Assessing pesticide transport and export dynamics from watersheds to Laizhou Bay, China

Pesticide residues from watersheds are increasingly exported into seawater, negatively impacting marine ecosystem health. However, existing watershed and flux models for pesticides are limited and insufficient. This study developed a Pesticides Export from Watersheds to Sea (PEWS) model to assess the transport processes of four common pesticides—atrazine, nicosulfuron, chlorpyrifos, and imidacloprid—in Laizhou Bay, China. The model was calibrated and analyzed using field surveys and geographic information system (GIS) data to evaluate the spatiotemporal distribution of pesticide fluxes within the watershed. The results indicated a decreasing trend in pesticide flux into Laizhou Bay from 2001 to 2023, with annual fluxes in 2023 recorded as 11.5 × 103, 0.1 × 103, 0.01 × 103, and 0.6 × 103 kg/yr for atrazine, nicosulfuron, chlorpyrifos, and imidacloprid in 2023, respectively. The seasonal pattern followed an inverted U-shape, with fluxes peaking in summer and declining in winter. Spatially, pesticide fluxes were highest in agricultural areas, with atrazine and nicosulfuron fluxes peaking in the downstream regions of the Xiaoqing River watershed, imidacloprid fluxes highest in the Mi and Bailang Rivers, and chlorpyrifos fluxes more prominent in the upstream regions of the Weihe River. The study further revealed that less than 3 % of the total pesticide load was exported into the sea, while 1 %–4 % was absorbed by crops, and over 90 % remained trapped in soil and river systems. The PEWS model couples Global NEWS and SWAT frameworks to simulate watershed-scale pesticide export. By quantifying pesticide export fluxes from watersheds to the sea, this study contributes to the development of effective strategies for controlling pesticide pollution.

1.Introduction

Pesticides, including atrazine, nicosulfuron, chlorpyrifos, and imi- dacloprid, are widely used in global agricultural production (Gao et al., 2019; Snow et al., 2020), with an annual application of approximately 3 million tons, according to the Food and Agriculture Organization (FAO: http://www.chinapesticide.org.cn/). However, their extensive use has led to ecological degradation and human health risks (Andrade-Rivas et al., 2023). In response, an increasing number of countries and inter- national organizations have prioritized biological security, developing protection programs to reduce pesticide residues in soil and water(Craparo et al., 2023; Daniel et al., 2020; Salazar-Ord and Sayadi, 2011). The quantities and application methods of these pesticides vary by region and policy (Zhong et al., 2015), and the Ministry of Agricul- ture adopted a “zero growth” action plan regulating pesticide usage in China by 2020 (Gu et al., 2025). Despite these efforts, quantitative es- timations of pesticide loads remain inadequate, yet they are crucial for effective control and management (Wolfand et al., 2019).Watershed models for pesticide transport and flux estimation have gradually advanced (Tran et al., 2025; Tu et al., 2023). One of the earliest models, CREAMS, was developed in the 1980s to simulate hy- drological, erosion, and pesticide transport processes at small

• Corresponding author at: Key Laboratory of Marine Chemistry Theory and Technology, Frontiers Science Center for Deep Ocean Multispheres and Earth System, Ministry of Education, Ocean University of China, Qingdao 266100, China.
  E-mail address: likeqiang@ouc.edu.cn (K. Li).
  https://doi.org/10.1016/j.marpolbul.2025.118576
  Received 27 April 2025; Received in revised form 21 July 2025; Accepted 8 August 2025
  Available online 16 August 2025
  0025-326X/© 2025 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.

agricultural field scales (Crowder et al., 1985). Later, models such as Agricultural Non-Point Source Pollution Model (AGNPS) (Pantone and Young, 1992), Hydrological Simulation Program—Fortran (HSPF) (Al- Abed and Whiteley, 2002), and the Soil and Water Assessment Tool (SWAT) (Di Luzio et al., 2002) were introduced to evaluate pesticide export and fluxes, particularly for widely used pesticides like glyphosate and atrazine, and their impacts on aquatic environments. The AGNPS model focuses on small watersheds and single events, incorporating runoff and soil erosion processes but lacks long-term time-series simu- lation (Kinnell, 2000). The widely used SWAT and HSPF models excel in hydrological and geochemical processes, yet they require substantial data inputs and involve complex parameterization (Luo and Zhang, 2009; Parker et al., 2007). The Global NEWS model, as a type of watershed model, emphasizes export processes and source-sink re- lationships, developed by UNESCO-IOC in 2010 (Mayorga et al., 2010). In contrast to process-based models such as SWAT, the Global Nutrient Export from Watersheds (NEWS) model adopts a simplified structure with lower data requirements and fewer parameters, making it partic- ularly suitable for large-scale, inter-basin, and data-scarce watershed studies. The Global NEWS2 model has been widely used to simulate annual or seasonal fluxes of dissolved inorganic nitrogen (DIN) and dissolved inorganic phosphorus (DIP) in several major watersheds, including the Amazon, Mississippi, Yellow, Yangtze, and Pearl Rivers (Jiao et al., 2023). However, there are only Global NEWS models asso- ciated with nutrients (e.g., nitrogen, phosphorus, silica) loads and fluxes in watersheds (Qu and Kroeze, 2012; Strokal et al., 2016).
Despite advancements, existing watershed models for pesticides (including herbicides, insecticides, and fungicides) still face challenges due to variability in pesticide structures and behaviors (Chen et al., 2024). The diversity in pesticide physicochemical properties, such as partition coefficients and half-lives, significantly affects their fate in the environment, yet is often oversimplified in current models (Six et al., 2002). Additionally, key biogeochemical processes such as foliar inter- ception (Villamizar and Brown, 2017), soil adsorption, river transport, and degradation are not fully integrated across multiple media. Furthermore, most models are limited to individual watersheds, making them weak in inter-basin or multi-watershed applications at larger spatial scales.
To address these gaps, this study developed the Pesticides Export from Watersheds to Sea (PEWS) model, based on Global NEWS2, to simulate large-scale, inter-basin pesticide export in Laizhou Bay, China. Which is a semi-closed bay in Bohai Sea, imported a large number of pesticides from Shandong Province, with higher pesticide residue levels than other coastal waters (Wang et al., 2025; Zhen et al., 2019). We enhanced the Global NEWS model framework by incorporating pesticide-specific parameters (e.g., degradation half-lives in crops, soil, and rivers) and environmental drivers (e.g., temperature and

precipitation) to better capture the seasonal dynamics of pesticide export. Moreover, to improve the regional applicability and achieve finer temporal resolution, we calibrated the model using field observa- tions of four representative pesticides and conducted monthly simula- tions oftheir export from multiple small watersheds around Laizhou Bay during 2001-2023. Additionally, we analyzed blocking capacities along the pesticide transport pathways from source to sea. The findings pro- vide scientific and technological support for pesticide pollution pre- vention and management in coastal areas.