Polycyclic aromatic hydrocarbons in surface sediments of Lingshui County lagoons: distribution, source analysis, and ecological risk assessment Authors: Bingxin Zhong, Zhifu Wang, Baijuan Yang, Weihua Feng, Nina Ma, Baoxing Huang, Hengtao Xu, Duian Lv Published: 2025 ABSTRACT AbstractThe Lingshui County lagoons represent Hainan’s first marine special conservation area, yet little is known about the distribution and source characteristics of PAHs in their surface sediments. This study employed principal component analysis, PMF, and MERM-Q modeling to determine the distribution, sources, and potential ecological risks of PAHs in Lingshui Lagoon’s surface sediments. The data indicate: (1) Spring PAH concentrations in the study area ranged from 18.12 to 111.30 ng/g (dry weight), with a mean concentration of 60.83 ng/g. Hexacyclic PAHs accounted for the highest proportion (52.43%), followed by tetracyclic PAHs (21.04%). (2) PMF model analysis revealed that PAHs in the lagoon primarily originate from fossil fuel and biomass combustion. (3) The PAH concentrations in the lagoon’s surface sediments indicate moderate to low pollution levels, with relatively minimal potential ecological risk. These findings provide a reference for identifying PAH sources in Lingshui Lagoon’s surface sediments and establish a scientific foundation for improving the lagoon’s environmental quality.Supplementary InformationThe online version contains supplementary material available at 10.1007/s10653-025-02954-3.Keywords: PAHs, Lagoon, PMF, Distribution characteristics, Ecological risk assessment Introduction Polycyclic aromatic hydrocarbons (PAHs) are aromatic hydrocarbons containing two or more benzene rings. They are a major class of persistent organic pollutants and are known for their toxicity, teratogenicity, and carcinogenicity (Said et al., 2023; Vijayanand et al., 2023). PAHs are produced by the incomplete combustion of organic matter. Based on their sources, PAHs can be categorized into two main types: natural sources, such as forest fires, natural oil seeps, coal deposits, diagenetic formation, and volcanic eruptions; and anthropogenic sources, including agricultural, mining, and industrial activities (Wang et al., 2024). PAHs emitted into the atmosphere can penetrate and spread into marine ecosystems through dry and wet deposition and air–water exchange pathways. Due to their low solubility and hydrophobic lipophilicity, PAHs bind to particulate matter in water bodies and subsequently settle into sediments, where they are resistant to degradation. Over time, they accumulate to form pollution sources and cause environment contamination. PAHs pose a serious threat to human health and environmental systems. Exposure to polycyclic aromatic hydrocarbons in fish and shellfish may lead to oxidative stress, DNA damage, and reproductive function alterations (Sadauskas-Henrique et al., 2024). Research (Honda & Suzuki, 2020; Said et al., 2020) has shown that PAHs can cause endocrine disruption and tissue-specific toxicity in organisms. They can also induce severe neurotoxic effects, such as impairing fetal neurodevelopment (Zhang et al., 2019a). Among PAHs, fluoranthene affects the stability of soil aggregates (Mo et al., 2023), thereby reducing soil permeability, aeration, and water-holding capacity (Yang et al., 2021). Studying their behavior and ecological impacts helps us understand both the contributions and threats these compounds pose to marine biodiversity. Currently, screening for highly efficient microbial strains and plant species remains a key focus for remediating PAHs residues in the environment (Tazdaït & Salah-Tazdaït, 2021). This study investigates the spatial distribution of PAH concentrations in surface sediments of Lingshui County’s lagoon and identifies pollution sources, providing crucial information on the status of Hainan’s first marine protected area. Previous research in this region has primarily focused on heavy metals, eutrophication, and anthropogenic impacts on the lagoon ecosystem. To date, no studies have examined PAHs levels in sediments of Lingshui County’s lagoon. Only (Zhang et al., 2024) have analyzed organic pollutants in the water column of Xincun Lagoon. The lagoon’s shoreline features dense residential areas and a thriving marine industry. Historically, substantial amounts of automobile exhaust, oil spills from vessels, and wastewater entered the marine ecosystem through surface runoff and atmospheric wet and dry deposition. Since 2015, the national government has implemented a bay remediation initiative targeting Lingshui County’s lagoon, launching ecological restoration projects. Economic development has begun shifting toward tourism, with the lagoon serving as the cornerstone of this economic transformation. Therefore, determining the current levels of these compounds following ecological remediation is crucial. Our objectives are to (1) determine the concentration levels of PAHs in lagoon sediments after implementing ecological remediation, (2) identify potential sources of PAH contamination, and (3) assess the potential ecological risks posed by PAHs in the lagoon. Materials and methods Study area and sampling The lagoons of Lingshui County are situated within the Li Autonomous County of Hainan Province, China, adjacent to the South China Sea. These semi-enclosed shallow water bodies are connected to the ocean, surrounded by land, and relatively small in area. There are two lagoons in total. The western lagoon is designated as Xincun Lagoon (18.395°N to 18.445°N, 109.970°E to 110.036°E), while the eastern lagoon is referred to as Lian Lagoon (18.405°N to 18.444°N, 110.043°E to 110.069°E). Xincun Lagoon measures 5.50 km in length from east to west and approximately 4.50 km in width from north to south, covering a water area of 19.43 square kilometers. The total shoreline length within the bay is 20.41 km, with a lagoon entrance approximately 250 m wide. The water depth distribution follows a pattern of shallow edges, moderate depth in the central region, and deeper southern areas compared to the shallower northern section. The average water depth at the entrance is 5.70 m, with a maximum depth of 11.20 m (Li et al., 2024). Lian Lagoon extends approximately 4.39 km in length, with a maximum width of about 2.82 km and a water area of 7.92 square kilometers. It connects to the South China Sea through a narrow approximately 60 m wide, exhibiting an irregular morphology characterized by a canstricted inlet and a broader basin. In terms of hydrodynamic conditions, the tidal regime in Lingshui Lagoon is characterized by an irregular mixed diurnal pattern, with an average tidal range of 69 cm and a maximum range of 155 cm (Ou et al., 2022). The currents are predominantly tidal, displaying tidal bore characteristics (Yipu et al., 2022): flood currents flow eastward, while ebb currents move westward. Residual currents are generally weak, though slightly stronger in the southern region. Flow directions vary spatially, with southeastward movement in the northern and central sections and westward movement in the southern section, at velocities ranging from 10 to 25 cm/s. Wave activity is primarily southerly in origin. Due to effective sheltering, even during periods of strong northeast winds generating large waves in the open sea, wave conditions within the lagoon remain relatively calm. The study area and sampling locations are illustrated in Fig. 1. Extensive mangrove forests are present near sampling points 6, 8, 30, 31, 34, and 45. Points 1–4, 10, 11, 23, 25, 29, and 39 are located near residential areas. Detailed site-specific information is provided in Annex 1. Schematic diagram of the study area and sampling points Surface sediment samples from the Lingshui County lagoons were collected between April 8 and 10, 2025, using a grab sampler to obtain the upper 0–2 cm layer of seabed sediment. Samples were stored in sealed, amber glass bottles with ground glass stoppers. During transport, samples were maintained under dark conditions at temperatures of 4 °C or lower and promptly delivered to the laboratory. Prior to analysis, experimental samples were freeze-dried, homogenized, sieved, and stored at − 20 °C until extraction and further processing. Analysis The gas chromatography–mass spectrometry system employed was a G3171A model, manufactured by Agilent Technologies (China) Co., Ltd. The organic reagents utilized in the experiment (e.g., hexane and dichloromethane) were obtained from Beijing Bailinwei Technology Co., Ltd., with liquid chromatography-grade purity. The 16 PAHs are naphthalene (Nap), acenaphthylene (Acy), acenaphthene (Ace), fluorene (Flu), phenanthrene (Phe), anthracene (Ant), fluoranthene (Flo), pyrene (Pyr), benzo[a]anthracene (B[a]A), chrysene (Chr), benzo[b]fluoranthene (B[b]F), benzo[k]fluoranthene (B[k]F), benzo[a]pyrene (B[a]P), indeno[1,2,3-cd]pyrene (IcdP), dibenzo[a,h]anthracene (D[a,h]A), and benzo[g,h,i]perylene (B[g,h,i]P). For the experiment, 4, 4-difluorobiphenyl was chosen as the internal standard. Sample pretreatment and detection The pretreatment methods for sediments in this study were mainly based on the national standards HJ 805-2016 and HJ 950-2018, with minor modifications (Tepe et al., 2022). The specific treatment methods are as follows: The collected sediment samples were placed in an enamel or glass tray, foreign objects such as branches, leaves, and stones were removed, and the samples were mixed thoroughly. After processing in a freeze dryer for 48 h, the samples were ground and crushed. After drying, the sediments were sieved through a 100 mesh screen. The collected sediment samples were placed in an enamel or glass tray, foreign objects such as branches, leaves, and stones were removed, and the samples were mixed thoroughly. After processing in a freeze dryer for 48 h, the samples were ground and crushed. After drying, the sediments were sieved through a 100 mesh screen. The sediment was accurately weighed to 10 g. 80.0 μL of the substitute standard solution were added to the prepared sediment sample. A mixture of 10 mL hexane and dichloromethane (v:v = 2:1) was used for ultrasonic extraction for 30 min. A small amount of quartz glass wool was placed in the funnel neck, an appropriate amount of anhydrous sodium sulfate was added, and the extract was filtered through the funnel. The extraction process was repeated twice more with the sediment to obtain a total extract volume of 30 mL. The sediment was accurately weighed to 10 g. 80.0 μL of the substitute standard solution were added to the prepared sediment sample. A mixture of 10 mL hexane and dichloromethane (v:v = 2:1) was used for ultrasonic extraction for 30 min. A small amount of quartz glass wool was placed in the funnel neck, an appropriate amount of anhydrous sodium sulfate was added, and the extract was filtered through the funnel. The extraction process was repeated twice more with the sediment to obtain a total extract volume of 30 mL. The extract was concentrated using a rotating evapotator to approximately 2 mL. The concentrated solution was transferred to a graduated concentration vessel using a disposable pipette. The bottom of the rotating evaporator flask was rinsed twice with a small amount of hexane-dichloromethane mixed solvent, all the concentrated solutions were combined, and further concentration was carried out to approximately 0.5 mL using nitrogen blowing. The temperature was set to 40 °C, an appropriate amount of internal standard solution was added to ensure the internal standard concentration matches that in the calibration curve, and the mixture was diluted to 1.0 mL with a hexane-dichloromethane mixture. The solution was filtered through a 0.22 μm membrane into a sample vial for analysis. The extract was concentrated using a rotating evapotator to approximately 2 mL. The concentrated solution was transferred to a graduated concentration vessel using a disposable pipette. The bottom of the rotating evaporator flask was rinsed twice with a small amount of hexane-dichloromethane mixed solvent, all the concentrated solutions were combined, and further concentration was carried out to approximately 0.5 mL using nitrogen blowing. The temperature was set to 40 °C, an appropriate amount of internal standard solution was added to ensure the internal standard concentration matches that in the calibration curve, and the mixture was diluted to 1.0 mL with a hexane-dichloromethane mixture. The solution was filtered through a 0.22 μm membrane into a sample vial for analysis. Quartz sand was used as a substitute for the actual sample, and the blank sample was prepared following the same steps as for sediment sample preparation. Quartz sand was used as a substitute for the actual sample, and the blank sample was prepared following the same steps as for sediment sample preparation. The gas chromatograph/mass spectrometer used equipment from Agilent Technologies, Inc. The chromatographic column is a quartz capillary column with the following specifications: length 30 m, inner diameter 0.25 mm, membrane thickness 0.25 µm, and a stationary phase of 5% phenyl-methyl polysiloxane. Under GC reference conditions, the injection port temperature was set to 280 °C, using a splitless injection method with an injection volume of 1 μL, and the column flow rate was maintained at 1.0 mL/min. The column temperature starts at 80 °C and was held for 2 min, then increased at a rate of 20 °C per minute to 180 °C and held for 5 min, followed by an increase at a rate of 10 °C per minute to 290 °C and held for 5 min. For mass spectrometry reference conditions, an electron impact source (EI) was selected, with an ion source temperature of 230 °C and an ionization energy of 70 eV. The interface temperature was set to 280 °C, the quadrupole temperature to 150 °C, and the mass scan range covered 45–450 amu. A solvent delay time of 6 min was set, and data were collected using full scan mode (SCAN). Quality assurance and quality control All glassware used in the experiment was sterilized at 300 °C for 12 h to prevent the introduction of external contamination. During the analytical testing process, laboratory blank samples were set up to detect any potential contamination introduced during the experiment. To ensure the reliability and accuracy of the data, two sets of parallel samples were analyzed for each sample, ensuring that the relative error was less than 15%. In terms of quantitative analysis, the internal standard method was used as the reference method. The standard curve consisted of six concentration levels (0, 4, 10, 50, 200, and 500 ng/mL), and the correlation coefficients for the curve fitting of all 16 target compounds exceeded 0.99. Additionally, the recovery rates of the target compounds ranged from 71 to 119%, with detection limits for each compound ranging from 0.08 to 0.61 ng/mL, fully meeting the detection requirements. To ensure the accuracy of the analytical results of the samples, the mass fraction of PAHs in the actual samples was calibrated using solvent blanks. The detection limit was calculated based on a signal-to-noise ratio of 3 times the blank measurement, while the quantification limit was calculated based on a signal-to-noise ratio of 10 times. Multiple measurements were averaged. Source identification We selected PMF version 5.0 for data analysis. Detailed information regarding the PMF model is available on the U.S. Environmental Protection Agency website. The fundamental equation of the PMF model is: The n × m dimensional analytical component concentration matrix X is decomposed into two factor matrices: G is the n × p factor contribution matrix, F is the p × m factor spectrum matrix, and an n × m residual matrix E is generated. Here, n denotes the number of samples, m represents the number of target compounds, and p indicates the sources of different PAHs. The equation can be further optimized as (Xue et al., 2024): where xij denotes the concentration of the jth component in the ith sample. gik is the loading of the ith sample on the kth factor. fkj is the contribution weight of the kth factor to the jth component. eij is the error term of the observation xij, i.e., the portion that the model fails to explain. The minimization fit quality parameter Q is used as a detection optimization metric and the objective function Q is defined as: The uncertainty directly affects the calculated weights of the component mass concentration data entered into the PMF model, calculated as follows (Zhang et al., 2019b). where Uij represents the uncertainty of the jth chemical species for the ith sample, Concentration represents the mass concentration of the analyzed component, MDL (Method Detection Limit) represents the detection limit of the analyzed component, and EF (Error Fraction) represents the error coefficient of the analyzed component, i.e., the proportion of error to the concentration of the analyzed component. The first term EF×Concentration2 represents the uncertainty associated with the concentration of the component, while the second term 0.5×MDL2 represents the uncertainty not associated with the concentration of the component. In this paper, the EF was set between 0.1 and 0.2. The uncertainty data are lprovided in the appendix. Ecological risk assessment indicators Sediment Quality Guidelines (SQGs) provide a scientific tool for assessing the adverse effects of sediment contaminants on aquatic organisms. The threshold effect concentration (TEC) and potential effect concentration (PEC), based on consensus, have been successfully employed for assessing the toxicity of PAHs in sediments (Chizhova et al., 2025; Long et al., 1995; Macdonald et al., 1996). The Mean Effect Range Median Quotient (MERM-Q) method is used to assess comprehensive ecological risks. This approach determines the combined ecological toxicity of PAHs by calculating the MERM-Q for each individual PAH component. The formula for MERM-Q is as follows: In this formula, ci represents the concentration of PAHs, ERMi represents the ERM value corresponding to PAHs,ci/ERMi represents the ERM-Q value corresponding to each individual PAHs, and n represents the number of PAHs types involved in the calculation. Results and analysis Concentrations and distributions of PAHs The concentration of PAHs does not exhibit a strong normal distribution and does not meet the requirements for Kriging interpolation. Based on statistical theory and incorporating inverse distance weighting with cross-validation, ArcMap software was employed to spatially predict the concentration distribution of 16 PAHs in Lingshui county lagoons, and to draw the spatial distribution contour lines of PAH concentrations in Lingshui county lagoons. The concentration distribution of 16 PAHs in surface sediments at different sampling sites in Lingshui county lagoons is shown in Fig. 2. The total PAH concentration ranged from 18.12 to 111.30 ng/g (dry weight), with an average concentration of approximately 60.83 ng/g. The total PAHs concentration in Xincun Lagoon ranged from 23.11 to 99.21 ng/g (dry weight), with an average concentration of 54.37 ng/g. As shown in Fig. 2, the distribution of PAH concentrations differs between the two lagoons. Xincun Lagoon generally exhibits a trend of increasing concentration from the shoreline toward the lake center. In the areas around points 37, 38, and 41, the average water depth is shallow, and the sediments contain a large amount of sand, gravel, snail shells, clam shells, and coral debris. The PAHs measured in the surface sediments are lower than those in the surrounding areas. In contrast, the total PAH concentration range in Lian Lake is 18.12–111.30 ng/g (dry weight), with an average concentration of 67.95 ng/g. Areas with high PAH concentrations account for a relatively larger proportion, indicating more severe pollution compared to Xincun Lagoon. The overall PAHs distribution in Lian Lake shows an increasing trend from the northeast to the southwest. The small lake on the southwest side is connected to Lian Lagoon by a small sluice gate. During high tide, seawater flows into the small lake, and during low tide, seawater flows out of the small lake. At low tide, the two water bodies are disconnected, with no exchange of water. Pollutant particles carried by seawater and pollutants emitted from surrounding residential areas settle and accumulate in the small lake, forming a sandy polluted area with relatively high PAH concentrations. Additionally, PAH concentrations at the mouths of Xincun Lagoon and Lian Lagoon exhibit a trend of initially increasing and then gradually decreasing. Based on grain size data, PAH concentrations in the surface sediments of Xincun Lagoon and Lian Lagoon generally show a negative correlation with grain size. Smaller particles have larger specific surface areas, providing more space for organic pollutants to adhere, and thus are more likely to adsorb organic pollutants. The distribution of PAHs in sediments is also influenced by organic matter. Distribution map of PAH concentrations in surface sediments of lagoons The measured range of 16 PAHs in lagoon sediments (18.12–111.30 ng/g dw) was significantly lower than values reported in some literature, such as sediments from the Euphrates River basin (Grmasha et al., 2023a) (5940–9727 ng/g dw), the Tigris River (Grmasha et al., 2023b) (5619–12,795 ng/g dw), and Lake Balaton (Grmasha et al., 2024a) (257–620 ng/g dw). Additionally, we investigated the proportion of PAH concentrations based on the number of benzene rings at different stations in the lagoons, as shown in Fig. 3. In this study, the primary composition pattern of PAHs in surface sediments was as follows: Lian Lagoon was dominated by hexacyclic PAHs, while Xincun Lagoon primarily contained tetracyclic, pentacyclic, and hexacyclic PAHs. Overall, the PAH composition followed the pattern of six-ring (52.43%) > four-ring (21.04%) > five-ring (16.73%) > three-ring (7.09%) > two-ring (2.71%), with the composition and proportion of individual PAHs being relatively similar across the entire study area. The distribution of low-molecular-weight (LMW) PAHs (≤ three rings) in sediments can indicate recent pollution, as LMW PAHs are more easily degraded than high-molecular-weight (HMW) PAHs (≥ four rings) and have lower lipophilicity. In contrast, HMW PAHs are more persistent and more readily adsorbed. As demonstrated in (Ambade et al., 2023) 2022 study, HMW PAHs exhibit strong adsorption properties toward soil. Most of the PAHs identified in this study belong to the HMW category (≥ 4 rings), which may be due to their greater persistence in soil. Furthermore, we found that HMW PAHs tend to accumulate more readily near pollution sources. Petroleum sources are rich in LMW PAHs, whereas pyrogenic sources exhibit the opposite pattern (Grmasha et al., 2023a). The higher content of HMW PAHs observed in this study suggests that PAHs in sediments likely originate primarily from pyrogenic sources. Proportion of PAH concentrations (a) and average proportion (b) of different numbers of benzene rings at each station in the lagoons Source analysis of PAHs The sources of PAHs in the environment are diverse, including industrial emissions from coal combustion and petroleum refining, unburned products in vehicle exhaust that contain high levels of PAHs, and the combustion of wood and natural gas, as well as natural sources such as forest fires and volcanic activity. To accurately assess the contribution of PAHs pollution sources in surface sediments of Lingshui county lagoons, this study will employ two methods: the characteristic compound ratio method and positive matrix factorization (PMF). Characteristic compounds ratio method The characteristic compounds ratio method, also known as the diagnosis ratio method, is a widely used analytical tool in environmental chemistry for idtntifying pollutant sources, assessing environmental behavior, and evaluating ecological risks, particularly in PAHs source determination (Soleimani et al., 2022). Four characteristic compound ratios are employed in PAHs analysis: IcdP/(IcdP + B[g,h,i]P), B[a]A/(B[a]A + Chr), Flo/(Flo + Pyr), and Ant/(Ant + Phe). Since Ant and Phe are prone to decomposition in natural environments, the first three characteristic compound ratios are primarily used as criteria for qualitative PAH source analysis. The criteria (Ali-Taleshi et al., 2021a; Soleimani et al., 2022; Yuan et al., 2021) are as follows: IcdP/(IcdP + B[g,h,i]P) < 0.20 indicates contamination from petroleum product leaks; 0.20 < IcdP/(IcdP + B[g,h,i]P) < 0.50 indicates contamination from the combustion of petroleum and other liquid fossil fuels; IcdP/(IcdP + B[g,h,i]P) > 0.50 indicates contamination from the combustion of wood and coal; B[a]A/(B[a]A + Chr) < 0.20, from petroleum-related substance leaks; 0.20 < B[a]A/(B[a]A + Chr) < 0.35, from the combined effects of petroleum-related substance leaks and combustion products; B[a]A/(B[a]A + Chr) > 0.35, from combustion processes; Flo/(Flo + Pyr) < 0.4, generated by diagenesis; 0.40 < Flo/(Flo + Pyr) < 0.50, combustion of liquid fossil fuels; Flo/(Flo + Pyr) > 0.50, derived from wood and coal combustion; Ant/(Ant + Phe) < 0.10, petroleum leakage; Ant/(Ant + Phe) > 0.10, combustion sources. The source apportionment of PAHs characteristic compound ratios is shown in Fig. 4. As shown in Fig. 4, the IcdP/(IcdP + B[g,h,i]P) ratio primarily ranges between 0.20 and 0.50, indicating that the combustion of liquid fuels such as petroleum is one of the primary pollution sources. Additionally, some station values exceed 0.50, further indicating that the combustion of wood and coal also contributes to pollution to some extent; The B[a]A/(B[a]A + Chr) ratio is mostly above 0.35, indicating that combustion processes are the primary pollution source; The Flo/(Flo + Pyr) ratio is generally below 0.40, attributed to diagenetic processes, with a few sites between 0.40 and 0.50, further confirming that the combustion of liquid fossil fuels is also one of the primary pollution sources; The Ant/(Ant + Phe) value is greater than 0.10 at all stations except Station 42, indicating that combustion processes are the primary source. In summary, PAHs pollution in the surface sediments of Lingshui county lagoons primarily originates from the combustion of fossil fuels, coal, and biomass such as wood, as well as oil spills. Source apportionment diagram of PAHs characteristic compound ratios PCA Principal component analysis was canducted using IBM SPSS Statistics 27 software. The Kaiser–Meyer–Olkin test yielded a value of 0.69, indicating the data’s suitability for factor analysis, and Bartlett’s sphericity test was passed. After rotation, the eigenvalues for the first four factors were 8.12, 2.41, 1.52, and 0.91, respectively, with cumulative frequencies of 54.13, 70.23, 80.39, and 86.48%. The eigenvalue of the fourth factor was below 1. However, defining only three factors would group PAHs with significantly different sources into the same factor. Combining this with the inflection point analysis of the scatterplot, defining four factors is more appropriate. These four factors collectively explain 86.48% of the sample variance. PC1 accounts for 54.13% of the variance, with representative PAHs including Nap, Acy, Chr, B[b]F, and B[k]F, potentially showing overlap between two components. PC2 accounts for 16.10% of the variance, with significant loadings for B[a]P, IcdP, and B[g,h,i]P. Both PC1 and PC2 exhibit substantial loadings for HMW PAHs. PC3 and PC4 account for 10.16% and 6.09% of variance, respectively, with LMW PAHs dominating their loadings. Variance explanations and rotated factor loadings for each component are presented in Annex 2 and 3. PMF Based on the number of factors determined by PCA, this study employed the USEPA PMF 5.0 model to conduct quantitative source apportionment analysis of 16 PAHs in surface sediment samples collected from 42 sampling sites in Lingshui County lagoons. The analysis utilized 4 to 6 pollutant components for model calculations, with the “Robust” mode implemented to minimize the influence of outlier data on model outcomes. The appropriateness of uncertainty settings for the 16 PAHs was assessed through residual diagnostics and error estimation (BS, DISP, and BS-DISP analysis). Rotational testing (Fpeak) was subsequently performed to optimize model outputs, confirming the validity and interpretability of the results. The signal-to-noise ratio (S/N) for PAH concentrations exceeded 2.90, indicating satisfactory data quality. The PMF model was conFig.d for 100 iterative calculations, with DISP error estimation ultimately determining five source component factors. During model implementation, D[a,h]A exhibited significant interference during integration and failed to meet operational requirements. Following extensive model validation, D[a,h]A was classified as “bad” and excluded from model calculations, resulting in the inclusion of 15 PAHs in the final analysis. Optimal model performance was achieved when uncertainty settings for Acy, Ant, Pyr, Flu, B[b]F, and B[k]F were designated as “weak”. Model outputs demonstrated robust convergence, with Q (Robust) = 930.80 and Q (True) = 1009.00 (ratio ≈ 1.08). This ratio, being less than 1.50 and approaching unity, indicates satisfactory model convergence. Residual analysis revealed that most individual PAH compound residuals fell within the range of − 3 to 3 and followed a normal distribution, supporting model reliability. Bootstrap analysis showed over 92% factor mapping consistency between bootstrap and base factors, while all DISP output values were 0, both meeting model specifications. No Fpeak correction was required for model outputs. Detailed BS and DISP model output data are provided in the supplementary materials. Different pollution sources have some characteristic indicator factors. This study clarifies the correspondence between different pollution sources and their characteristic components. Table 1 summarizes the characteristic components of the main sources of PAHs pollution. Table 1Identified components of major sediment sourcessource categoryIdentified componentsReferencesGasoline fuel sourceB[b]F, B[a]P, D[a,h]A, B[g,h,i]P(Wang et al., 2016; Yunker et al., 2002)Diesel fuel sourceIcdP, B[k]F(Sofowote et al., 2008; Xu et al., 2021)Biomass fuel sourceAcy(Khalili et al., 1995)Petroleum sourceNaP, Chr(Liu et al., 2009; Wang, 2017)Coke sourceFlu, Phe(Sofowote et al., 2008)Natural gas fuel sourceB[a]A, Chr(Ali-Taleshi et al., 2021b)Coal combustion sourceAce, Phe, Ant, Flo, Pyr, B[a]A(Ji et al., 2025; Wang et al., 2015)Commercial phenolic oilAce, Flu, Phe, Ant, Flo, Pyr(Kohler et al., 2000; Wang et al., 2013) Identified components of major sediment sources Five pollution source factors were successfully extracted for quantitative analysis of pollution source contribution rates. The contribution ratios of each pollution source in the factor contribution rates are shown in Fig. 5. In Pollution Source 1, the two substances Nap and Acy account for the largest contributions, at 82.42 and 66.41%, respectively. Nap is the primary indicator factor for petroleum pollution (Liu et al., 2009), while Acy primarily originates from wood combustion and is the primary indicator factor for biomass combustion sources (Khalili et al., 1995). Therefore, Pollution Source 1 is identified as a mixed pollution source from petroleum and biomass combustion. Petroleum pollution and wood combustion in residential areas around Points 27–43 may be the source of this factor. Petroleum pollution is particularly severe in the area extending outward from the mouth of the Xincun Lagoon, where a large number of vessels are engaged in tourism transportation and fishing activities. Consequently, oil spills from vessels cause petroleum pollutants to enter the water body. PMF source component spectrum In Pollution Source 2, Phe, Ant, Flo, Pyr, and B[a]A have the largest contribution rates, and they align with the pollution indicator factors for coal combustion sources (Ji et al., 2025; Wang et al., 2015). Along the lagoon’s shoreline, numerous stilt-dweller restaurants operate, consuming substantial amounts of coal during business hours. Additionally, coal may have served as the primary fuel source for residents’ daily lives in the past. Therefore, Pollution Source 2 is identified as a coal combustion source. Pollution Source 3 primarily contains B[a]P, B[g,h,i]P, and B[a]A, with a small amount of B[b]F. B[a]P and B[g,h,i]P have the highest contribution rates, at 49.17 and 51.52%, respectively. Both are pollution indicators for gasoline combustion sources (Wang et al., 2016; Yunker et al., 2002), so Pollution Source 3 can be identified as a gasoline combustion source. Vehicle traffic may be a contributing factor. In Pollution Source 4, IcdP has the largest contribution rate, reaching 85.65%. IcdP is typically an indicator factor for diesel combustion products (Sofowote et al., 2008; Xu et al., 2021), so Pollution Source 4 can be identified as a diesel combustion pollution source. At points 7, 14, and 27–29, IcdP concentrations were relatively high, possibly related to vessel traffic. In Pollution Source 5, Flu has the largest contribution rate, followed by Phe. Typically, Flu and Phe are indicator factors for coke sources (Sofowote et al., 2008), so Pollution Source 5 is identified as a coke source. Pollution source 6 primarily contains Ace, Phe, Ant, Flo, and Pyr. Commercial phenolic oil is produced through the distillation of coal tar and contains up to 85% PAHs, with 16 PAHs accounting for 20–40% of the total weight of typical commercial phenolic oil. Among these, Ace, Flu, Phe, Ant, Flo, and Pyr are the primary PAHs components in commercial phenolic oil (Kohler et al., 2000; Wang et al., 2013). Therefore, Pollution Source 5 can be identified as commercial phenolic oil. After normalization, the average contribution rates of the five factors were as follows: petroleum sources and biomass fuel sources accounted for 18.74%, coal combustion sources accounted for 24.86%, gasoline combustion sources accounted for 23.96%, diesel fuel sources accounted for 19.41%, and commercial phenolic oil accounted for 13.03%. The transportation hubs at Xincun dock and the Monkey Island tourist pier at the estuary entrance experience heavy traffic. Considering the proportion of pollution source contributions, transportation-related emissions may be a primary cause of PAHs pollution in Lingshui County’s lagoon. Ecological risk assessment of PAHs Ecological risk of individual PAHs in surface sediments Based on the low effect range (ERL) and medium effect range (ERM) established by Long and Morgan (Long et al., 1995), in conjunction with the sediment quality guidelines (SQGs), the potential ecological risks were evaluated. These guidelines are derived from threshold concentrations for individual and total PAHs. Exceeding these concentrations may result in adverse effects on organisms inhabiting the sediments. Furthermore, the Threshold Effect Level (TEL) and Potential Effect Level (PEL) developed by (Macdonald et al., 1996) using expanded datasets,were emplyed to further elucidate the significance of these benchmark values. The ERL and TEL represent median values at which 10 and 15% of tested sediment samples, respectively, are expected to induce adverse effects, indicating concentrations where adverse biological effects are seldom observed (Chizhova et al., 2025). Consequently, concentrations below the ERL/TEL are unlikely to pose ecological risks and have minimal impacts on organisms. The ERM and PEL correspond to the 50th percentile of effect data, defining concentrations at which adverse effects are expected to occur frequently. PAH concentrations exceeding the ERM/PEL suggest frequent ecological risks and significant toxic effects on organisms, whereas concentrations between the ERL/TEL and ERM/PEL indicate occasional risks of moderate to low toxic effects. Using those methodologies, numerous studies have extensively assessed the environmental hazards posed by organic pollutants in marine sediments (Duker et al., 2024; Sreedevi & Harikumar, 2025). The benchmark values and assessment results for the ecological risk evaluation of individual PAHs are presented in Table 2. Table 2Benchmark values and assessment results for the ecological risk assessment of individual PAHsPAHsERLERMTELPELConcentration rangeBelow ERL/TELNap160210034.63910.34 ~ 5.07√Acy446405.87128ND ~ 0.36√Ace165006.7188.90.15 ~ 2.58√Flu1954021.2144ND ~ 1.49√Phe240150086.75440.36 ~ 7.82√Ant853110046.92450.23 ~ 2.95√Flo60051001131494ND ~ 8.20√Pyr665260015313980.30 ~ 11.99√B[a]A261160074.8693ND ~ 9.12√Chr38428001088460.48 ~ 15.56√B[b]F3201800––0.80 ~ 11.72√B[k]F2801620––0.61 ~ 8.88√B[a]P430160088.8763ND ~ 6.83√IcdP––––ND ~ 8.49–D[a,h]A63.42606.221353.37 ~ 100.63 × B[g,h,i]P4301600––ND ~ 10.12√ Benchmark values and assessment results for the ecological risk assessment of individual PAHs The results of the ecological risk assessment of individual PAHs in surface sediments from 42 sampling sites in Lingshui County lagoons indicate that the concentrations of all 16 PAHs at these sites were below the ERM/PEL thresholds, suggesting a low risk of significant ecological toxicity in the lagoon. The concentrations of Nap, Acy, Ace, Flu, Phe, Ant, Flo, Pyr, B[a]A, Chr, B[b]F, B[k]F, B[a]P, and B[g,h,i]P were all below the ERL/TEL valus at all sampling sites, indicating no ecological risk associated with these 14 PAHs in the lagoon. D[a,h]A concentrations fell between the ERL/TEL and ERM/PEL ranges, suggesting moderate to low toxic effects of D[a,h]A in the study area. Based on D[a,h]A concentration data, ecological risk differences were observed between Xincun Lagoon and Lian Lagoon. D[a,h]A concentrations in Xincun Lagoon were all below the ERL, posing minimal ecological hazards. However, at sampling points 6, 7, 9, 10, 11, 12, 13, 16, and 20 in Lian Lagoon, D[a,h]A concentrations exceeded the ERL, indicating moderate to low toxicity effects on organisms at these locations. The presence of D[a,h]A may be associated with high-temperature combustion from nearby residential areas and emissions of coal tar residues. Combined with PMF factor analysis, these findings align with local activity characteristics. The single sampling event (conducted April 8–10, 2025) limits temporal generalization of the results. This study did not provide ERL/TEL and ERM/PEL values for IcdP, which exhibits potential toxicity to biological organisms. IcdP was detected at 34 of the 42 sampling points in the lagoons, indicating that approximately 81% of the study area may harbor potential ecological risks due to IcdP. Comprehensive ecological assessment of PAHs in surface sediments Relying solely on ecological risk assessments based on individual PAHs is insufficient; a comprehensive ecological risk assessment of PAHs is also necessary to further clarify the ecological risk situation. McCready et al., (2006) further supplemented the aforementioned methods with a quantitative approach for analyzing the combined toxicity risk of multiple PAHs in sediments, known as the Mean Effect Range Median Quotient (MERM-Q) method. This method determines the comprehensive ecological toxicity of PAHs by calculating the MERM-Q for each individual PAH component. The toxicity effect levels and probabilities of toxicity corresponding to the comprehensive ecological toxicity values are presented in Table 3. Table 3Toxicity corresponding to MERM-Q values (Tang et al., 2017)MERM-Q rangeToxicity levelPossibility of toxicityMERM-Q < 0.1low toxicityless than 10%0.1 < MERM-Q < 0.5moderately low toxicityaround 30%0.5 < MERM-Q < 1.5moderate to high toxicityaround 50%MERM-Q > 1.5high toxicityaround 75% Toxicity corresponding to MERM-Q values (Tang et al., 2017) The MERM-Q value for a given site was calculated by dividing the sum of each PAH concentration divided by its corresponding ERM value by the number of PAHs included in the calculation. The MERM-Q values of PAHs at 42 sampling sites in Lingshui County lagoons are presented in Table 4. The results indicate that the MERM-Q values of PAHs in surface sediments at all sampling sites in the study area are far below 0.1, suggesting that the probability of bioaccumulative toxicity effects occurring in this region is less than 10%. Therefore, the potential ecological risk is relatively low. The D[a,h]A outlier data had minimal impact on the MERM-Q calculation and did not alter the qualitative final result. The likelihood of individual ecological risks in the lagoon area is higher than that of comprehensive ecological risks. Furthermore, the contribution rates of seven carcinogenic CPAH monomers to the total MERM-Q are as follows: D[a,h]A (89.12%), B[a]A (0.99%), Chr (0.96%), B[b]F (1.52%), B[k]F (1.13%), B[a]P (0.82%), and B[g,h,i]P (0.70%). The industrial sector should prioritize degradation methods for these seven carcinogenic PAHs (Grmasha et al., 2024b). Compared to the Euphrates River basin (Grmasha et al., 2023a), the Tigris River (Grmasha et al., 2023b), and Lake Balaton (Grmasha et al., 2024a) the ecological risks in these lagoons are at a low level. Table 4MERM-Q values of PAHs at 42 sites in Lingshui County lagoonsStation123456789MERM-Q0.0020.0030.0100.0090.0130.0170.0180.0070.019Station101112131415161718MERM-Q0.0260.0180.0180.0200.0160.0090.0200.0120.011Station192021222324252627MERM-Q0.0020.0220.0130.0020.0050.0040.0080.0060.007Station282930313233343536MERM-Q0.0060.0080.0050.0040.0070.0070.0070.0040.005Station373839414243MERM-Q0.0040.0040.0050.0030.0050.004 MERM-Q values of PAHs at 42 sites in Lingshui County lagoons Limitations and prospects This study examined only the seasonal distribution of PAH concentrations during spring. Since PAH concentrations are influenced by seasonal variations, the findings cannot be generalized to annual PAH levels in 2025, indicating temporal limitations. The PMF model results remain subject to certain subjectivity and limitations due to variable selection, model structure, and data uncertainties. The PMF allocation results presented here were preliminary and seasonally constrained. We employed indicators such as ERL/ERM, TEL/PEL, and MERM-Q for ecological risk assessment. These empirical SQGs have known probabilistic interpretation and regional transferability caveats. The findings of this study provide only a preliminary understanding of the spatial distribution characteristics and sources of PAHs in surface sediments of the Lingshui County lagoon in Hainan. Further in-depth research is required for ecological risk assessment. To improve the lagoon’s water environment, future efforts should integrate sediment PAH studies with environmental governance and pollution control. Exploring more effective pollution reduction measures will provide scientific basis for environmental improvement in similar lagoon areas. Conclusions This study investigated the concentrations, potential sources, and ecological risks of polycyclic aromatic hydrocarbons (PAHs) in springtime surface sediments from the Lingshui County lagoon in Hainan. Spring PAH concentrations ranged from 18.12 to 111.30 ng/g (dw), with a mean concentration of 60.83 ng/g. The detected PAHs were predominantly tetracyclic and hexacyclic compounds. PAH concentrations were elevated in the central region and mouth of the lagoon, potentially associated with hydrodynamic conditions and sediment characteristics. Comparative analysis revealed higher sediment contamination in Lian Lagoon compared to Xincun Lagoon. When compared with global lagoon systems, PAH levels in Lingshui Lagoon surface sediments were relatively low.The characteristic compounds ratio method and PMF model analysis indicate that the primary pollution sources in the lagoon are pyrogenic in origin, including petroleum, biomass, and coal combustion. Characteristic compounds ratio method results were consistent with PMF analysis findings. However, these source attribution results should be considered preliminary and season-specific. The PAH concentrations in Lingshui Lagoon sediments indicated moderate to low pollution levels. D[a,h]A exhibited potential ecological risk, though concentrations remained below established environmental thresholds. Nevertheless, local authorities should maintain vigilance regarding its potential environmental impacts. This research provides critical evidence for developing science-based management strategies and enhancing regional pollution prevention and control measures. Future investigations will examine seasonal variations in PAHs distribution patterns in lagoon sediments, with particular attention to climatic influences, hydrodynamic conditions, and sediment properties, to elucidate the effects of external environmental factors on PAHs distribution dynamics. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (DOCX 32 KB) Supplementary file2 (XLSX 19 KB) Acknowledgements The study was funded by the Pioneer and Leading Goose + X" S&T Program of Zhejiang (2025C02016), and supported by "Pioneer" and "leading Goose" R&D program of Zhejiang Province (2023C03130). Author contributions Bingxin Zhong:Performed the data analysis and wrote the manuscript. Nina Ma: Validated the model results. Baijuan Yang: Provided GC–MS instrument technical guidance. Zhifu Wang: Revise the paper and provide financial support. Weihua Feng: Proposed the plan and directed the article. Baoxing Huang,Hengtao Xu and Duian Lv : Reviewed the article. Data availability The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author. Declarations Conflict of interest The authors declare that they have no competing interests. The corresponding author affirms that all of the authors have read and approved the manuscript. Ethical approval The authors declare their responsibility for any ethical issues that may arise after the publication of this manuscript. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary file1 (DOCX 32 KB) Supplementary file2 (XLSX 19 KB) Data Availability Statement The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.