Residues and Uptake of Soil-Applied Dinotefuran by Lettuce (Lactuca sativa L.) and Celery (Apium graveolens L.)

Pesticides have been used for decades to protect agricultural products and increase productivity by controlling crop pests. However, the frequent application of pesticides on crops or soil leads to the accumulation of their residues in the environment, which will be subsequently absorbed by plants and finally translocated to edible parts. This study aimed to investigate the uptake of soil-applied dinotefuran and three major metabolites by lettuce and celery from the previous season’s applications and analyze their residues in soil and consumable parts. Dinotefuran was soil-applied at plant back intervals of 30 and 60 days (PBI-30 and PBI-60). Residues of dinotefuran and its metabolites in 50% and 100% mature lettuce leaves and celery shoots, soil after application, at planting, and at 50% and 100% plant maturity were estimated. Half-lives and bio-concentration factors were calculated. The uptake of dinotefuran by lettuce and celery ranged between 23.8% and 28% and between 51.73% and 53.06%, respectively. Respective half-lives (days) of dinotefuran applied on PBI-30 and PBI-60 were 1.33–1.54 and 0.91–2.16 in lettuce soil and 0.9–1.47 and 0.79–1.65 in celery soil. Residues were below Korean MRLs in PBI-60 and most PBI-30 samples. The calculated risk assessment parameters indicated that negligible risk could be expected. The current study recommends growing the next crop 60 days after harvesting the first crop, but not less than 30 days.


Introduction
Pesticides have been used for decades to protect agricultural products and increase productivity by controlling crop pests. Despite these merits, their residues present harm to the environment [1]. The frequent application of pesticides on crops or soil leads to the accumulation of their residues in the environment, which will be subsequently absorbed by plants and finally translocated to edible parts [2,3]. Investigation of the degree of crop uptake and movement of residues in agricultural soil and grown crops is needed to ensure the safety of agricultural products [4]. Uptake and translocation of pesticides in plants are affected by many factors, including plant species, characteristic of the compound, content of soil organic matter, and temperature [5,6].
Neonicotinoids are systemic insecticides, particularly nitroguanidine compounds, which are widely used in controlling insect pests. A widely used nitroguanidine compound is dinotefuran, developed by Mitsui Chemicals, Inc., Tokyo, Japan, in 2002 and first registered by the U.S. EPA in 2004 to control insect pests [7,8]. This compound has broad spectrum activity and is used to control a wide range of sucking insect pests. It is highly toxic to most pollinators. Dinotefuran can be applied on foliage, soil, nurseries, as well to paddy water through spraying, drenching, broadcasting, and pricking-in-hole treatments [7,9]. It is also applied as a systemic insecticide via trunk injection or infusion or soil injection or drenches [8,10]. The applied neonicotinoids, whether in soil or crops, can be absorbed and taken up by plant roots or leaves and distributed acropetally and

Soil Treatments
The experiments were designed following the OECD method [24] for rotational crops. The insecticide was prepared by thoroughly mixing with soil at a rate of 90 g/6 kg control soil (dinotefuran free) and spread over the subplots, followed by immediate irrigation to soak the dinotefuran into the soil. Then, after 30 and 60 days, the plants were grown.

Collection of Plant Samples
Sampling of the two crops was performed at 50% and 100% maturity. This corresponds to 21 and 28 days from seeding in lettuce and 45 and 75 days in celery. Control sets were harvested from untreated plots at the same periods. The plants were harvested using the method described by the Korean Rural Development Administration (RDA) [25] and transferred to the Bio-regulatory Chemical Laboratory, Kangwon National University for residue analysis. The collected samples were prepared and kept cool at −18 • C for residue analysis.
We confirm that all plant samples used in the current work comply with the IUCN Policy Statement on Research Involving Species at Risk of Extinction and the Convention on the Trade in Endangered Species of Wild Fauna and Flora.

Soil Sampling and Preparation
The first soil samples in both 60-PBI and 30-PBI plots were collected 2 h after application and before planting. Other soil samples were taken at planting and at 50% and 100% maturity (i.e., 21 and 28 days in lettuce soil and 45 and 75 days for celery soil). The untreated control plots were sampled at the same time. Sampling was carried out following the grid sampling method described by the Korean RDA [25]. The collected samples were immediately transferred to the Bio-regulatory Chemical Laboratory, Kangwon National University for residue analysis. Sub-samples were transferred to the RDA for soil physiochemical analysis following the method described by RDA [26] (Tables S2).  shaking for 15 min at 1400 g-force after adding the ceramic homogenizer. The modified original QuEChERS method was followed [27,28]. The QuEChERS kit (4.0 g magnesium sulfate, 1.0 g sodium chloride, 1.0 g sodium citrate tribasic dihydrate, 0.5 g sodium citrate dibasic sesquihydrate) was added to the extract, shaken for 2 min at 504 g-force, and centrifuged at 4 • C and 3584 g-force for 10 min. The supernatant was filtered at this step for MNG analysis. However, for dinotefuran, UF, and DN analysis, 1 ml of the supernatant was taken and cleaned up by adding it to the d-SPE kit (MgSO 4 150 mg, PSA 50 mg) and centrifuged at 7168 g-force for 10 min. The supernatant was filtered with a hydrophilic 0.20-nanometer filter. Finally, 500 µL of filtrate was taken, and then 400 µL of acetonitrile and 100 µL of methanol were added for the matrix-matching before instrumentation analysis.

Extraction and Cleanup of Dinotefuran, MNG, UF, and DN Residues from the Soil
Upon arrival to the laboratory, the samples were left to dry at room temperature (27.5 • C), crushed into small particles, and sieved through a 2 mm sieve. The extraction was performed by adding 10 mL of distilled water to 10 g of soil samples in a polyethylene tube and shaking after adding the ceramic homogenizer for 15 min at 1400 g-force (for dinotefuran), and for 30 s in a vortex mixer for the metabolites MNG, UF, and DN. Then, 20 mL of acidified acetonitrile (with 1% formic acid) was added and the mixture was shaken for 2 min at 504 g-force. However, for the metabolites, the acetonitrile was acidified with 2% formic acid and shaken for 30 min at 1400 g-force. Then, the QuEChERS kit was added to the tube containing the extract, shaken for two minutes at 504 g-force, and then centrifuged at 4 • C and 3584 g-force for 10 min. For MNG, the supernatant was filtered through a hydrophilic 0.20-nanometer filter, while for dinotefuran, UF, and DN, a further 1.0 mL of organic supernatant was taken and cleaned up by adding it to the d-SPE kit and centrifuged at 14336 g-force for 10.0 min. The supernatant was filtered through a hydrophilic 0.20-nanometer filter and 500 µL of the filtrate was taken, to which 400 µL acetonitrile and 100 µL methanol were added for the matrix-matching before analysis.

Method Validation
To verify the suitability of the analytical method, several parameters were tested, such as linearity of calibration curve, method limits of quantification (MLOQ), accuracy, and precision. The linearity of the six-point calibration curve over the range of 0.003 to 0.2 mg/kg was excellent, with a regression coefficient (R 2 ) of 0.99 for all compounds. The recovery rates of dinotefuran, MNG, UF, and DN at 0.01 and 0.1 mg/kg fortification levels ranged from 81.7% to 107.7%, 77.6% to 113.1%, 72.2% to 77.9%, and 78.5% to 109.0% respectively, in plant samples. On the other hand, the respective values in soil samples ranged from 90.7% to 100.9%, 78.1% to 113.1, 74.7% to 92.5%, and 87.2% to 109.4% (Table S3 and Figures S1-S3). The repeatability expressed as the relative standard deviation (RSD) of the analyzed samples (n = 4) was less than 10% for all analytes. The MLOQ was 0.01 mg/kg, calculated using the following equation:

LC-MS/MS Analysis
High-performance liquid chromatography (LC-20AXR series, Shimadzu, Japan) coupled with mass spectrometry (TSQ Quantum Ultra, Thermo Science, USA) was used for analysis of the samples. The column used was Kinetex 2.6 µm Polar C18 100 A (2.1 mm ID × 100 mm, 2.6 µm, Torrance, CA, USA). The mobile phase flow at volumes given in Table S3 under gradient conditions used (A) 5 mM ammonium formate in water and (B) 5 mM ammonium formate in methanol, to which 0.1% formic acid was added. The injection volume was 1.0 µL. MS/MS was analyzed by the multiple reaction monitoring (MRM) method using the electrospray ionization (ESI) positive mode. Spray voltage positive polarities were 3.8 kV (for dinotefuran and MNG) and 4 kV (for FU and DN). Capillary temperatures were 300 • C (for dinotefuran, FU, and DN) and 280 • C (for MNG). Sheath gas pressures were 45 units for dinotefuran, 20 units for MNG, and 30.0 units for FU and DN. Auxiliary gas pressures were 15 units for dinotefuran and 10.0 units for MNG, FU, and DN (Table S3).

Standard Calibration Curve
Stock solutions of analytical standards were prepared by accurately placing 10.18, 10.0, 10.0, and 10.15 mg of dinotefuran (98.7% pure), MNG (100.0% pure), UF (99.8% pure), and DN (98.5% pure), respectively, in 100 mL volumetric flasks, and then topping them up to the mark with methanol (99.9% pure). Different concentrations (0.02, 0.05, 0.1, 0.2, 0.5, and 1.0 mg/L) of dinotefuran and its metabolites (MNG, UF, and DN) were prepared by serial dilutions from the stock solution. Then, 1 µL aliquot of each concentration was injected into the LC/MS/MS, and the response was used for the construction of the standard calibration curve. The matrix-matching effect for dinotefuran was realized by taking 1.0 mL of dinotefuran working standard, drying under a gentle stream of nitrogen, then reconstituting it in 1.0 mL of untreated sample (control) and diluting it 10 times before analysis. The matrix-matching effect for the metabolites MGN, UF, and DN was achieved by adding 100 µL of their working standard solution to 500 µL of untreated sample and 400 µL of acetonitrile. The matrix effect (%) was calculated using the equation, Matrix effect (%) = Slope of calibration curve matrix Slope of calibration curve solvent − 1 × 100 (2)

The Calculation of Dinotefuran Total Residues
The total residues in the test samples were calculated as follows: Total residue of dinotefuran (mg/kg) = dinotefuran + (MNG residue ×1.71) + (UF residue × 1.28) + (DN residue × 1.29). The conversion factors were calculated as follows:

Half-Life Calculation
The half-lives of dinotefuran in soil were calculated using the first-order kinetic equation:

Storage Stability
The samples were spiked with dinotefuran, MNG, UF, and DN at a concentration of 0.01 and 0.1 mg/kg for lettuce and celery samples, and 0.01, 0.1, and 1.0 mg/kg for soil samples. The storage stability of the dinotefuran and its metabolites were conducted under cool storage at −20 • C for 7 months (Table S4).

Bio-Concentration Factor
The bio-concentration factor was calculated according to the method described by Yuan et al. [29] and Mckone and Maddalena [30] as follows: where BCF: bio-concentration factor; CFC: concentration in the fresh crop; ICS: initial concentration in the soil at the seeding time.

Risk Assessment
In this study, human exposure to dinotefuran was estimated to assess chronic and acute health risks according to the following equations [31].
Chronic Health Risk Assessment EDI = (C×FI)/BW RQc = EDI/ADI where EDI is the estimated daily intake of dinotefuran from lettuce and celery (mg/kg/d), C is the median residual concentration of dinotefuran in lettuce and celery (mg/kg), FI is the daily intake of lettuce and celery (g/day), BW is the average body weight (60 kg) for the Korean population [32], ADI is the acceptable daily intake, and RQc is the chronic exposure risk. The potential risk of dinotefuran was evaluated according to a commonly used risk ranking criterion. For RQc values < 1, low or negligible risk can be expected; RQc values higher than 1 indicate higher risk.

Acute Health Risk Assessment
where ESTI is the estimated short-term dietary intake of dinotefuran via lettuce and celery, HR is the highest residue of dinotefuran (mg/kg) in lettuce and celery samples, LP is the large portion, referring to the highest daily amount of lettuce and celery intake (kg/day), RQa is the acute risk quotient of dinotefuran via consumption of lettuce and celery, and ARD is the acute reference dose of dinotefuran (mg/kg/day, bw). RQa values <1 are considered acceptable and do not pose a health risk in the short-term exposure; RQa values >1 can create an unacceptable risk.

Results
The current work investigated the soil residues, dissipation, and uptake of soil-applied dinotefuran by two common rotational minor crops (lettuce and celery) grown in two important production sites (Pyeongtaek and Chuncheon). The results are summarized in the following sub-sections.

Soil Characteristics
Soil characteristics of two rotational crops (lettuce and celery) in the two experimental sites (Pyeongtaek and Chuncheon) were studied for proper interpretation of the dissipation studies of dinotefuran. The results of soil physiochemical analysis revealed that both soils of the two studied areas were sandy. The Pyeongtaek soil was characterized by a pH value of 5.5, cation exchange capacity (CEC) of 13.31 cmolc/kg, and an organic matter percentage of 2.63. In contrast, Chuncheon soil has a pH value of 6.5, CEC of 18.49 cmolc/kg, and an organic matter percentage of 5.16 (Table S2).

Soil Residues and Dissipation of Dinotefuran in the Soils of Lettuce and Celery
The levels of dinotefuran and its major metabolite residues (mg/kg) remaining in soil of lettuce in Pyeongtaek and Chuncheon are given in Table 1A. The levels of dinotefuran and its major metabolite were below the detection limit (<0.01 mg/kg) in the control soils of both locations throughout all sampling dates, increased following the insecticide treatments, and later generally progressively decreased following the planting of the lettuce at the two locations (Table 1A), although the levels apparently decreased at faster rates in Chuncheon (66.29% to 84.2%) than Pyeongtaek (53.33% to 64.7%) ( Figure 1). Generally, there is a slight difference between the PBI-30 and PBI-60 residue levels in both locations. The pre-planting level in Chuncheon (0.68 mg/kg at PBI-30 and 0.60 mg/kg at PBI-60) is approximately 1.5 times higher than in Pyeongtaek (0.9 mg/kg at PBI-30 and 0.89 mg/kg at PBI-60).  On the other hand, the soil residues in the celery control soil (Table 1B) were similar to those of lettuce soil (<0.01 mg/kg) throughout all sampling dates at the two locations. The soil level increased following the insecticide treatments at both application dates at the two locations (3.18 and 2.14 mg/kg for the PBI-30 and PBI-60 in Pyeongtaek, and 0.48 and 0.61 mg/kg for the PBI-30 and PBI-60 in Chuncheon). The levels progressively decreased following various growth stages but at a faster rate in Pyeongtaek (88-95.5%) than in Chuncheon (54.0-81.9%) ( Figure 2). Generally, there is a slight difference between the PBI-30 and PBI-60 residue levels in both locations. The pre-planting level in Pyeongtaek (3.18 mg/kg at PBI-30 and 2.14 mg/kg at PBI-60) is approximately 6.63 times higher than in Chuncheon (0.48 mg/kg at PBI-30 and 0.61 mg/kg at PBI-60). Generally, the pre-planting residue levels in celery soils are far greater than those of lettuce soils (Table  1). On the other hand, the soil residues in the celery control soil (Table 1B) were similar to those of lettuce soil (<0.01 mg/kg) throughout all sampling dates at the two locations. The soil level increased following the insecticide treatments at both application dates at the two locations (3.18 and 2.14 mg/kg for the PBI-30 and PBI-60 in Pyeongtaek, and 0.48 and 0.61 mg/kg for the PBI-30 and PBI-60 in Chuncheon). The levels progressively decreased following various growth stages but at a faster rate in Pyeongtaek (88-95.5%) than in Chuncheon (54.0-81.9%) ( Figure 2). Generally, there is a slight difference between the PBI-30 and PBI-60 residue levels in both locations. The pre-planting level in Pyeongtaek (3.18 mg/kg at PBI-30 and 2.14 mg/kg at PBI-60) is approximately 6.63 times higher than in Chuncheon (0.48 mg/kg at PBI-30 and 0.61 mg/kg at PBI-60). Generally, the pre-planting residue levels in celery soils are far greater than those of lettuce soils (Table 1).   On the other hand, the soil residues in the celery control soil (Table 1B) were similar to those of lettuce soil (<0.01 mg/kg) throughout all sampling dates at the two locations. The soil level increased following the insecticide treatments at both application dates at the two locations (3.18 and 2.14 mg/kg for the PBI-30 and PBI-60 in Pyeongtaek, and 0.48 and 0.61 mg/kg for the PBI-30 and PBI-60 in Chuncheon). The levels progressively decreased following various growth stages but at a faster rate in Pyeongtaek (88-95.5%) than in Chuncheon (54.0-81.9%) ( Figure 2). Generally, there is a slight difference between the PBI-30 and PBI-60 residue levels in both locations. The pre-planting level in Pyeongtaek (3.18 mg/kg at PBI-30 and 2.14 mg/kg at PBI-60) is approximately 6.63 times higher than in Chuncheon (0.48 mg/kg at PBI-30 and 0.61 mg/kg at PBI-60). Generally, the pre-planting residue levels in celery soils are far greater than those of lettuce soils (Table  1).   The results of total residue (dinotefuran + main metabolites MNG, UF, and DN) levels of dinotefuran remaining in the lettuce soil are given in Table 2 and Figure 3. Since the levels of the three metabolites throughout most of the sampling dates were below the detection Agriculture 2022, 12, 1443 9 of 20 limit (<0.01 mg/kg), the total residue levels of dinotefuran follow a similar distribution pattern as those of dinotefuran alone, as described above (Table 1). There are significant differences in the residue levels throughout the various sampling dates ( Table 2). All levels were significantly different from the control (<0.01 mg/kg) at the two locations, and significantly increased following the insecticide treatments in both locations (0.68 and 0.90 mg/kg for the 30 days and 0.60 and 0.89 mg/kg for the 60 days in lettuce soil of Pyeongtaek and Chuncheon, respectively), and later generally progressively decreased following planting of the crops at the two locations (Table 2), although the levels apparently decreased at faster rates in Chuncheon (66.29% to 84.2%) than in Pyeongtaek (53.33% to 64.7%). Generally, there is slight difference between the PBI-30 and PBI-60 residue levels in both locations. The pre-planting level in Chuncheon (0.90 mg/kg at PBI-30 and 0.89 mg/kg at PBI-60) is approximately 1.5 times higher than in Pyeongtaek (0.60 mg/kg at PBI-30 and 0.68 mg/kg at PBI-60).  both locations and on the two application dates (3.18 and 2.14 mg/kg for the PBI-30 and PBI-60 in Pyeongtaek, and 0.48 mg/kg and 0.61 for the PBI-30 and PBI-60 in Chuncheon). The levels progressively decreased at the various growth stages but at a faster rate in Pyeongtaek (88-95.5%) than Chuncheon (54.0-81.9%) (Figure 2). Generally, there is slight difference between the PBI-30 and PBI-60 residue levels in both locations. The pre-planting residue level in Pyeongtaek (3.18 mg/kg at PBI-30 and 2.14 mg/kg at PBI-60) is approximately 6.63 times higher than in Chuncheon (0.48 mg/kg at PBI-30 and 0.61 mg/kg at PBI-60). Generally, the pre-planting residue levels in celery soils are far greater than those in lettuce soils (Table 2).   Figure 3. Dinotefuran residues (mg/kg) remaining in the soil at 50% and 100% plant maturity. The total soil residue distribution in celery control soil ( Table 2) is similar to that of lettuce soil (<0.01 mg/kg). The soil level increased following the insecticide treatment in both locations and on the two application dates (3.18 and 2.14 mg/kg for the PBI-30 and PBI-60 in Pyeongtaek, and 0.48 mg/kg and 0.61 for the PBI-30 and PBI-60 in Chuncheon). The levels progressively decreased at the various growth stages but at a faster rate in Pyeongtaek (88-95.5%) than Chuncheon (54.0-81.9%) (Figure 2). Generally, there is slight difference between the PBI-30 and PBI-60 residue levels in both locations. The pre-planting residue level in Pyeongtaek (3.18 mg/kg at PBI-30 and 2.14 mg/kg at PBI-60) is approximately 6.63 times higher than in Chuncheon (0.48 mg/kg at PBI-30 and 0.61 mg/kg at PBI-60). Generally, the pre-planting residue levels in celery soils are far greater than those in lettuce soils ( Table 2).
The half-lives of dinotefuran in Pyeongtaek lettuce soil were 1.54 days and 2.16 days in PBI-30 and PBI-60, and in Chuncheon soil, they were 1.33 days and 0.91 days for PBI-30 and PBI-60, respectively. In comparison, the half-lives of dinotefuran in Pyeongtaek celery soil were 0.79 and 0.78 days for PBI-30 and PBI-60, while in Chuncheon soil, they were 1.47 and 1.18 days for PBI-30 and PBI-60, respectively. The half-lives were found following the first-order kinetics model.
The metrological (daily temperature, RH%, and precipitation) factors had no effect on the dissipation and uptake of dinotefuran (Table S5).

Uptake of Dinotefuran and Its Major Metabolites Detected in Lettuce Leaves and Celery Shoots
Dinotefuran levels and residues of three major metabolites in lettuce leaves are presented in Table 3A. Residues of UF and DN in lettuce leaves were below the detection limits in both PBI-30 and PBI-60 treatments; MNG was the only metabolite detected at both maturity stages at the two locations. The MNG level in lettuce leaves of PBI-30 is almost constant, around 0.023 mg/Kg in both locations, while its level in PBI-60-treated plots is slightly higher and decreases as plants reach 100% maturity (Table 3A).
On the other hand, three metabolites (MNG, UF, and DN) were detected in celery shoots (Table 3A). The levels of the three metabolites decreased as celery approached 100% maturity in both PBI-30 and PBI-60 at the two locations. Generally, the levels of the PBI-30 are slightly higher than those of PBI-60, especially in Pyeongtaek, while in Chuncheon, the levels are almost similar in the two PBIs (30 and 60). The residue levels were relatively higher (14-55%) in Pyeongtaek than in Chuncheon; however, the levels later progressively decreased as plants approached 100% maturity at the two locations (Table 3).  The levels of dinotefuran in the control sets on both sampling dates in the two crops and at the two locations are below the detection limit (<0.01 mg/kg). There are no significant differences between the residue levels in lettuce leaves measured at the two maturity stages of PBI-30 at the Chuncheon site, and the levels significantly decrease at the 100% maturity stage of PBI-60 at the two locations (Table 4). On the other hand, there is a significant decrease in the residue levels in celery shoots at 100% maturity of both PBI-30 and PBI-60 at the two locations ( Table 4). The levels in celery shoots are 63.85% higher than those in lettuce leaves (Table 4 and Figure 4). Generally, levels of dinotefuran in PBI-60 are slightly higher than those in PBI-30, although the differences are not significant. The residue levels in lettuce leaves in Pyeongtaek are 14-55% higher than in Chuncheon, while those in celery shoots in Pyeongtaek are 140-150% higher than in Chuncheon.

Dinotefuran Metabolites Analysis
The results indicate the presence of three dinotefuran metabolites-MNG, UF, and DN-in lettuce leaves, celery shoots, and/or their soils ( Figure S4). MNG was the only metabolite detected in the celery soil at the Chuncheon site. All metabolites in Pyeongtaek soil were below the detection limits. MNG was the only metabolite detected at both sites in lettuce leaves at 0.030-0.037 mg/kg in Pyeongtaek and 0.020-0.027 in Chuncheon. However, in celery shoots, the three metabolites (MNG, UF, and DN) were detected at measurable levels at both sites. MNG residues detected in celery shoots were 0.013-0.046 mg/kg in Pyeongtaek and 0.010-0.023 mg/kg in Chuncheon. The UF residues ranged from 0.013 to 0.020 mg/kg in Pyeongtaek and from <0.01 to 0.020 mg/kg in Chuncheon. The respective values for DN residues ranged from 0.036 to 0.060 mg/kg and <0.01 to 0.050 mg/kg in Pyeongtaek and Chuncheon, respectively. Generally, the levels of the PBI-30 are slightly higher than those of PBI-60, especially at the Pyeongtaek site, while in Chuncheon, the levels are almost similar in the two PBIs (30 and 60). These levels were relatively higher and later generally progressively decreased following plant manuring (100%) at the two locations ( Table 3). The residue levels in Pyeongtaek were 14-55% higher than those at the Chuncheon site.

Storage Stability Studies of Dinotefuran and Its Major Metabolites
The results of storage stability are given in Table S4. The recoveries (%) of dinotefuran and its three major metabolites in plants samples after storage stability test ranged from 73.6 ± 3.3 to 105.7 ± 6.5 and from 72.6 ± 2.7 to 107.7 ± 3.3 for 0.01 and 0.1 mg/kg spiked samples, respectively. The recoveries (%) in spiked soil samples ranged from 74.7 ± 1.9 to 113.1 ± 1.0 for 0.01 mg/kg, from 72.1 ± 0.2 to 109.4 ± 2.6 for 0.1 mg/kg, and from 76.4 ± 1.7 to 99.4 ± 0.4 for 1.0 mg/kg.

Bio-Concentration Factor of Dinotefuran in Lettuce Leaves and Celery Shoots
The results of the bio-concentration test showed an average residue level of dinotefuran in 100% mature lettuce leaves of 0.217 mg/kg for PBI-30 and 0.216 mg/kg for PBI-60, while these values in celery shoots were 0.230 mg/kg and 0.297 mg/kg for PBI-30 and PBI-60, respectively. The percentage translocation of dinotefuran to the 100% mature lettuce leaves was 23.8 in PBI-30 and 28 in PBI-60. On the other hand, the respective translocation values of dinotefuran in celery shoots were 53.06% and 51.77%.

Dinotefuran Metabolites Analysis
The results indicate the presence of three dinotefuran metabolites-MNG, UF, and DN-in lettuce leaves, celery shoots, and/or their soils ( Figure S4). MNG was the only metabolite detected in the celery soil at the Chuncheon site. All metabolites in Pyeongtaek soil were below the detection limits. MNG was the only metabolite detected at both sites in lettuce leaves at 0.030-0.037 mg/kg in Pyeongtaek and 0.020-0.027 in Chuncheon. However, in celery shoots, the three metabolites (MNG, UF, and DN) were detected at measurable levels at both sites. MNG residues detected in celery shoots were 0.013-0.046 mg/kg in Pyeongtaek and 0.010-0.023 mg/kg in Chuncheon. The UF residues ranged from 0.013 to 0.020 mg/kg in Pyeongtaek and from <0.01 to 0.020 mg/kg in Chuncheon. The respective values for DN residues ranged from 0.036 to 0.060 mg/kg and <0.01 to 0.050 mg/kg in Pyeongtaek and Chuncheon, respectively. Generally, the levels of the PBI-30 are slightly higher than those of PBI-60, especially at the Pyeongtaek site, while in Chuncheon, the levels are almost similar in the two PBIs (30 and 60). These levels were relatively higher and later generally progressively decreased following plant manuring (100%) at the two locations ( Table 3). The residue levels in Pyeongtaek were 14-55% higher than those at the Chuncheon site.

Storage Stability Studies of Dinotefuran and Its Major Metabolites
The results of storage stability are given in Table S4. The recoveries (%) of dinotefuran and its three major metabolites in plants samples after storage stability test ranged from 73.6 ± 3.3 to 105.7 ± 6.5 and from 72.6 ± 2.7 to 107.7 ± 3.3 for 0.01 and 0.1 mg/kg spiked samples, respectively. The recoveries (%) in spiked soil samples ranged from 74.7 ± 1.9 to 113.1 ± 1.0 for 0.01 mg/kg, from 72.1 ± 0.2 to 109.4 ± 2.6 for 0.1 mg/kg, and from 76.4 ± 1.7 to 99.4 ± 0.4 for 1.0 mg/kg.

Bio-Concentration Factor of Dinotefuran in Lettuce Leaves and Celery Shoots
The results of the bio-concentration test showed an average residue level of dinotefuran in 100% mature lettuce leaves of 0.217 mg/kg for PBI-30 and 0.216 mg/kg for PBI-60, while these values in celery shoots were 0.230 mg/kg and 0.297 mg/kg for PBI-30 and PBI-60, respectively. The percentage translocation of dinotefuran to the 100% mature lettuce leaves was 23.8 in PBI-30 and 28 in PBI-60. On the other hand, the respective translocation values of dinotefuran in celery shoots were 53.06% and 51.77%.

Chronic Health Risk Assessment
In this study, the average lowest concentrations of dinotefuran in lettuce and celery (Table 4) were used for long-term consumer exposure. The daily intake of lettuce and celery in the Republic of Korea is 0.00719 and 0.0006 kg/day, respectively. The results showed that the EDI of dinotefuran was 0.00451 mg/kg/d for lettuce, and that for celery was 0.000057 mg/kg/d. The RQc values were 0.2255 for lettuce and 0.00285 for celery. RQc values for both plants are less than 1, which indicates an expected low or negligible risk.

Acute Health Risk Assessment
The highest concentrations were used to calculate the acute health risk assessment. The RQa values were 0.5752 and 0.02352 for lettuce and celery, respectively. RQa values were <1, which is considered acceptable and does not pose a health risk in short-term exposure.

Discussion
We investigated the residue distribution and uptake of dinotefuran in two common rotational minor crops (lettuce and celery) and at two important production sites (Pyeongtaek and Chuncheon) in South Korea. Residues of this insecticide and other pesticides may be translocated to subsequent crops from previous applications in the primary crop and thus may pose serious hazards to consumers of the next crop (secondary crop), especially if the latter is consumed fresh or uncooked, as leafy vegetables such as lettuce and celery often are. To understand the residual characteristic of the pesticides in the soil, the determination of their residues in certain soil is needed [29]. In the current study, the bare soil was treated with dinotefuran insecticide before crop cultivation (30 and 60 days before planting lettuce and celery) to ensure uptake and translocation of the candidate pesticide by the growing plants, to allow the estimation of metabolites, and predict the dissipation of the tested compound as outlined by the Organization for Economic Co-operation and Development (OECD) [24]. A dinotefuran level greater than 0.5 mg/kg was adopted in the current experiment as a violation level since RDA monitoring data of pesticide residues at 508 sites of agricultural soil found dinotefuran residues at 73 sites, and only two of them had residue level levels >0.5 mg/kg (RDA 2020 unpublished data).
The results indicated that the residue levels of dinotefuran were below the detection limit in the control soils of both locations (<0.01 mg/kg), increased following the insecticide treatments at both locations, and later generally progressively decreased following the planting of the crop at the two locations ( Table 2 and Figures 1-3). The current findings are consistent with Masahiro et al. [33]. Fantke and Juraske [34] and Kwak et al. [35] reported a declining tendency of dinotefuran and its metabolites in greenhouses and agricultural soils following the cultivation of plants. Such a decrease may be explained by plant uptake and/or degradation by various biological and environmental factors, as noted by previous authors [36][37][38] who attributed significant dissipations of various pesticides from different chemical groups in the soil to microbial action or other environmental factors. Furthermore, Fantke and Juraske [34] attributed the pesticide's tendency to decline in soil to the substance solubility. Less soluble pesticides are washed off and/or leached into the soil at a muchreduced rate, and thus, their half-lives increase, in contrast to the soluble compound, which had a decreased half-life.
The results of soil analysis indicated differences in the soil physical and chemical properties between the two sites, which could affect the growth of the crops and the fate of dinotefuran at the two locations. The Chuncheon soil pH is close to natural (6.5) and this allows maximum availability of soil nutrients. It has been reported that most plant nutrients are optimally available to plants within the pH range of 6.5 to 7.5, and this pH range is generally compatible with plant root growth [39]. In contrast, Pyeongtaek soil pH is more acidic (5.5), which may impede crop growth and hinder optimum propagation of useful soil microbes, as well as cause a slow release of dinotefuran and subsequent uptake by the plant [28,40]. The difference in soil physicochemical properties may explain the difference in uptake in the two crops on the two application dates (PBI 30 and PBI 60) (Figure 4). Dinotefuran has high water solubility (39.83 g/L) and low sorption to the soil particles. The latter could be attributed to its high solubility and low (−0.549) octanol-water (log K OW ) partition coefficient, which indicates its potential for leaching [7,40]. Despite its higher solubility, dinotefuran has had higher sorption and thus tends to accumulate in soil [40,41]. Generally, pesticide immobilization is based on physicochemical interactions between the pesticides and the soil organic matter, which affect fate, soil leaching, runoff, uptake, and bioavailability and decrease the concentration of the pesticide [12,42]. The microbial degradation mechanism of dinotefuran is not yet fully understood [43]. Dinotefuran was reported to degrade by 31% after three weeks of incubation with white-rot fungus (Phanerochaetesordida YK-624), a recalcitrant degrader of organic pollutant, under ligninolytic conditions, while it was found more stable under nonligninolytic conditions. Moreover, some genera of bacteria-Methylotenera, Ramlibacter, and Rubrivivax-were reported to degrade dinotefuran [44]. Neonicotinoids, the chemical group to which dinotefuran belongs, were reported [34] to have faster biodegradation rates in soil with a higher organic carbon content. The degradation rate of neonicotinoids in soil was reported [45] to be affected by changes in soil pH and microbial metabolism [46]. Furthermore, plant groundcover could alter the soil physical properties, decrease the soil clay content, increase the soil organic matter, and affect the distribution and dissipation patterns of pesticides in general [36,47].
The initial level of dinotefuran is higher in celery soil than in lettuce soil, and this may be explained by the difference in application rates, which were higher in celery [25]. Generally, there is no clear difference between the PBI-30 and PBI-60 soil residue in both locations, and this may be explained by physiochemical properties of the soil and crop and the variety grown [36,48].
Residues in lettuce soils disappear at a relatively faster rate in Chuncheon soil than in Pyeongtaek, as reflected by their levels and half-lives, whereas the dissipation in celery soil was faster in Pyeongtaek than in Chuncheon. The lowest half-lives (0.78-0.79 days) were observed in the celery soil of Pyeongtaek. The different patterns may be attributed to the physicochemical properties of dinotefuran, soil organic content, and the physiological characteristics of plants, which influence the translocation and dissipation of pesticides [29]. Chuncheon soil had a higher level of organic matter content, soil CEC, and pH value (6.5), which favor more diversity in the population of microorganisms, especially soil bacteria, which form a complex with soil organic matter, thus leading to faster degradation rates, as mentioned before [39,44,49]. The current results are in line with previous work conducted by Kwak et al. [35,45], who found that the half-lives of dinotefuran in radish soil treated with 2.01 and 9.35 mg/kg of initial residue were 6.2-8.9 days. Half-lives and dissipation rates of pesticides vary according to the plant species, environmental conditions, and the physicochemical properties of the targeted compound [50], especially water solubility [34].
Statistical analysis of the metrological data such as daily temperature, RH%, and precipitation indicated no significant effect on the dissipation and/or uptake of dinotefuran at both sites, and this was expected since the experiments were conducted under controlled conditions in greenhouses. This result is consistent with results of Ishag et al. [38,51], who reported that environmental factors did not affect the degradation of some pesticides.
Leafy vegetables are widely consumed in South Korea as kimchi and as raw leaves wrapping meat [22,23]. Consuming vegetables containing pesticide residues may pose severe health hazards through either acute or chronic toxicities depending on the level, frequency, and mode of exposure [1]. The levels measured at 50% and 100% maturity showed that this insecticide can be absorbed and move acrobatically and thus may contaminate the edible parts. Previous studies showed that dinotefuran is characterized by low log P OW of −0.549, high water solubility (39.8 g/L), and lower sorption to soil organic matter [7]; therefore, it can be absorbed and translocated by plants roots or leaves [7,10,12,35] and contaminate edible parts, thus posing a great risk to human health through the food chain [2,29,33,52]. Dinotefuran and imidacloprid were reported to translocate from soil to the plant and were detected in the nectar of woody landscape plants [53]. Translocation studies on dinotefuran and imidacloprid in ash trees, hemlock, walnut, avocado, and ilex trees proved that dinotefuran tends to translocate more quickly and rapidly degrade in woody plant tissues than imidacloprid [13,33,36,52,53]. A recent study conducted by Yuan et al. [29] showed the translocation of measurable levels of some pesticides in spinach from treated soil.
The total residue levels of dinotefuran and its main metabolites in the control sets at both sampling dates in the two crops and the two locations are below the detection limit (<0.01 mg/kg). Generally, the levels of dinotefuran residues in plants significantly decrease when the two crops approach 100% maturity in both PBI-30 and PBI-60 treatments at the two locations (Table 4). The reduction in pesticide residues through plant growth may be explained by crop dilutions, as reported by Hwang et al. [5] and Yuan et al. [29]. The growth dilution effect is one of the primary factors playing a substantial role in reducing the residue concentration [5]. Many studies reported the reduction in pesticide residue by crop dilutions [5,29,54].
The absorption of residual pesticides from contaminated soil by plants relies on the physicochemical properties of the compound [4], the interval between application and harvest, the route of absorption, and abilities of the pesticides to penetrate and translocate in the plant [33,55,56]. Dinotefuran residues in celery shoots are much higher than those in lettuce leaves, and this might be explained by the differences in the initial amount applied to both crops, which rely on the authorized rate of application [25]. Dinotefuran residues detected in lettuce at both experimental sites are much lower than Korean MRLs (25 mg/kg and 0.05 mg/kg) in stalk and stem vegetables, and below the 0.6 mg/kg tolerance level set by the Ministry of Food and Drug Safety, MFDS [57], and FAO Codex [58]. On the other hand, the residue levels in celery shoots at both experimental sites are much lower than that established in the FAO Codex [58]. This might indicate the significance of the current findings for the safe production of lettuce and celery using dinotefuran by adhering to a PBI of at least 30 days. Generally, the residue levels in lettuce leaves grown in Pyeongtaek are higher than those grown in Chuncheon, and there is a decreasing trend of residue level in the two crops at the Chuncheon site in PBI-60 compared to PBI-30, while at the Pyeongtaek site, the opposite is true. This might be explained by differences between initial amounts applied [25] to the two crops and/or differences in the soil physicochemical properties and environmental conditions [8,36] at the two locations, as mentioned earlier.
The current work reported the detection of measurable levels of three metabolites, MNG, DN, and UF (Table 3 and Figure S4). These metabolites were reported as major metabolites of dinotefuran by previous authors [7,51,54]. The presence of measurable residues of MNG in lettuce leaves and celery shoots detected in the current work is consistent with Turner [7], who reported that MNG is the major dinotefuran metabolite in the soil. The metabolites MNG, UF, BCDN, and DN were found in rotational crops with levels of 0.01 mg eq/kg and 0.035 ppm in immature and mature leaf samples. In addition, the metabolites MNG and DN were detected as major degradation products in water and sediment [59,60]. MNG can be formed by cleavage of tetrahydromethyl or tetrahydrofuran portion of the parent compound, while DN could be formed by the loss of the nitro group from the original compound. Subsequent hydrolysis of the nitroimino moiety leads to the formation of UF. This process can occur in plants or soil conditions, as well. The FAO manual [61] and Turner [7] also reported the formation of these metabolites in plants and soils. In another study, Zhang et al. [36] reported that the degradation of neonicotinoids was mainly via nitrate reduction, cyano hydrolysis, and chloropyridinyl dechlorination reactions.
The results of a seven-month low-temperature storage stability test indicated the stability of dinotefuran in plant and soil samples at various spiked concentrations (0.01 MLOQ (0.002 mg/kg), 0.1 MLOQ (0.02 mg/kg), and 10 MLOQ (0.2 mg/kg) of dinotefuran and its metabolites). This finding is in line with the FAO [61], which reported the stability of dinotefuran and its metabolites in various plant species and soils, and the EPA [8], which reported the stability of dinotefuran and its metabolites (MNG, UF, and DN) in various conditions. The residues of dinotefuran in summer-treated ilex leaves were found at higher concentrations in the foliage collected in the spring of the following year [11]. Furthermore, they also stated that dinotefuran might be more persistent than is generally believed.
The bio-concentration factor showed percentages of translocation of dinotefuran to the 100% mature lettuce leaves of 23.8% in PBI-30 and 28% in PBI-60. On the other hand, the translocation percentages of dinotefuran in celery shoots were 54% and 52%, respectively. This high bio-concentration factor may be explained by the low log P OW , high water solubility, and lower sorption to soil organic matter, meaning it can be readily absorbed and translocated by plants [7].
Similar results were reported in radish, which was found to uptake 4.9-16.7% of dinotefuran from treated soil after intervals of 10 to 70 days [35]. Other translocation studies on dinotefuran, other pesticides, and/or their metabolites in various crops reported similar findings of quick translocation through plant tissues [13,29,33].
The studied risk assessment parameters, RQc and RQa, reveal acceptable safety to consumers of lettuce and celery, as indicated by the calculated risk values of less than one. [31,32]. Celery displayed the lowest RQc and RQa values, and this can be explained by its LP value of 0.06 g/d compared to the LP value for lettuce of 7.19 g/d [54].

Conclusions
The results of the current study indicate that dinotefuran residues in lettuce leaves and celery shoots are influenced by intervals between application and harvesting periods, as well as by the physiochemical properties of dinotefuran and soil. We detected measurable levels of dinotefuran residues in the edible parts of lettuce and celery. The total residual amount of dinotefuran and its metabolites in lettuce leaves and celery shoots grown in Pyeongtaek and Chuncheon were lower than the tolerance level set by MFDS (2021) and/or Codex (2013). The disappearance of dinotefuran in Chuncheon lettuce soil was faster than that in Pyeongtaek, and vice versa in celery soil, as indicated by their half-lives. The shortest half-lives were recorded in the celery soil of Pyeongtaek. Dinotefuran soil residues were similar at both sites before application. Generally, there was no clear difference between the PBI-30 and PBI-60 soil residues at both locations. Based on that and since dinotefuran is applied as a slow-release granule formulation, we recommend growing minor crops 60 days after harvesting the primary crop, but not before 30 days in the worst case scenario, to ensure that dinotefuran residues are lower than the MRL tolerance levels in lettuce leaves and/or celery shoots. The values of the risk assessment parameters, RQc and RQa, for both plants were <1, indicating low or negligible expected risk. The current findings contribute to pesticide residue safety management in rotational crops. Future work in other rotational crops is suggested.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/agriculture12091443/s1, Table S1: The chemical structure formula of dinotefuran and its metabolites; Table S2: Soil properties of the experimental sites; Table S3: LC/MS/ MS operating conditions for the analysis of dinotefuran and its metabolites in lettuce, celery, and soil samples; Table S4: Recovery (%) of dinotefuran and its metabolites after 7 months of storage at −20 • C; Table S5: Meteorological data* during the experimental period; Table S6: Irrigation regime of lettuce and celery; Figure S1: Calibration curves of dinotefuran and its metabolites in lettuce; Figure S2: Calibration curves of dinotefuran and its metabolites in celery; Figure S3: Calibration curves of dinotefuran and its metabolites in soil; Figure S4: Flowchart of metabolite formation in plants and/or soils of lettuce and celery; Figure S5: Experimental design for the study; Figure S6: Excerpt chromatograms showing dinotefuran residues in lettuce soil of PBI-30, PBI-60, and control sets; Figure S7: Excerpt chromatograms showing dinotefuran residues in the celery soil of PBI-30, PBI-60, and control sets; Figure S8: Excerpt chromatograms of the results of storage stability test of dinotefuran and its metabolites (MNG, DN, and UF).