Occurrence, air-seawater exchange, and ecological risk of pesticides in the southern Bohai Sea, China

Introduction：

Pesticides are synthetic or naturally occurring chemical agents extensively employed in modern agriculture to eradicate or prevent pathogens, insect infestations, weeds, and other harmful organisms affecting crops and forest ecosystems, while also serving plant growth regulation purposes (Cao et al., 2024). Pesticide residues are usually classified into four categories: organochlorine pesticides, organophos- phorus pesticides, organic amine pesticides, and pyrethroid pesticides (Ma, 2022; Zhang et al., 2023a). These agrochemicals have extended degradation half-lives (> 180 days) and significant bioaccumulation potential in aquatic ecosystems, with specific compounds officially lis- ted as persistent organic pollutants (POPs) under the Stockholm Convention (UNEP, 2001). Notably, most pesticide residue precursors are volatile (Zheng et al., 2022), indicating that volatilization and sub- sequent environmental release inevitably occur during manufacturing,

transportation, and application.
It has been demonstrated that environmental pollutant levels exhibit a strong correlation with human activities (Lu et al., 2022a). For example, there was a significant positive correlation between population density and the concentration of pollutants in the Pearl River Delta (Zeng et al., 2017). The detection rate of pesticide residues is high in coastal areas, such as Bohai Bay and Laizhou Bay, and the surrounding rivers, and is closely related to agricultural production (Dang et al., 2021; Hu et al., 2009; Li et al., 2018). As an important agricultural product production base in China, Shandong has the highest annual usage of chemical pesticides in the country, approximately 110,000 tons (National Data CHN, National Bureau of Statistics, 2024). Adjacent to the Shandong Peninsula, the Bohai Sea is a shallow, semi-enclosed marginal sea characterized by substantial riverine inputs and limited water exchange (Gu et al., 2022; Song et al., 2024). Additionally, the Bohai Sea serves as a vital fishery base in China due to its rich resources,

faces serious pesticide pollution pressures (Wang et al., 2025; Yang et al., 2025b). Consequently, the dilution of land-based pollutants is inhibited, presenting an elevated ecological risk. Studies have shown that the pesticide concentrations in river water in the Shandong ranged from 19.1 to 74.0 ng/L, and the range of pesticide concentrations in the northern Shandong coastal area ranged from 0.2 to 14.3 ng/L (Cao et al., 2024). Consequently, the southern Bohai Sea exhibits some of the highest pesticide contamination levels globally.
Influenced by environmental factors and the inherent characteristics of pollutants, the dynamic exchange of pollutants occurs at the air- seawater interface through diffusion or direct transport by atmo- spheric bubbles or sea spray (Stanley and Bell, 2025)—a process known as air-seawater exchange. By studying the process of air-seawater ex- change, the potential risks of pollutants to ecosystems and human health can be further understood, providing a scientific basis for the control and monitoring of pollutants. The air-seawater exchange fluxes of γ-Hexa- chlorocyclohexanes (HCH), dieldrin, and chlorpyrifos were all negative, ranging from - 3600 to - 900 pg/m2/day, - 6400 to - 400 pg/m2/day and - 1400 to - 200 pg/m2/day, respectively, and all compounds were significantly less than 1, indicating that the ocean was receiving atmo- spheric deposition (i.e., net air-to-sea transfer during the sampling period; Bigot et al., 2016). However, current research has mainly focused on POPs such as polycyclic aromatic hydrocarbons (PAHs) and polychlorinated biphenyls (PCBs), and the research areas are mainly concentrated in the polar regions and the Pacific Ocean (Fu et al., 2023; Na et al., 2020; Zhang et al., 2021). Notably, the interplay between inshore hydrodynamic conditions and pesticide transformation path- ways during air-seawater exchange remains poorly constrained, limiting the predictive capabilities for coastal ecological risk assessment.
Previous research on pesticides in the Bohai Sea covered estuaries, the atmosphere, and marine areas (Liu et al., 2018; Zhao et al., 2025). However, there is a lack of comprehensive understanding of the levels and transport patterns of pesticides. The lack hinders a comprehensive understanding of human influences on coastal pesticide cycling, particularly in marine environments where intensive agricultural and industrial activities converge. Therefore, the sources and distribution

characteristics of pesticides in the marine environment of the Bohai Sea and their relationship with human activities are current scientific issues that require urgent study. If the pesticides in the inshore waters of the Bohai Sea mainly originate from agricultural activities and river inputs, the distribution of their concentrations will be characterized by high inshore and low offshore concentrations and will be negatively corre- lated with the distances between areas of intensive agricultural activities and the mouths of rivers. The present study aimed to: 1) quantify con- centrations of pesticides in gaseous and aqueous phases within the southern Bohai Sea coastal area and its surrounding rivers; 2) elucidate the spatial distribution of pesticides in water and assess their environ- mental impacts; 3) reveal the air-seawater exchange dynamics of pes- ticides; 4) evaluate potential ecological risks of pesticides.

2. Methods and materials
   2.1. Sample collection
   In May 2024, 30 seawater samples were collected in the southern Bohai Sea, China, including 9 sampling sites (Sites 1–9) in Bohai Bay (BB), 5 sampling sites (Sites 10–14) in the Yellow River (YR), 14 sam- pling sites (Sites 15–28) in Laizhou Bay (LB), and 2 sampling sites (Sites 29–30) in the western of the Temple Island Archipelago (MW). A paired sampling design was implemented across 12 major rivers in the southern Bohai Sea, with synchronized collection of river water and atmospheric samples (n = 12 each) at estuarine locations. Two river water samples were collected from the upper reaches of the Weifang and Mi rivers (Fig. 1B; Specific information about samples collection is described in SI). Detailed GPS coordinates and meteorological parameters, which include temperature, relative humidity, atmospheric pressure, and wind speed, were documented at each sampling site (Table S1 & S2).
   Briefly, river water and seawater samples (20L) were collected using a stainless seawater sampler. For gas phase, atmospheric samples (100 m3) were collected using high-volume solid-phase extraction. To ensure reproducibility, samples were performed in triplicate. The samples un- derwent filtration through GF/F glass fiber membranes (0.45 μm, 142

mm; Whatman) to eliminate particles, followed by concentration via XAD2/4 adsorption columns and storage at _ 20 ◦ C pending laboratory analysis.

2.2. Sample pre-treatment and instrumental analysis
The sampling method employed high-volume solid-phase extraction technology for organic analysis (Liu et al., 2024; Shi et al., 2024; Yang et al., 2025a). The concentration for the calibration curve was set to 10, 20, 50 μg/L for analyzing the pesticides, and the R values were greater than 0.99. For high-volume solid-phase extraction sampling columns (extraction columns), all were dried in a 45 ◦ C thermostat for 96 h to ensure complete removal of water from the extraction columns before

H(salt,T) = HT × 10Kssalt

where ΔHvap is the enthalpy of vaporization of the pollutant and Ks[salt] denotes the Setschenow constant of the pollutant (L/mol). It is calcu- lated as follows (Ma et al., 2018):
KS = 0.04 log KOW + 0.114 (4-1)
Uncertainty analysis of the air-water fugacity ratios:

, 、， , 、 fi , 、2 , 、2 , 、2 , 、2 ,1/2

sample extraction. Briefly, the extraction column was placed in a stainless-steel sleeve and eluted sequentially with 45, 25, and 25 mL of dichloromethane (DCM). Extraction was performed after 12, 2, and 2 h of shaking. The extracts from each shake were purified using silica gel columns to obtain the eluate. The combined purified eluates were concentrated to 1 mL at 40 ◦ C under mild nitrogen, replaced with 10 mL of n-hexane, reconcentrated to 1 mL, and transferred to a chromatog- raphy vial. The sample extracts were analyzed for 221 pesticides using a GCMS-TQ8050 NX gas chromatography-mass spectrometer (GC–MS/ MS, Shimadzu, CHN). Chromatographic separation was performed on an SH-RXI 5SIL MS capillary columns (30 m × 2.5 mm, 0.25 μm) using research-grade helium as a carrier gas. Pesticides were analyzed and detected based on the ion information and retention time (Table S3).
The actual concentrations (Ct) of the pollutants were calculated as follows:
Ct = (Cm _ Ck)Vm /VS RR (1-1)
where Cm denotes the concentration on the machine, Ck denotes the concentration of the sample blank, Vm denotes the volume on the ma- chine, Vs denotes the volume of the sample, and RR denotes the recovery rate of the pollutant.

2.3. Air-seawater exchange
Air-seawater exchange was calculated based on concentration of pesticides in the estuary and atmospheric composition. The FR and flux describe the process of air-seawater exchange, representing the driving force and the actual rate, respectively. Integrating these approaches can reveal the transmission direction and mechanism of pesticides and provide a key scientific basis for marine pollutant management.
2.3.1. Air-seawater exchange fugacity ratio
FR of a pollutant describes the potential direction of diffusion or transport of the pollutant at the air-seawater interface.
FR = fa /fw = CaRT/CWH(T,salt) (2-1)
where f wand fadenote the pollutant FR in the seawater and atmosphere, respectively. Ca and Cw denote the pollutant concentrations in the at- mosphere (pg/m3) and seawater (pg/L), respectively. T denotes the temperature (K). R denotes the molar gas constant (8.31 Pa*m3/mol), and H(T，salt)denotes the Henry's constant corrected for salinity and temperature. It is calculated as follows (Lohmann et al., 2013; Sahsuvar et al., 2003):

Where δH/H, δCw/Cw, δCa/Ca and δT/T denotes the relative standard deviations (RSD) of H, Cw, Ca and T. We hypothesis the RSD of H and T was 50% and 5%.
2.3.2. Air-seawater exchange fluxes
The air-seawater exchange flux (Faw, pg/m2/day) was calculated using a modified version of the Whitman two-film resistance model (Bigot et al., 2016; Liu et al., 2016):
Faw = KOL Cw _ CaRT/H(T,salt)
Where KOL denotes the air-seawater exchange mass transfer coefficient of the pollutant, and Ca and Cw denote the pollutant concentrations in the atmosphere and seawater (pg/m3), respectively. Where KOL is calculated as follows:
1/KOL = 1/KW + RT/KaH(T,salt) (7-1)
where Ka and Kw denote the mass transfer resistance (cm/s) in the at- mosphere and seawater, respectively. Ka and Kw are defined as:
Ka = Da /DH2O,a × Vwater,a (8-1)
Da = 1.55/M0.65 cm2 /s (8-2)
DH2O, a is 0.27 cm2/s, and M denotes the molecular weight of the contaminant.
Vwater,a = 0.2 × V + 0.3 cm/s (9-1)
V indicates the measured wind speed (m/s).
Kw = (SC W/600)_ asc × Vco2,w (10-1)
SC W = VW/DW (10-2)
where Vw denotes 0.00893 cm/s assuming 25 ◦ C and 1.013 Pa, and asc is 0.5 for wind speeds higher than 4.2 m/s and asc is 0.67 for wind speeds lower than 4.2 m/s.
Dw = 2.7 × 10_ 4 /M0.71 cm2 /s (11-1)
Vco2,w = (0.79 × V _ 2.68) × 10_ 3 cm/s (11-2)
This formula applies when the wind speed is between 4.2 and 13 m/ s; otherwise, V CO2 = 0.65 × 10_ 3 cm/s.
Uncertainty analysis of air-water exchange flux:

(12-1)
We hypothesis the RSD of Kol, H and T was30%, 50% and 5%, respectively.
2.4. Ecological risk assessment
In the present study, the risk quotient (RQ) index was employed to assess the ecological risk of the pesticides (Cao et al., 2024; Hernando et al., 2006; Li et al., 2018; Ren et al., 2023; Zhang et al., 2022):
RQ = MEC/PNEC (13-1)
Where MEC denotes the concentration of pesticides in the seawater sample. PNEC is the predicted no-effect concentration for pesticides which from US EPA ECOTOX and the assessment factor is 1000.
2.5. Statistical analysis
Data are expressed as average ± standard deviation (SD). The Shapiro-Wilk test was employed to assess data normality. Moreover, one-way analysis of variance (ANOVA) was used to compare the dif- ferences in different environmental media or regions after the data met homogeneity of variance (Levene test) (SPSS 27.0, IBM, USA). Statistical analyses and graphics were generated using Origin 2022 (Origin Lab, USA). The “Kriging interpolation” tool (ArcGis 10.8, Environmental Systems Research Institute, USA) was used to simulate the spatial dis- tribution of pesticides.
2.6. Quality assurance and quality control
To avoid errors caused by experimental consumables, all of them have been processed. For example, the experimental glassware was ul- trasonically cleaned with Milli-Q water and dried in an oven. The stainless steel and GFFs heated in a muffle furnace at 450 ◦ C for more than 4 h to remove organic interferences. The glassware was rinsed three times with DCM prior to use.
For the water samples, the recovery of individual pesticides ranged from 41% to 128%. For the atmospheric samples, the recovery of indi- vidual pesticides ranged from 31% to 117%. 20 L of Milli-Q water served as a seawater field blank, assessing potential contamination during processing. With the exception of permethrin, which had a blank con- centration of 3.3 pg/L, all other pesticide blanks in the water samples were close the detection limit. Air field blanks are collected by briefly exposing the columns to the atmosphere at the sampling sites. The whole blank for pesticides in atmospheric samples were close to the detection limit. To avoid errors due to randomness, data were recorded only when detection rate of the samples above 50%. The method detection limit (MDL) was calculated as the average of the blank values plus three times the standard deviation (Cao et al., 2024; Ma et al., 2018). The MDLs for individual pesticides in the water and atmospheric sample ranged from 0.0 to 3.3 pg/L and 0.0 to 1.5 pg/m3, respectively.

3. Results and discussion
   3.1. Distribution and concentration of pesticides in surface seawater, river water and atmosphere
   The present study revealed significant environmental prevalence of pesticides in the southern Bohai Sea ecosystem, with detectable residues across coastal seawaters, rivers, and the atmosphere. Notably, the three

environmental media exhibited distinct pesticides detection profiles, reflecting phase partitioning governed by compound-specific physico- chemical properties. Specifically, a total of 26 pesticides were detected across three environmental media: surface seawater (17.0 pg/L–8.7 ng/ L), river water (5.4 pg/L–33.3 ng/L), and atmosphere (3.2 pg/m3–1.8 ng/m3). Based on chemical structure, these pesticides were categorized into four groups: organochlorines (OCPs), organophosphates (OPPs), pyrethroids (PYRs), and Others (Table S4).
Notably, the three environmental media exhibited distinct pesticides detection profiles, reflecting phase partitioning governed by compound- specific physicochemical properties. For example, boscalid was detected only in water samples, while molinate was detected only in atmospheric samples. This may be attributable to boscalid's higher octanol/air partition coefficient (Koa) and enthalpy of vaporization, coupled with a lower Henry's law constant (HLCs), whereas molinate exhibits the opposite pattern for these properties (Table S5). A high Koa and enthalpy of vaporization indicates low volatility, meaning the pollutant is not readily volatilized into the air (Paluselli et al., 2018). Futhermore, A high HLCs indicates pollutants are readily volatilized into the air (Guillaume et al., 2025). Six pesticides were detected in all three media: permethrin (seawater: 2.4 ± 2.0 ng/L, river water: 4.4 ± 0.8 ng/L, at- mosphere: 0.3 ± 0.2 ng/ m3), procymidone (seawater: 1.8 ± 0.6 ng/L, river water: 2.2 ± 0.5 ng/L, atmosphere: 0.5 ± 0.2 ng/ m3), phorate sulfoxide (seawater: 1.0 ± 0.2 ng/L, river water: 2.4 ± 0.8 ng/L, at- mosphere: 1.3 ± 0.4 ng/ m3), quinoxyfen (seawater: 0.2 ± 0.1 ng/L, river water: 0.9 ± 0.3 ng/L, atmosphere: 0.4 ± 0.2 ng/m3), dichlobenil (seawater: 0.1 ± 0.0 ng/L, river water: 0.2 ± 0.2 pg/L, atmosphere: 12.6 ± 7.1 pg/m3), pendimethalin (seawater: 0.1 ± 0.1 ng/L, river water: 0.2 ± 0.2 ng/L, atmosphere: 0.1 ± 0.2 pg/m3; Table S4).
3.1.1. Distribution and concentration of pesticides in the surface seawater
Among the 221 target pesticides monitored, 11 pesticides were detected in >50% of the samples. Dichlobenil, β-HCH and Permethrin were the widely detected pesticide (100%; Table S4). We found PYRs were the most abundant pesticide class in seawater, in contrast to OCPs, which were the lowest (Fig. 2A). Furthermore, a comparison of con- centrations across all pesticides identified permethrin as the dominant compound (Fig. 2B). The concentration s of pesticides in the seawater of Bohai Sea were similar in magnitude to 2023 levels (Cao et al., 2024). Compared with other areas, the concentrations of pesticides were 1–5 orders of magnitude greater than in Fildes Bay (Luarte T. et al., 2024), yet remained below the levels of pesticides in the South Sea, China (Zheng et al., 2023) and Arctic waters (Ding et al., 2023). Notably, permethrin was the main contaminant in the present study, yet its concentration was substantially below that detected in Lake Taihu (Wang et al., 2023), possibly because permethrin is widely utilized in urban and domestic environments (Wang et al., 2024). Consequently, the concentrations of pesticides in the southern Bohai Sea occupy an intermediate position globally. The detected HCH isomers had different frequencies of detection and concentrations. Notably, the β-HCH exhibited the highest concentrations of HCH in the surface seawater because the α, γ-HCH are easily degraded and δ-HCH is the most easily adsorbed. (Yu et al., 2014).
Pesticide concentrations exhibited pronounced spatial heterogeneity across the study area, with a distinct contamination gradient. The analysis showed significant differences in pesticides concentrations across sampling areas in the southern Bohai Sea (ANOVA; p < 0.05). Specifically, the Σ11Pesticides were significantly more abundant in LB (5.7 ± 1.6 ng/L) than in YR (5.2 ± 1.1 ng/L), BB (4.6 ± 1.8 ng/L), and MW (4.3 ± 1.2 ng/L). Laizhou Bay exhibited the highest concentrations of pesticides, peaking near the Xiaoqing River estuary (Fig. 3A). One possible reason for this discrepancy is that Laizhou Bay is a closed bay with a large number of river inputs, including the Xiaoqing River, traversing major agricultural regions of Jinan, Zibo, Binzhou and Wei- fang, which have high concentrations of pesticides (Cao et al., 2024). Another possible reason is that the spatial distribution of pollutants is also closely linked to ocean current movements (Zhang et al., 2023b). Influenced by the Bohai Sea Warm Current, the MW area exhibits rela- tively low pollution levels, while due to the northward flow of the Laizhou Bay Cyclonic Circulation, pesticide concentrations in the YR region diffuse northward (Cui et al., 2024; Song et al., 2019).