celery under greenhouse conditions in China. The field trials were designed in accordance with the Guideline on Pesticide Residue Trials [35]. Two trials were made to compare the degradation safety of pesticide residues in celery after applying different pesticide doses.
For the first trial, clothianidin, imidacloprid, abamectin, and chlorantraniliprole granules were applied onto the soil at the celery roots at the recommended dose and 1.5-fold the recommended dose. Pesticide labeling information and recommended dose was shown in Table S1. Representative samples of the celery were collected randomly at 14, 21, 28, and 50 days after pesticide application to analyze for clothianidin. Representative samples of the celery were collected randomly at 14, 21, 28, 60, and 77 days after pesticide application to analyze for imidacloprid and abamectin. Chlorantraniliprole was applied 30 days before celery was harvested. Representative samples of celery were collected randomly at 2 h and 1, 3, 5, 7 and 14 days after spraying with chlorantraniliprole.
For the second trial, representative samples of celery were collected randomly at 2 h and 1, 3, 5, 7 and 14 days after the spraying with thiamethoxam, abamectin, imidacloprid, lambda-cyhalothrin and acetamiprid. Table S2 shows pesticide labeling information and recommended dose combined with 2 g·ha −1 Se nanoparticles. After pesticide spray application, the celery also received a spray application of 2 g·ha −1 Se nanoparticles for each experimental plot, while control plots (with no Se nanoparticles application) were also included. The area of each plot was 20 m 2 , with three replicates. A buffer zone was utilized between the different treatment varieties to avoid cross-contamination. At least 1 kg of celery was harvested from each plot, homogenized and maintained at −20 • C before further analysis.

Method Validation
The performance of the proposed analytical method was verified on celery using the following parameters: linearity, limit of quantification (LOQs), accuracy, and precision. This method validation study was undertaken by comparing matrix-matched standard curves with celery matrix extracts. The LOQ is the lowest spiked level of the validation meeting these method performance acceptability criteria. The linearity of pesticide residue analysis method should be above 0.99, the recovery rate of pesticide should be between 70 and 120%, and the RSD should be between 0 and 20%. The accuracy and precision of the analytical method was measured from recovery tests with 3 spiked levels of pesticides (0.01, 0.05 and 0.1 mg kg −1 ) in 5 replicates. The concentration of pesticide residue was achieved with acceptable accuracy by the application of a complete analytical method in accordance with the European criteria guidelines established by SANTE/12682/2019 [36].

Extraction and Purification
Up to three kilograms of celery was collected in an S-shaped pattern across each plot, then homogenized to a pulp with a blender and stored at −18 • C until analysis. Soil samples were also collected, sieved to 1 mm and stored at −18 • C. Fifteen grams of each celery sample were weighed into a 50 mL plastic centrifuge tube, and 30 mL acetonitrile and 5-7 g sodium chloride were added [29]. The tube contents were homogenized for 1 min using a high-speed tissue homogenizer at 16,099× g and then centrifuged for 1 min using a high-speed centrifuge at 11,180× g until the acetonitrile phase and the aqueous phase separated. The pretreatment extraction method for pesticide residues in soil samples was the same as that of the celery extraction. According to the different analytical instruments or detectors required to analyze each pesticide, further processing was as follows: (1) 0.5 mL of the upper acetonitrile extract was added into a solution of 0.5 mL methanol and water (1:1). The mixture was eddy mixed and filtered through a 0.22 µm organic microporous filtration membrane for LC-MS/MS.
(2) 10 mL of the supernatant was reduced to near dryness for purification in a 150 mL round bottom flask in a water bath at 40 • C. Around 30 mL of acetonitrile and toluene (3:1) mix is used as an eluent solution for solid phase extraction. First, 5 mL of the eluent solution was used to pre-wash the graphite amino series-to-solid phase purification column. When the level of eluent solution reached the small column filter, about 1.5 mL of eluent solution was used to collect the residue from the round bottom flask bottle and then quickly transferred to the purification column. This step was repeated 3 times (each time, the new eluent was only transferred when the level of eluent solution reached the small column filter). The remaining eluent was poured into the round bottomed flask and gradually transferred to the purification column. After purification, the effluent (collected in another 150 mL round bottomed flask) was evaporated to near dryness in a 40 • C water bath. 2 mL of n-hexane was added, and the solution was passed through a 0.22 µm organic microporous filtration membrane before GC-MS/MS. Lambda-cyhalothrin was analyzed by GC-MS/MS, the remaining six pesticides were analyzed by UPLC-MS/MS.

LC-MS/MS Conditions
Pesticide extracts from celery and soil samples were analyzed using UPLC Xevo TQ-S micro (Waters, Milford, MA, USA). Chromatographic column: BEH C18 column 100 A (50 mm× 1.7 mm, particle size: 5 µm), Waters Corporation, USA; column temperature was 40 • C, and the liquid chromatography mobile phase and gradient elution conditions are shown in Table 1. Sampling volume: 10 uL; Ion source: ESI, scanning mode: positive and negative ion scanning; ion source temperature was 150 • C, desolventing temperature was 350 • C, N 2 gas flow rate was: 650 L/h, N 2 conical gas flow rate was 250 L/h, and the mobile phase flow rate was 0.20 mL/min. The mass spectrometry parameters for 6 target compounds, such as qualitative ion pair, quantitative ion pair, collision energy (Ce), and the conical voltage, are shown in Table S3. Ion source: electron bombardment ion source (EI), temperature 230 • C; Four-stage rod temperature: 150 • C; Emission current 50 µA; Masshunter workstation was used for instrument control and data processing. Scanning method: multiple reaction monitoring (MRM). The parameters of lambda-cyhalothrin mass spectrum are shown in Table S3.

Dissipation Kinetics of Pesticides
The pesticide dissipation dynamics of celery for the two treatments were evaluated in this study. Residue dissipation behaviors of the 7 pesticides were analyzed using a first order kinetic equation as follows [37,38]: In this equation, C t is the residual concentration of pesticide at t (day), C 0 (mg·kg −1 ) is the initial concentration of pesticide, and k is the dissipation rate constant. The t 1/2 is pesticide half-life of pesticide dissipation.

Risk Assessment
To verify the limits for safe application of clothianidin, imidacloprid, abamectin, chlorantraniliprole, thiamethoxam, lambda-cyhalothrin, and acetamiprid, the health risk was quantified based on RQc [39] for chronic dietary exposure risk probability and RQa for acute dietary exposure risk probability, respectively.
Here, NEDI (mg·kg −1 bw) is defined by the national estimated daily intake, STMR i is the supervised trial median residue and F i is the dietary reference intake of a specific food for the general population [40]. ADI (mg·kg −1 bw) was the acceptable daily intake and ARfD (mg·kg −1 bw) was acute reference dose. bw refers to the mean body weight of a person in China (63 kg).

Method Validation
According to methods approved by the Ministry of Agriculture and Rural Affairs of China (GB/T 20769-2008), clothianidine and imidacloprid residues in celery and soil, and thiamethoxam and acetamiprid residues in celery are only a few of the 450 pesticides and related chemical residues that are routinely analyzed in fruits and vegetables using LC-MS/MS. Abamectin residues [31] in celery and soil, and chlorantraniliprole residues [41] in celery were determined from literature methods. Lambda-cyhalothrin was analyzed in celery according to GC-MS/MS methods [42] (GB 23200.113-2018, Ministry of Agriculture and Rural Affairs of China) which can determine 208 pesticide residues and their metabolites in plant origin foods.
The analytical methods were verified by linearity, limit of quantification (LOQ), accuracy, and precision. Linearity was determined using a standard solution with a mix of five pesticide concentrations (0.01 to 5.0 mg·kg −1 ). Linearity calibration curves were constructed between the integrated chromatographic peak area and the corresponding concentration. All standard curves had good linear relationship, and the correlation coefficients (R 2 ) were ≥0.99.
The field trial pesticides were added to the standard matrix to confirm the extraction efficiency [43,44]. In order to evaluate accuracy and precision, recovery experiments were performed. Blank samples were replicated 5 times with 3 spiked pesticide concentration levels (0.01 to 0.1 mg·kg −1 ). Before extraction, 3 concentrations of clothianidam, imidacloprid, thiamethoxam, acetamiprid, chlorantraniliprole, and abamectin were added into a homogenized celery sample to obtain a spiked sample. A spiked celery sample with lambda-cyhalothrin was analyzed separately. The precision of the analysis method was measured by the relative standard deviation (RSD) [40]. The clothianidin, imidacloprid, thiamethoxam, and acetamiprid pesticide recovery levels from celery were 72.0 to 80.0%, 97.4 to 101.2%, 93.6 to 105.2%, and 80.2 to 90.3% at each of the three spiked concentration levels, respectively. The abamectin, chlorantraniliprole, and lambda-cyhalothrin pesticide recoveries from celery were 65.5 to 69.8%, 94.1 to 115.8% and 92.3 to 105.1%, respectively ( Table 2). The clothianidin, imidacloprid and abamectin pesticide recoveries from soil were 79.2 to 83.1%, 85.2 to 106.2% and 68.1 to 70.2%, respectively ( Table 3). The results showed that, except for the low recovery rate of abamectin, the above methods met the requirements of Guide to Pesticide Residue Tests in China [45] (NY/T 788-2018).  The results showed that the limits of quantification for thiamethoxam, imidacloprid, clothiamidam, acetamiprid, lambda-cyhalothrin, abamectin, and chlorantraniliprole were 0.01 mg·kg −1 . The analytical method showed good sensitivity, accuracy, and precision, and was found to be suitable for the determination of clothiamidam, imidacloprid, thiamethoxam, acetamiprid, chloranterianamide, abamectin, and lambda-cyhalothrin in celery and soil.

Residue Levels in Celery
The limit of quantitation (LOQ) was assigned as 0.01 mg kg −1 . A maximum residue limit (MRL) for clothianidin, imidacloprid, abamectin, and chlorantraniliprole in celery was set at 0.04, 6, 0.03, 7 mg kg −1 according to Codex regulations (http://www.fao. org/fao-who-codexalimentarius/codex-texts/dbs/pestres/pesticides/pt/ (accessed on 5 April 2021)) and 0.01, 0.5, 0.01, 0.06 mg kg −1 for EU regulations (https://ec.europa. eu/food/plant/pesticides/eu-pesticides-database/mrls/?event=search.pr (accessed on 5 April 2021)).Clothianidin and chlorantraniliprole are presently unregistered for use on celery and leafy vegetables, so a further objective of the study was to verify the safe use of these two pesticides for celery. Table 4 shows the initial and final residue levels of clothianidam, imidacloprid, abamectin, and chlorantraniliprole at the recommended dose and 1.5-fold the recommended dose. The initial residue content of clothianidin was 0.108 mg kg −1 at 1.5-fold the recommended dose of 23.6 g a.i. ha −1 , which was higher than 0.06 mg kg −1 at the recommended dosage of 16.8 g a.i. ha −1 . The final residue level of clothianidin was less than LOQ at 1.5-fold the recommended dose of 23.6 g a.i. ha −1 , which was lower than 0.013 mg kg −1 at the recommended dose of 16.8 g a.i. ha −1 . The initial residue level of imidacloprid was 0.029 mg kg −1 at 1.5-fold the recommended dose of 84.4 g a.i. ha −1 , which was higher than 0.007 mg kg −1 at the recommended dose of 56.2 g a.i. ha −1 . However, the final residue level of imidacloprid was 0.404 mg kg −1 at the recommended dose of 56.2 g a.i. ha −1 , which was higher than 0.246 mg kg −1 at 1.5-fold the recommended dose of 84.4 g a.i. ha −1 , the specific reason for this anomaly is yet unknown. The initial residue level of abamectin was 0.033 mg kg −1 at 1.5-fold the recommended dose of 19.7 g a.i. ha −1 , which was higher than 0.021 mg kg −1 at the recommended dose of 16.9 g a.i. ha −1 . The initial residue level of chlorantraniliprole was 4.958 mg kg −1 at 1.5-fold the recommended dose of 10.1 g a.i. ha −1 , which was higher than 2.701 mg kg −1 at the recommended dose of 6.8 g a.i. ha −1 . The final residue level of clothianidin was 1.847 mg kg −1 at 1.5-fold the recommended dose of 10.1 g a.i. ha −1 , which was higher than 1.784 mg kg −1 at the recommended dose of 6.8 g a.i. ha −1 . All celery with 1.5-fold the recommended dose of clothianidin, abamectin and chlorantraniliprole had higher residue levels than at the recommended dose that except for imidacloprid, for both the initial and final residues in celery. PHI: the pre-harvest interval. "*" represents a temporary limit value. GR = granules, SC = Suspension concentrate. Different letters across treatments indicate significant differences at p < 0.05.
Foods 2021, 10,1987 9 of 20 Table 5 shows the initial and final residues of thiamethoxam, abamectin, imidacloprid, acetamiprid, and lambda-cyhalothrin applied to celery at the recommended dose (control treatment) and the same pesticides combined with an application of 2 g·ha −1 Se nanoparticles (Se nanoparticles treatment). After the application of Se nanoparticles, the initial residue levels of thiamethoxam, abamectin, imidacloprid and acetamiprid of celery collected after 2 h were found to be lower than the Chinese maximum residue limit (MRL), although lambda-cyhalothrin was higher than the Chinese maximum residue limit (MRL) in both treatments. The application of Se nanoparticles to celery significantly reduced the residue levels of imidacloprid and lambda-cyhalothrin after 14 days. Se nanoparticles had no significant effect on residue digestion of acetamiprid after 7 days. With the exception of the control treatment lambda-cyhalothrin, the final residue levels of the other pesticides in Se nanoparticles treated celery were below the maximum residue limits after 14 set by China, Codex Alimentarius Commission and the European Union. The recommended doses for thiamethoxam, abamectin, imidacloprid, lambda-cyhalothrin and acetamiprid were 1.0 g a.i. ha −1 , 1.6 g a.i. ha −1 , 6.8 g a.i. ha −1 , 0.7 g a.i. ha −1 , and 2.0 g a.i. ha −1 , respectively. With the exception of acetamiprid, the original deposition amount (at 2 h) of thiamethoxam, abamectin, imidacloprid, and lambda-cyhalothrin in celery treated with Se nanoparticles were higher than the control treatment. The original deposition amount (at 2 h) of thiamethoxam, abamectin, imidacloprid and lambda-cyhalothrin in celery treated by Se nanoparticles were 0.131 mg kg −1 , 0.351 mg kg −1 , 4.613 mg kg −1 and 0.709 mg kg −1 , respectively, which were higher than 0.101 mg kg −1 , 0.219 mg kg −1 , 4.446 mg kg −1 , 0.687 mg kg −1 in the control treatment. The original deposition amount of acetamiprid in celery treated with Se nanoparticle was the same as in the control treatment. A possible reason is that the application of Se nanoparticles changed the celery's surface texture and increased the adhesion of thiamethoxam, abamectin, imidacloprid, and lambda-cyhalothrin.
Apart from acetamiprid, the application of Se nanoparticles to celery shortened the residue half-lives of thiamethoxam, abamectin, imidacloprid and lambda-cyhalothrin. Under greenhouse conditions, the residue levels of celery samples at both recommended dose and 1.5-fold the recommended dose after 7 days were lower than the Chinese MRL. According to residue monitoring throughout the field experiment, safe harvest of celery was possible only 22 days and 31 days after the application of clothianidin granules and chlorantraniliprole spray at 1.5-fold the recommended dose, respectively. These more persistent individual residue behaviors may arise from different external conditions such as crop species or climate conditions (temperature, viscosity) at spraying time [40]. Pre-harvest withholding intervals are usually set at seven days to ensure that pesticide residues are below the MRL [48]. However, half-lives of lambda-cyhalothrin and thiamethoxam have previously been noted from 7.01 d to 17.3 d [30], the half-life of abamectin was found to be 1.02 days in strawberry [49]. The dissipation rates of acetamiprid and lambda-cyhalothrin have previously been shown to be a first-order rate kinetic equation, with the half-lives of these chemicals in pomegranate fruits from 9.2 to 13 days and 13.5 to 17 days in the leaves [50].  Table 8. According to the principle of maximum dietary risk, chronic dietary risk assessment was conducted for MRLs of celery. The ADI and ARfD were obtained from the MRLs of Pesticides in Food [51] (GB 2763-2019) and the JMPR report, respectively. When calculating the NEDI value corresponding to each food group, the MRLs referred to Chinese standards. The STMR of imidacloprid, abamectin, clothianidin, and chlorantraniliprole in celery were used as the reference MRL's for national chronic dietary risk assessment where celery belongs to the dark vegetable group. STMR was used to calculate the NEDI of clothianidin, imidacloprid, abamectin and chlorantraniliprole at the recommended dose and 1.5-fold the recommended dose. Abamectin and its isomeric 8, 9-Z avermectin B 1 a, a photolytic product of avermectin B 1 a, were defined as chemical residues in dietary risk assessment [52]. As illustrated in Table 8, the dietary risk of celery was acceptable for consumers apart from celery treated with abamectin granules. The application of clothianidam, imidacloprid, and chlorantraniliprole at the recommended dose and 1.5-fold the recommended does not raise a dietary risk to ordinary Chinese consumers. The chronic dietary risk assessment of thiamethoxam, acetamiprid, imidacloprid, abamectin, and lambda-cyhalothrin in celery after the recommended dose and spraying with 2 g·ha −1 Se nanoparticles is shown in Table 9. STMR was used to calculate the recommended dose and NEDI of thiamethoxam, acetamiprid, imidacloprid, abamectin, and lambda-cyhalothrin after spraying with 2 g·ha −1 Se nanoparticles. As illustrated in Table 9, the dietary risk for celery was acceptable for consumers, apart from abamectin emulsion sprays. Abamectin is considered to be a dangerous pesticide with a variety of cytotoxic and genotoxic effects on non-target organisms [53]. The rate of dissipation in abamectin was previously noted to be 6 days after application on squash fruits [47]. Table 9. Risk quotient and risk probability of 5 pesticides in celery after pesticide application at the recommended dose combined with 2 g·ha −1 Se nanoparticles. This study also assessed the risk of acute dietary exposure to imidacloprid, abamectin, acetamiprid, thiamethoxam, and lambda-cyhalothrin in celery. According to the JMPR report, the ARfD values of imidacloprid, acetamiprid, clothianidin, thiamethoxam, and lambda-cyhalothrin were 0.4 mg·kg −1 bw, 0.1 mg·kg −1 bw, 0.6 mg·kg −1 bw, 1.0 mg·kg −1 bw, 0.02 mg·kg −1 bw, respectively. ARfD of abamectin and chlorantraniliprole was not calculated for these two pesticides. The highest residues (HR) for clothianidin and imidacloprid in celery for these experimental trials were 0.014 mg·kg −1 , 0.032 mg·kg −1 , 2.81 mg·kg −1 , and 4.06 mg·kg −1 at the recommended dose and 1.5-fold the recommended dose, respectively. The RQa of clothianidin, imidacloprid, and chlorantraniliprole in celery were all less than 100%. The HR of thiamethoxam, acetamiprid, imidacloprid, and lambda-cyhalothrin were 0.101 mg·kg −1 , 0.131 mg·kg −1 , 0.042 mg·kg −1 , 0.040 mg·kg −1 , 4.446 mg·kg −1 , 4.613 mg·kg −1 , 0.687 mg·kg −1 , and 0.709 mg·kg −1 after spraying with 2 g·ha −1 Se nanoparticles, respectively. The RQa of thiamethoxam, acetamiprid, imidacloprid, and lambda-cyhalothrin in celery were all less than 100%, respectively. These results suggest there is a negligible risk of acute dietary exposure to clothiamethoxam, imidacloprid, thiamethoxam, acetamiprid, and lambda-cyhalothrin at these application rates.

Conclusions
GC-MS/MS and UPLC-MS/MS were used to measure the recovery, precision, accuracy and degradation behavior of lambda-cyhalothrin, clothianidin, imidacloprid, abamectin, chlorantraniliprole, thiamethoxam, and acetamiprid residues in celery. Results showed that, apart from acetamiprid, the half-lives of thiamethoxam, abamectin, imidacloprid and lambda-cyhalothrin were 2.4, 0.5, 1.2, 4.2 days and 1.4, 0.2, 0.9, and 3.7 days for the control and a further Se nanoparticles treatment (2 g·ha −1 ), respectively. According to STMR, chronic and acute dietary exposure risks were assessed for clothianidin, imidacloprid, abamectin, chlorantraniliprole, thiamethoxam, lambda-cyhalothrin, and acetamiprid in celery. The dietary risk of celery was acceptable for consumers apart from abamectin, whether emulsion sprays or granules were used. This study provides a valuable reference for the safe and reasonable application of recommended doses of imidacloprid, abamectin, thiamethoxam, and lambda-cyhalothrin for celery cultivation in China.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/10 .3390/foods10091987/s1, Table S1: Pesticide labeling information and recommended dose, Table S2: Pesticide labeling information and recommended dose combined with 2 g·ha −1 Se nanoparticles, Table S3: Instrument parameters and their physicochemical properties.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.