Research Studies
Study Tracker
Characterization of fluorine pollution and evaluation of multi-media environmental health risk in Daihai watershed: Based on coupled uncertainty-sensitivity modeling.
Abstract
Full-text study online at
https://www.sciencedirect.com/science/article/pii/S0147651325017828?via%3Dihub
Highlights
- Fluoride pollution in Daihai threatens ecosystems and human health.
- Multi-source analysis reveals diverse fluoride migration pathways.
- Daihai fluoride is unstable, with high mobility and bioavailability.
- Health risk model with uncertainty analysis gives probabilistic risk values.
Excessive fluoride levels have become a critical concern in contemporary society, with China ranking among the nations most severely affected by fluoride pollution. To investigate the health and ecological risks associated with excessive fluoride, 80 surface water samples, 80 sediment samples, 152 groundwater samples from rural drinking wells surrounding the lake, and 64 tributary samples flowing into Lake Daihai were consecutively collected between 2023 and 2024. A comprehensive spatiotemporal analysis of fluoride concentrations in Lake Daihai’s water body, sediments, and surrounding groundwater was conducted using ArcGIS and Origin software. The potential ecological risks of Lake Daihai’s surface water and sediments were evaluated using the integrated pollution index method, the entropy method, and the modified Nemero index method. A human health risk model was applied to evaluate the health risks of groundwater around Lake Daihai. Monte Carlo uncertainty analysis and the Crystal Ball sensitivity analysis system were combined to estimate the probability of exceeding non-carcinogenic risk thresholds and to identify sensitivity levels for fluoride-related ecological risks across different population groups. The results indicate that the average fluoride concentration in Lake Daihai’s water bodies is 6.22 mg/L, while groundwater in surrounding rural drinking wells ranges from 0.11 to 3.78 mg/L, exceeding the national average of 0.66 mg/L. Surface water fluoride concentrations exhibited an overall seasonal pattern of spring increase, summer decrease, and autumn rebound. Sediment fluoride concentrations showed a gradient pattern of higher in the northwest, lower on the east/southwest shores. Groundwater fluoride concentrations generally followed a pattern of higher on the east/west shores, lower on the north shore. The surface water bodies of Lake Daihai face potential ecological risks, with over 99 % of samples exhibiting a hazard quotient (HQ) greater than 1 and an average HQ of 3.12. This finding indicates a state of moderate risk and high instability, characterized by potential mobility and bioavailability. The ecological risk of fluoride in sediments is generally classified as mild to moderate contamination. For groundwater samples, the non-carcinogenic risk threshold exceeded standards in 85 % of infants (THQ > 1.0), 60 % of adults, and 30 % of children and adolescents. Uncertainty model analysis indicates non-carcinogenic health risk exceedance probabilities of 74.25 % for infants, 28.18 % for children, 24.82 % for adolescents, and 42.47 % for adults, with infants facing the highest non-carcinogenic fluoride risk. Sensitivity analysis reveals F– contribution as the primary factor for non-carcinogenic risks in infants (73.9 %) and children (30.6 %).
Keywords
Fluoride overload
Health risk models
Entropy weight method
Non-carcinogenic health risks
Occurrence pattern
1. Introduction
In the late 19th century, the relationship between fluoride and human health was first identified. Fluoride exposure at concentrations of 0.5–1.5 mg/L was observed to promote bone and tooth development and has been reported to provide potential benefits in the treatment of microbial infections, inflammation, cancer, kidney stones, and other conditions (Sharma et al., 2017). However, as research advanced, the health risks associated with fluoride gradually became evident. In the 1930s, health risks associated with fluoride began to receive scientific attention. Churchill and his team were the first to demonstrate that variations in fluoride concentrations in drinking water could influence dental health. In 1932, Moller and Gudjonsson (1932) documented the first cases of occupational fluorosis in workers at a Swedish cryolite manufacturing plant. In 1971, the World Health Organization released the report “ Fluoride and Human Health ”, which proposed a maximum acceptable concentration of fluoride in global drinking water of 1.5 mg/L (with China’s standard being 1.0 mg/L). Fluoride contamination has emerged as a global public health concern. Approximately 50 countries worldwide have reported cases of fluorosis, placing over 200 million people at risk. Fluoride contamination primarily originates from two sources: natural processes, such as the weathering of mineral rocks and the leaching of fluoride from soil, and human activities, including industrial discharges from facilities such as aluminum plants. Kudo et al. (1987) conducted a comparative analysis of fluoride levels in five rivers in the French Alpine region and identified industrial effluent from aluminum plants as the primary source of fluoride pollution in these waterways. In a spatiotemporal study of fluoride in the Yongding River, Wang et al. (2020) analyzed the spatiotemporal characteristics of fluoride concentrations in the Yongding River and conducted a risk assessment, concluding that mineral rocks are the natural source of fluoride in the river, whereas the discharge of industrial waste is a major cause of fluoride pollution.Additionally, some scholars have carried out in-depth investigations into the spatial distribution and hydrogeochemical mechanisms. Kitalika et al. (2018), through analyzing the correlations between river fluoride concentrations and ions in different seasons, as well as hydrochemical characteristics and mineral lithology, revealed that the variability of fluoride-bearing minerals leads to different spatial distributions of fluoride, and that the seasonal differences are mainly controlled by groundwater recharge and rainy-season runoff. Wen et al. (2013), based on the regional climatic characteristics of arid and semi-arid areas in northern China, discussed the hydrogeochemical mechanisms of fluoride enrichment and provided suggestions for future water resources management.At the same time, the occurrence forms of fluorine in soils and sediments have also received attention.Data from the China National Environmental Monitoring Centre (CNEMC) indicate that soil fluoride concentrations across China range from 50 to 3467 mg/kg, with a mean value of 478?mg/kg (95 % CI: 191–1012 mg/kg). This average is notably lower than the fluoride levels reported for Moroccan soils, which range from 410 to 1263 mg/kg and have a mean of 785 mg/kg (Edmunds and Smedley, 2012).In terms of microscopic occurrence forms, Liu et al. (2023) found that the content relationship of occurrence forms in fluoride-bearing sediments in Wuliangsuhai, Inner Mongolia, China, was exchangeable fluorine < water-soluble fluorine <Fe/Mn oxide-bound fluorine <organically bound fluorine <residual fluorine.
Despite the significant progress in understanding the distribution patterns and risk assessment of fluoride pollution, several areas still warrant further investigation. Most studies have focused on a single medium (surface water or groundwater), whereas systematic and multi-media coupled research on fluorine cycling in closed basins remains scarce.For example, in their study on Wuliangsuhai in Inner Mongolia, Liu et al. (2023) analyzed the different occurrence forms of fluorine, but the quantitative contribution of these sediment forms to the surface water fluoride concentration has not yet been clarified. In another closed-basin study, Lafayette et al. (2020) revealed covariate relationships between fluoride and arsenic contamination in shallow groundwater and sediments at Laguna El Cuervo, Mexico, using principal component analysis, yet the study lacked explicit multi-method integration and uncertainty assessment. Qu et al. (2025) investigated fluoride enrichment and migration processes in a typical closed lake (Daihai Lake); however, systematic quantitative analyses of the reciprocal feedback mechanisms governing fluoride migration among surface water, groundwater, and sediments remain underreported. Therefore, adopting a systematic approach to closed lakes-one that comprehensively considers material transport relationships among surface water, sediments, and surrounding groundwater; integrates multi-method assessments (e.g., indicator systems, multi-medium fluoride pollution risk models); incorporates uncertainty analyses and sensitivity risk probabilities; and quantifies risk probabilities and sensitivities—remains an urgent research priority that requires further advancement.
This study attempts to explore the following aspects. First, by selecting the slightly saline lake in the Daihai Lake basin of Liangcheng County, Inner Mongolia, and the surrounding rural drinking water wells as the investigation subjects, a comprehensive spatiotemporal analysis of fluoride concentrations in Daihai Lake water, sediments, and surrounding groundwater was carried out using ArcGIS and Origin software. Second, combined with the comprehensive index analysis method, entropy method, potential ecological risk method, and health risk model, a multi-media comprehensive evaluation of the surface water of Daihai Lake and surrounding groundwater was conducted. Subsequently, Monte Carlo uncertainty analysis and Crystal Ball sensitivity analysis were introduced to enhance the reliability of risk assessment, with particular attention to health risk differences among different age groups (infants, children, adolescents, and adults) and the quantification of non-carcinogenic risk probability for the infant population. Finally, the multi-media migration and feedback mechanisms of fluoride pollution in the Daihai Lake watershed were analyzed. This work aims to provide new perspectives for the systematic study of the health impacts and environmental behavior of fluoride and to provide a reference for developing more targeted pollution prevention and control strategies.
2. Materials and methods
2.1. Overview of the study area
The Daihai Lake basin is a narrow faulted basin, with a major axis of about 45 km and a minor axis of about 14 km. It was formed by crustal uplift and subsidence movements during the Pliocene to Quaternary period and developed into a lake in the early Pleistocene. Daihai Lake (40°29’~40°37’N, 112°33 ~112°46’E) is located in Liangcheng County, Ulanqab City, Inner Mongolia Autonomous Region. It is the third largest inland lake in Inner Mongolia, a typical inland closed lake, and also a typical saline lake, undertaking important functions of ecological regulation, soil and water conservation, and water conservation in the Inner Mongolia Autonomous Region. There are 22 main inflow river channels around Daihai Lake, of which 8 carry water, namely the Muhua River, Tiancheng River, Gongba River, Wuhao River, Buliang River, Yuanzi Gully, Gechou Gully, and Shiyao Gully. Surface runoff, groundwater runoff, and precipitation recharge are the main sources of water supply for Daihai Lake, and water surface evaporation is currently the primary discharge pathway. Since the mid-1970s, the climate has entered a warm and arid period. Due to continuous drought and persistent human interception of inflow water, the lake surface area has continued to shrink rapidly, from 200 km2 in the 1950s to 45 km2 in 2023. The storage capacity decreased to 519 million m3, with an average water depth of about 6 m (Fig. 2S). With the reduction of lake volume, solutes continued to concentrate, accelerating the salinization process and causing fluoride concentrations to continue to rise. Under the dual pressures of water resource shortage and harsh meteorological conditions, the water quality and aquatic ecological security of the lake are facing unprecedented challenges.
2.2. Sample collection and testing methods
According to the Technical Specifications for Surface Water Monitoring (HJ-T91), 10 sampling points (SW1-SW10) were set in the lake to collect surface water and sediment samples. Based on the distribution of rural drinking water wells around Daihai Lake, 19 groundwater sampling points (GW1-GW19) were selected to collect groundwater samples. The layout of the sampling points is shown in Fig. 1.
Fig. 1. a.Spatial distribution map of improved drinking water fluoride concentration in fluorosis areas of China; b.Topographic map of the study area; c.Locate the study area; d.Sampling point map.
Between April and November 2023, samples were collected on the 15th day of each month, with a total of 80 surface water samples, 80 sediment samples, 152 groundwater samples from surrounding rural drinking water wells, and 64 tributary inflow river samples being collected. A handheld GPS device (Model: GARMIN EXTRA 10) was used to precisely locate the sampling sites. A stratified water sampler was employed for the collection of surface water, surrounding groundwater, and tributary samples, while a grab-type sediment sampler was used to collect lake sediment samples. The freshly collected sediments were naturally air-dried at room temperature in the laboratory, then ground and passed through a 100-mesh (0.149 mm) sieve. After removing residual plant debris and stones, the air-dried sediment samples were stored in sealed polyethylene bags for subsequent analysis.
The fluoride concentration in the water was measured using the electrode method, and the electrode used was a Mettler perfectION composite F (fluoride) ion electrode. The water sample was filtered through a 0.45 um glass fiber membrane, 5?ml of filtered water was taken, an ionic strength adjuster was added, and the fluoride ion electrode was inserted to measure the fluoride ion concentration in the water. For the fluoride forms in sediments, a sequential extraction method was used for determination, and each sample was tested three times and the average of the three was taken as the final result (the error of the three analytical results was <5 %). In addition, before starting a new batch each time, the fluoride ion concentration in a blank sample (deionized water) was measured first and subtracted from the actual data afterwards.
2.3. Methodology for groundwater risk assessment in and around Lake Daihai
2.3.1. Ecological risk assessment
Ecological risk assessment has received widespread academic attention as an effective approach for evaluating the extent of damage caused by harmful elements to the natural environment. In the 1980s, the U.S. Environmental Protection Agency (USEPA) introduced the concept of ecological risk based on health risk assessment and simultaneously proposed methods for identifying risk sources and receptors. In the 1990s, the USEPA developed an ecological risk assessment framework, which was subsequently updated and expanded in 1998. A variety of ecological risk assessment methods are currently employed, including the entropy method, the integrated pollution index method (Marselina et al., 2025), the improved Nemerow index method (Hu et al., 2023), and the potential ecological risk assessment methodology. Each ecological risk assessment method has distinct advantages and limitations. To obtain the most accurate evaluation results, a comprehensive assessment should be conducted by integrating multiple methods for comparative analysis.
(1) Entropy weighting method
Ecological risk assessment models include the entropy method, the SSDs model, and the PERA model, among others. Data availability is one of the key factors in selecting an assessment model (Del Signore et al., 2016). Among these, the entropy method requires relatively limited toxicological data and provides strong operational feasibility (Liu et al., 2022). Therefore, the entropy method was employed to quantitatively characterize the fluoride pollution status of surface water in Lake Daihai, with the model formula expressed as follows:(1)HQ=EC/WQC where HQ denotes the hazard quotient, EC represents the fluoride exposure concentration, and WQC signifies the water quality criterion, defined as the chronic benchmark value. In this study, a value of 2 mg/L was adopted based on the research of Okkerman et al., 1991. Owing to the absence of a standard HQ value for fluoride, the HQ value for zinc—determined according to the pollutant concentration limits under different water quality conditions specified in the Surface Water Environmental Quality Standard (GB 3838-2002)—was used as the reference standard in this study. When the fluoride exposure concentration exceeds the water quality standard, potential ecological hazards may occur. The specific HQ ranges are defined as follows: HQ <0.1, 0.1 <HQ <1, 1 <HQ <10, and HQ >10, corresponding to no risk, low risk, medium risk, and high risk in the assessment area, respectively (Liu et al., 2022).
(2) Integrated Pollution Index method
The Comprehensive Pollution Index is a widely used and comparable evaluation tool. It calculates pollution indices for individual pollutants and applies statistical methods to integrate them into a composite water quality index. This approach not only considers the independent impact of each pollutant but also reflects the relative importance of groundwater in the overall water quality status through weight allocation. The Comprehensive Pollution Index (P) is expressed as:(2)Pi=CF/SF(3) where Pi denotes the single-factor pollution index of the pollutant, CF represents the fluoride content detected in Lake Daihai sediments, and SF is the evaluation standard for fluoride pollutants (425 mg/kg) (Chen et al., 1991). The grading criteria for the single-factor pollution index are defined as follows: The single-factor pollution index method classifies pollution indicators into four levels. When the single-factor pollution index (Pi) <1, the pollution level is classified as unpolluted; 1 <Pi <2 indicates mild pollution; 2 UPi <3 indicates moderate pollution; and Pi >3 indicates severe pollution.
Where P denotes the comprehensive pollution index, and n represents the number of evaluation items. The classification criteria for the comprehensive pollution index are defined as follows: When P<1, water quality is considered slightly polluted; 1 <P <1.5 indicates mild pollution; 1.5 <P<2 signifies moderate pollution; 2 <P the single-factor pollution<P <2.5 denotes relatively severe pollution; and P > 2.5 indicates severe pollution.
(3) Improved Nemerow index method
The traditional Nemero index method often overemphasizes the influence of a single dominant pollutant when assessing fluoride contamination in sediments, while failing to adequately account for the relative weight of each pollutant in the overall water quality evaluation. In contrast, the improved Nemero index method comprehensively incorporates both the concentration of each pollutant and its relative importance in water quality assessment. This approach avoids biases in sediment quality evaluation that may arise from neglecting fluoride pollutants, which pose significant hazards despite occurring at lower concentrations. This method more accurately reflects the degree of fluoride contamination in sediments, thereby providing a more scientific and comprehensive basis for sediment management and pollution control (Hu et al.,2023). The calculation formula is given as follows:(4)
(5)(6)Where denotes the correlation ratio of the i-th pollutant factor;
represents the standard value of each pollutant factor; smax is the standard value of the ith pollutant factor; m indicates the number of evaluated pollutant factors, and F– corresponds to the F value of the pollutant factor with the highest weight. P denotes the Nemero comprehensive pollution index, with the grading criteria defined as follows: when P???0.7, the grade is I, indicating a safe pollution level; 0.7 < P???1.0, Grade II, caution; 1.0?<?P???2.0, Grade III, mild pollution; 2.0 <?P???3.0, Grade IV, moderate pollution; and P > 3.0, Grade V, severe pollution.
(4) Potential ecological risk assessment methodology
Potential ecological risk assessment of sediment fluoride occurrence forms follows the stability risk assessment code (RAC) proposed by Singh et al. (2005). When the stability coefficient (SAC, defined as the ratio of water-soluble and exchangeable fluoride to total extractable fluoride) is <1 %, it is extremely stable; 1 % <SAC <10 % indicates stable; 10?% <SAC <30 % indicates moderately stable; 30 % <SAC <50 % indicates unstable; SAC > 50 % indicates highly unstable.
Where ERb denotes the risk index and Tr represents the fluoride toxicity response factor. In this study, fluoride was treated as a single factor, with fluoride abundance values serving as corrected abundance values, yielding a Tr value of 1. Cbio, CPbio, and CNbio denote the measured fluoride concentrations (mg/kg) for the three fluoride species at each sampling point, while w0 represents the reference value for total fluoride content in sediments (mg/kg). The fluoride content evaluation standard value of 425 mg/kg was adopted in this study. The ecological risk classification criteria are summarized in Table 1 (Pan et al.,2018).
Table 1. Ecological grade division standard.
| Fluoride content (mg/kg) | ER | Pollution level |
|---|---|---|
| wi <4 2 5 | ER <0.6 | pollution-free |
| 425 <wi <800 | 0. 6 <ER <1 | Slight pollution |
| 800 <wi <1600 | 1 <ER <2 | Moderate pollution |
| wi > 1600 | 2 <ER <4 ER >4 |
Highly polluted Heavy pollution |
2.3.2. Health risk assessment methods
This study employed the human health risk model proposed by the U.S. Environmental Protection Agency (USEPA, 1989) to quantify the health risks associated with fluoride contamination. Fluoride poses non-carcinogenic risks to humans primarily through oral ingestion and dermal absorption. The risk assessment was based on fluoride concentrations in drinking water for four distinct age groups. Owing to differences in behavioral and physiological characteristics among age groups, the population was categorized into infants (1–3 years), children (4–10 years), adolescents (11–19 years), and adults (20–70 years). The average daily dose (ADD) of F– was calculated using simulated doses and exposure parameters for these four age groups within the study population.(8)ADDing(mg/kg-body weight/day)=?(C×IR×EFing×EDing)/(BW×ATing)(9)ADDder (mg/kg-body weight/day)=?(C×SA×KP×L×ETder×EFder×EDder×0.001)/(BW×ATder)(10)HQing=ADDing/RfDing(11)HQder=ADDder/RfDder
Where HQing and HQder represent the hazard quotients for ingestion and dermal exposure pathways, respectively, and THQ denotes the total noncarcinogenic hazard quotient.(12)THQ=HQing+HQder
When THQ = 1, it represents the threshold for non-carcinogenic risk, whereas THQ >1 indicates that non-carcinogenic substances pose an unacceptable risk to human health. When THQ <1, the risk posed by non-carcinogenic substances to human health is considered acceptable. The ADD and HQ values were calculated using parameters from four different age groups in the study population (Table S1) (Botle et al.,2020).
3. Results and discussion
3.1. Fluoride concentration levels and statistical differences in groundwater around Daihai Lake
Fluoride concentrations in Lake Daihai during spring (April–May), summer (June–August), and autumn (September-November) ranged from 4.79 to 7.66?mg/L, 5.43–7.47?mg/L, and 5.38–6.30?mg/L, respectively. The corresponding mean values were 6.04, 6.67, and 5.84?mg/L, all exceeding the Class V threshold (1.5?mg/L) specified in the Surface Water Environmental Quality Standard (GB 3838-2002) (Table S2). Fluoride concentrations in groundwater ranged from 0.11 to 3.78?mg/L, with a mean value of 1.13?mg/L. This concentration exceeded the national mean (0.66?mg/L) (Edmunds and Smedley,2012), and some sites surpassed China’s guideline value for drinking water (1.0?mg/L), although it remained below the World Health Organization (WHO, 2011) guideline limit of 1.5?mg/L (Table S3). Compared with other regions, the maximum fluoride concentration in groundwater in this area was relatively moderate. The mean fluoride concentration in surface sediments was 808.08?mg/kg, with residual fluoride as the predominant form, accounting for 91.63–94.50?%. This was followed by water-soluble fluoride (3.09–4.87?%), organically bound fluoride (0.52–1.03?%), fluoride bound to iron–manganese oxides (0.24–0.77?%), and exchangeable fluoride (0.64–1.46?%). Fluoride concentrations in sediments were 2–3 orders of magnitude higher than those in surface water, indicating that sediments act as a significant potential reservoir for fluoride.
Fig. 2. a.Spatial Distribution of Fluoride Concentrations in Surface Water;b.Spatial Distribution of Fluoride Concentrations in Sediments;c.Spatiotemporal Distribution of Fluoride Concentrations in Surface Water and Surrounding Groundwater;d.Spatiotemporal Distribution of Fluoride Concentrations in Sediments.
Variance analysis results indicated that fluoride levels differed significantly (p?<?0.01) or highly significantly (p?<?0.001) among the four media (Fig. 3S). Sediments exhibited the highest fluoride content, functioning as an important accumulation reservoir in the region (Liu et al., 2023). Surface water showed the second-highest concentrations, regulated by both external inputs and sediment release (Tang et al., 2021). Groundwater had the lowest fluoride levels overall; however, specific sampling sites such as GW3, GW12, and GW17 exhibited localized enrichment, indicating the presence of high-value zones. Tributary waters showed the lowest total fluoride concentrations, but slight enrichment at individual points suggested that they act as source-input corridors and transitional pathways, facilitating transport rather than storage during fluoride migration.In surface water (Fig. 4S), SW6 showed the highest levels, followed by SW3 and SW4, while SW1, SW2, and SW10 had the lowest. Significant differences (p?<?0.01) were observed between SW3 and SW1, SW2, SW10; between SW6 and SW1, SW2, SW10; and between SW4 and SW10. Regarding groundwater (Fig. 5S), GW3 (downstream runoff zone), GW12 (southeastern urban-industrial composite zone with lake-groundwater interaction), and GW17 (southwestern relatively enclosed zone) were subject to strong inputs and prolonged residence time due to human discharge, evaporation, and recharge accumulation. These sites showed highly significant differences compared with GW6, GW9, GW10, GW15, and GW18, which are located in recharge corridors with high water exchange rates.In sediments (Fig. 6S), SW3 is located in the west-southwest inflow deceleration zone and a semi-enclosed bay area, characterized by fine particle enrichment, strong adsorption/co-precipitation, and low water exchange. Consequently, it showed highly significant differences compared with all other sampling points (Horowitz, 1991). Overall, the significant and highly significant spatial differentiation of fluoride distribution in Daihai essentially reflects the spatial opposition between external input intensity and hydrodynamic dilution capacity, and this mechanistic contrast is more stable than internal water body fluctuations.
Fig. 3. a.PCA ordination of fluoride and environmental parameters across different water and sediment media; b.Redundancy Analysis Ranking Diagram of the Relationship Between Fluoride Speciations in Daihai Sediments and Surface Water Fluoride Concentrations; c.Correlation Heatmap of Sediment Fluoride Forms, Physicochemical Parameters, and Surface Water Fluoride Concentrations.
Fig. 4. a. Daihai groundwater single-factor risk index;b. Daihai surface water fluorine ecological risk boxplot.
Fig. 5. Fluoride Health Risks in Groundwater of the Daihai Lake Basin.
Fig. 6. Uncertainty modelling results for total non-carcinogenic risk in groundwater around Daihai Lake.
3.2. Spatio-temporal evolution of fluoride in groundwater at Lake Daihai and its surroundings and multi-medium feedback mechanisms
3.2.1. Spatial and temporal distribution characteristics of fluoride in groundwater at Lake Daihai and its surrounding areas
The aforementioned variance analysis revealed a stable spatial pattern of fluoride distribution in Daihai Lake. This section further examines the temporal dynamics of this pattern and its underlying driving mechanisms. Fig. 2 illustrate the spatiotemporal distribution of fluoride in surface water, sediments, and surrounding groundwater in the Daihai Lake region. Temporally (Fig. 2c), fluoride concentrations in surface water exhibited an overall trend of “spring increase-summer decrease-autumn rebound”. Notably, SW1, SW4, and SW6 exhibited distinct short-term peaks in April, whereas sediment concentrations did not significantly increase during the same period. This spatiotemporal decoupling suggests that short-term external inputs (e.g., spring snowmelt runoff from adjacent agricultural areas), rather than sediment release, were the primary sources of fluoride at these sites (Su et al.,2023). The absence of a significant increase in sediment fluoride concentrations suggests a limited short-term adsorption or retention capacity for external fluoride inputs, or that adsorption capacity had already reached saturation (Soliman et al.,2025). Spatially (Fig. 2a), the high-value core zone was located in the northwest–central area, centered around SW6 and extending toward SW3. Contour lines formed concentric patterns, indicating persistent high-concentration hotspots. The transition zone extended northeastward along the central region, with SW9 and SW10 falling within the transitional contour area, lower than the high-value core but higher than the eastern lake. Low-value zones were primarily distributed along the southwestern margin (SW1, SW2) and eastern regions (SW5, SW7), showing a stepwise decline in concentration. The overall pattern exhibited a gradient of “higher in the northwest, lower in the east/southwest,” suggesting stronger external inputs and weaker dilution in the northwest and northern inflow zones (Tian et al.,2017), whereas the eastern and southwestern margins experienced stronger exchange and better dilution(Qin et al.,2007). This finding aligns with the variance analysis: inflow/retention zones represented by SW3 and SW6 formed high-value hotspots, whereas the eastern and southwestern margins were low-value zones dominated by dilution. A total of 22 tributary inflow channels are present around Daihai Lake, among which 8 carry water. According to the 2023 monitoring data, the water quality of these inflow rivers generally met the Class III standard; however, water quality deteriorated during certain periods, even declining to an inferior Class V level, particularly in Daheyan River, Tiancheng River, Gongba River, Wuhao River, Buliang River, Suodai Gully, Wusumu River, and Changchong Gully. The 8 tributaries exhibited fluoride concentrations ranging from 0.21 to 0.82?mg/L, with an annual fluoride load of 1.27 t/a. The average fluoride concentration ranged from 0.41 to 0.67?mg/L, all below the national Class III surface water limit (1.0?mg/L). Among them, Muhua River and Tiancheng River showed stable concentrations with limited variability, whereas the Gongba River, although having a relatively low mean value (0.47?mg/L), exhibited a maximum concentration of 0.82?mg/L, indicating a short-term high-fluoride input risk.To examine the relationship between tributary input and surface water fluoride levels, ANOVA (analysis of variance) was conducted on fluoride ion concentrations. The results showed a statistically highly significant difference between the two (p?<?0.001), indicating that fluoride ion concentrations in surface water were significantly elevated relative to tributaries (6.20?±?0.60?mg/L vs. 0.49?±?0.12?mg/L). This suggests that tributary input exhibits a statistically significant contribution to the increase in surface water fluoride levels. Further analysis of the spatial distribution patterns revealed that this significant correlation was predominantly concentrated in the western inflow zone (e.g., Gongba River and Wusumu River), where the high-fluoride zone of surface water (SW3) exhibited a clear spatial overlap with the high-fluoride zone of sediments (SW1-SW3).This overlap indicates that tributary inflows enhanced local fluoride accumulation by simultaneously increasing surface water concentrations and strengthening sediment storage(Amri et al.,2023). Conversely, at sites distant from inflow points (SW4, SW6), elevated fluoride concentrations could not be entirely attributed to tributary inputs but were more likely associated with intra-lake processes such as evaporative concentration and groundwater discharge (Qu et al.,2025).
As shown in Fig. 2d, sediment fluoride concentrations exhibited oscillatory patterns of increase and decrease with low-amplitude fluctuations. This indicates dynamic bidirectional exchange processes at the sediment–water interface, where fluoride release and adsorption were modulated by environmental factors rather than occurring unidirectionally. Concurrently, seasonal fluctuations in water-soluble and exchangeable fluoride components revealed periodic release-fixation cycles (Liu et al., 2023, Zhu et al., 2025). Spatially, total sediment fluoride exhibited a stable “high in west-moderate in center-low in southeast” pattern (Fig. 2b). SW3, located at the confluence of west-southwest inflows with deceleration zones and semi-enclosed bays, exhibited fine-grained enrichment, strong adsorption/co-precipitation with Fe/Al oxides and organic matter, and low water exchange rates, thereby forming a long-term reservoir (Ding et al., 2018). SW4 occupied the transitional zone between the lake center and the north-central region, characterized by intermediate hydrodynamics and grain size conditions (Wu et al., 2018). It was influenced by both upstream input and dilution/resuspension, resulting in moderate levels. SW1 and SW2, although influenced by western inflows, exhibited lower accumulation than SW4 due to nearshore erosion and unstable external pulses. SW8, located in a relatively open and hydrodynamically active area with coarse sediments, experienced strong dilution and resuspension, which prevented long-term enrichment, resulting in low values (Bu et al., 2020). As shown in Fig. 2a, groundwater fluoride concentrations along the northwest shore peaked in November. This was attributed to the cessation of irrigation following late October harvests, during which residual water containing fertilizers flowed into the northwest depression. Subsequently, the formation of permanent frost in early winter restricted shallow groundwater flow within this area, concentrating residual fluoride (Nogara et al., 2025, Xie et al., 2021). In contrast, due to prolonged drought in late summer, which caused shallow-phase evaporation to reach its annual maximum, concentrations along the southeast shore peaked in September (Du et al., 2022). Spatial analysis further indicated that fluoride concentrations in groundwater were higher along the eastern and western shores of Lake Daihai, whereas the northern shore exhibited lower concentrations. The eastern shore’s groundwater flow retention zone was susceptible to evaporation and concentration, leading to fluoride enrichment. On the western shore, deep groundwater upwelling transported fluoride-rich fluids from deeper geological structures. By contrast, the northern shore, adjacent to mountainous areas, received direct precipitation recharge, resulting in rapid water renewal, short residence times, and minimal fluoride accumulation (Cao et al., 2024).
In summary, the spatiotemporal mechanisms of fluoride in Lake Daihai can be described as follows: The northwest–central region (SW3, SW6) lies within a deceleration and low-exchange inflow zone, forming a stable high-value core over the long term. The eastern and southwestern nearshore areas constitute active dilution and exchange zones, maintaining low concentrations. Sediments act as long-term reservoirs, periodically releasing fluoride in water-soluble and exchangeable forms during high-temperature periods and reabsorbing it during low-temperature periods, leading to seasonal decoupling from surface water. Groundwater along the shoreline provides lateral recharge to nearshore surface water during the dry season (April-June), whereas reverse seepage may occur during the wet season (July-September). These processes are further modulated by evaporation-concentration and freeze-thaw dynamics, respectively, shaping the peaks observed on the northwest shore in November and the southeast shore in September. This corroborates the stable spatial pattern revealed by variance analysis, characterized spatially by “higher levels in the northwest and central regions, lower in the east,” and temporally by alternating external pulses and sediment-water interface processes for groundwater fluoride. Collectively, it integrates groundwater recharge through infiltration, surface water dilution from external sources, and sediment release into a spatially and temporally closed system.
3.2.2. Ecological risk assessment of fluoride in groundwater at Lake Daihai and surrounding areas
To elucidate the migration characteristics and driving mechanisms of fluoride within the multi-medium system of the Daihai Basin, this study first applied PCA to key physicochemical parameters. The resulting PCA ordination plot (Fig. 3a) indicates that surface water (SW), groundwater (GW), tributaries (TRIB), and sediments (Sediment) exhibit markedly differentiated distribution patterns across major physicochemical parameters. PC1 and PC2 collectively accounted for 91.6?% of the total variance. PC1 (76.7?%), primarily influenced by TDS (mg/L) and SAL (ppt), represented salinity buildup and solute accumulation tendency resulting from evaporation-driven concentration and reduced lake replenishment capacity. This reflects the combined response to intensified evaporation and progressive lake shrinkage in recent years (Ren et al.,2022). Daihai Lake is located in a typical arid-semi-arid climate zone, with an annual average precipitation of approximately 395?mm. It experiences prolonged sunshine, with an annual mean of 3000?h, and strong solar radiation. Evaporation reaches 1820?mm, approximately five times the precipitation, resulting in intense evaporative concentration processes(Chen et al., 2020; Schulz et al., 2020)(Fig. 1S). Remote sensing data reveal that from 1980 to 2020, the water level of Daihai Lake dropped by 9?m, and its surface area decreased from 148.34?km2 to 48.52?km2-only 36.15?% of its original extent (Wang et al., 2022)(Fig. 2S). In 2023, the lake surface area was approximately 45?km2, with a water storage volume of 519 million m3 and an average depth of about 6?m (Wu et al., 2025). This long-term hydrological evolution directly amplified the solute enrichment trend revealed by PC1 in PCA, indicating that fluoride migration processes are strongly constrained by climate-driven forcing. The persistent reduction in lake surface area has diminished the dilution capacity of inflowing rivers while intensifying evaporative concentration effects (Ren et al., 2022). The combined impact of these factors has led to fluoride accumulation in the lake water. PC2 (14.9?%) exhibits an approximately opposite distribution between F– and pH, reflecting that under similar salinity conditions, F– shows a strong sediment-directed transport tendency, indicating a pronounced fluoride accumulation trend. Surface water, however, does not directly align along the F– axis, suggesting that its fluoride content is not solely derived from sediment release but is influenced by evaporation and TDS background levels. Against this backdrop, groundwater and tributary samples predominantly cluster on the negative side of PC1, exhibiting lower salinity and TDS levels, indicating that they maintain low-salinity recharge characteristics with relatively stable chemical compositions; surface water samples cluster on the positive side of PC1, indicating that they operate under evaporation-dominated conditions. However, their distribution along the negative direction of PC2, rather than directly along the F– axis, suggests that although fluoride in surface water exists under high-salinity conditions, it remains constrained by interfacial processes such as pH, exhibiting a transitional and phased response state. Sediments also occupy the high-TDS region but exhibit a more pronounced shift toward the F– vector direction, showing a clear accumulation tendency. This indicates that under evaporative concentration conditions, sediments are progressively assuming the role of long-term fluoride storage.
Based on the overall trends revealed by PCA, redundancy analysis (RDA) was further applied to elucidate the interaction mechanisms between surface water and sediments and to investigate the occurrence characteristics of fluorine in sediments. This study further examined the relationships between different fluorine forms and environmental factors. The RDA biplot (Fig. 3b) shows that the vectors of F-res (residual fluoride), F-ws (water-soluble fluoride), and F-total (total fluoride) are nearly collinear and extend along the same direction. Although F-ws is typically considered a highly mobile and reactive fluoride component, it appears nearly collinear with F-res and F-total in the biplot. This suggests that exogenous water-soluble fluoride continuously enters the interface system, gradually transforms into sediment-bound phases, and ultimately accumulates as F-res (Moirana et al., 2021). In contrast, the vectors for F-extract (extractable fluoride) and F-org (organically bound fluoride) exhibit slightly divergent orientations, indicating higher sensitivity to interface conditions such as pH (Fowler et al., 2024). These fluoride fractions are prone to release fluctuations under environmental disturbances, suggesting that they possess higher potential for migration. The F-SW vector is positioned near the origin, showing alignment with TDS and SAL while remaining nearly orthogonal to sediment vectors. This indicates that short-term fluctuations in surface water fluoride content are more readily driven by external inputs and evaporation-induced concentration rather than being solely controlled by background sediment release (Jia et al., 2019). Furthermore, iron/manganese-bound fluoride (F-Fe/Mn) occupies a distinct position in the RDA space, suggesting that it may intermittently influence fluoride migration under specific reduction-reoxidation conditions, acting as an intermittent buffer phase in the overall migration process (Miao et al., 2006). These patterns are corroborated by the correlation heatmap (Fig. 3c), F-res and F-total exhibit a high correlation (r?=?0.99), jointly defining long-term sediment fluoride accumulation, while F-ws and F-extract (r?=?0.86) demonstrate responsiveness to interface conditions, constituting a potential releasable fluoride sub-pool. In contrast, F-SW shows generally low correlations with sediment forms (r?<?0.1) but maintains moderate correlations with TDS and SAL, indicating that it is more strongly regulated by evaporation-induced enrichment processes.
In summary, fluoride in the Daihai Basin does not migrate in a single direction among tributaries, surface water, sediments, and groundwater. Instead, under the combined influence of intensified climatic evaporation and interfacial chemical processes, it forms an interactive feedback cycle (Fig.7S). Tributaries and groundwater primarily serve as dilution and replenishment sources, supplying relatively low-salinity background inputs to the lake body. Surface water exhibits a tendency for fluoride enrichment under evaporative concentration conditions but, constrained by interface conditions such as pH, shows a tendency toward directed migration into certain sediments. Sediments gradually evolve into long-term regional fluoride reservoirs under high-salinity conditions. Residual forms control the baseline fluoride reserves, whereas active components-including water-soluble, extractable, organically bound, and iron-manganese-bound fractions-can be re-released into the water body under specific conditions through adsorption-desorption and redox interface regulation. This structure indicates that fluorine migration in Daihai Lake follows a cyclical pathway that progresses from external input to water body enrichment, then to sediment fixation, and finally to interfacial release. Against the backdrop of intensifying aridification and increasing lake closure, this cycle has gradually shifted from a supply-dominated to an accumulation-feedback-dominated evolutionary pattern, revealing the endogenous maintenance mechanism of fluorine pollution in closed lakes in arid and semi-arid regions.
3.3. Fluoride risk assessment of groundwater in and around Daihai Lake
3.3.1. Ecological risk assessment of fluoride in surface water of Daihai Lake and surrounding groundwater
Excessive fluoride concentrations in aquatic systems directly affect lake organisms, thereby threatening overall ecosystem health. Fluoride levels in Daihai Lake exceed the Class V standard (1.5?mg/L) specified in GB3838–2002 by more than twofold, raising serious concerns regarding their ecological impact. Using the entropy method and China’s freshwater chronic biological reference value (2?mg/L) as the water quality benchmark, the potential ecological risk of fluoride in Daihai Lake surface water was assessed (Fig. 4a). The surface water exhibited potential ecological risk, with more than 99?% of samples showing a hazard quotient greater than 1 and an average value of 3.12, indicating a moderate risk level. Risk levels showed no significant temporal variation across the sampling period. Except for 2?% of samples in July classified as low risk, all other monthly samples fell into the medium-risk category. The integrated pollution index method, combined with the single-factor pollution index method, indicated that groundwater fluoride levels in Daihai Lake were predominantly classified as moderate to light pollution during September, October, and November. In September, the proportion of heavily polluted samples was the highest, reaching 45?%. Based on fluoride concentration monitoring results, the maximum composite pollution index for groundwater reached 3.4, with an average of 1.9, indicating moderate pollution (Fig. 4b).
3.3.2. Ecological risk assessment of sediments in Daihai Lake
Using the modified Nemero pollution index and sediment pollution assessment standards, the comprehensive pollution index was calculated to be 2.73. Among the fluoride species in sediments, water-soluble and exchangeable forms exhibited weak binding forces and relatively high mobility, resulting in rapid release and high biological activity. A higher proportion of extractable fluoride was correlated with increased fluoride bioavailability in sediments and elevated potential risk levels. The stability risk assessment criteria were applied to evaluate the potential ecological risks of fluoride in Daihai Lake sediments. As shown in Fig.8S, water-soluble fluoride in surface sediments of Daihai Lake ranged from 29.51 to 42.87?mg/kg, with an average of 32.36?mg/kg. Consequently, the stability range of fluoride in surface sediments of Daihai Lake was 71.69?%-79.71?%, with an average of 73.89?%, indicating an extremely unstable state. This suggests that fluoride in Daihai Lake sediments exhibits overall poor stability, with potential mobility and bioavailability, as well as a certain risk of desorption and release. Therefore, fluoride poses an ecological threat to the Daihai Lake ecosystem. A more detailed analysis evaluated the potential ecological risks of fluoride pollution in different species within Daihai Lake sediments, with results shown in Fig.9S. Based on the speciation-based fluoride pollution assessment, the ecological risk index for fluoride in Daihai Lake sediments ranged from 0.83 to 1.77. Over time, the ecological risk index showed a gradual trend, with the majority of sampling points indicating moderate risk and a few indicating mild contamination. The potential risk assessment results for all sampling points remained above the safety threshold.
3.4. Health risk assessment of groundwater around Daihai Lake
3.4.1. Non-carcinogenic risk assessment
The results of the health risk assessment (Table S4) indicate that non-carcinogenic effects varied across the four age groups. Specifically, non-carcinogenic risks through dietary exposure yielded HQing values ranging from 0.7135 to 2.5312 for infants, 0.4281–1.5187 for children, 0.4116–1.4603 for adolescents, and 0.4893–1.7357 for adults. The HQder values ranged from 0.0000 to 0.0001 for infants, 0.0003–0.0010 for children, 0.0004–0.0015 for adolescents, and 0.0005–0.0019 for adults. The non-carcinogenic risk values for fluoride in groundwater around Daihai Lake ranged from 0.8155 to 2.5314 for infants, 0.4896–1.5197 for children, 0.4709–1.4619 for adolescents, and 0.5598–2.5314 for adults. The corresponding mean values were 1.5735, 0.9447, 0.9087, and 1.0801, respectively. Health risks from fluoride decreased with increasing age. Furthermore, additional analysis revealed that THQ values in groundwater samples exceeded the threshold (THQ > 1.0) in 85?% of infants, 60?% of adults, and 30?% of both children and adolescents. These findings indicate that fluoride contamination in the study area poses significant health risks to humans, with infants being particularly vulnerable to adverse health effects from high-fluoride groundwater consumption.
The THQ values for non-carcinogenic health risk assessments across different age groups and months are presented in Fig. 5. The figure indicates that infants in the study area may be exposed to non-carcinogenic health risks throughout the year. For children, adolescents, and adults, periods of elevated health risks primarily occurred in April and June. However, compared with other months, the non-carcinogenic risk in June was significantly higher, despite fluoride concentrations in the water not being particularly elevated. This phenomenon is closely related to elevated summer temperatures, which increase water intake and concurrently enhance opportunities for skin contact and respiratory exposure, thereby elevating total exposure doses. In addition, profuse sweating during summer alters electrolyte balance, heightening the body’s sensitivity to fluoride. Under thermal stress, reduced antioxidant capacity permits even low concentrations of fluoride to induce fluorosis (Basha and Sujitha, 2012). Synergistic interactions with other seasonal pollutants further reduce the safety threshold for fluoride.
3.4.2. Uncertainty analysis of risk assessment
In 1983, the U.S. National Research Council standardized the procedures for risk assessment, outlining four phases: hazard identification, dose-response evaluation, exposure assessment, and risk characterization. Exposure assessment entails substantial computational uncertainty, which makes it difficult to obtain precise and comprehensive data during the evaluation process. Consequently, the concept of “uncertainty” was introduced into risk assessment. Based on global scholarly consensus, the unknown factors in health risk assessment can be classified into three categories: (i) scenario uncertainty, which arises from vague or inaccurate descriptions of the surrounding environment or population conditions; (ii) model uncertainty, resulting from errors or biases inherent in the model itself; and (iii) parameter uncertainty, arising from errors in the input data of various models (USEPA,2002). To address parameter uncertainty, sensitivity analysis is employed to quantify the influence of each input variable on the outcome. Badeenezhad et al. (2023) demonstrated that Monte Carlo-based probabilistic risk assessment improves the accuracy and reliability of health risk estimates by accounting for variability and uncertainty in input parameters.
A review of relevant literature (Ganyaglo et al., 2019) indicates that the primary indicators considered for model parameter uncertainty are intake rate (IR), body weight (BW), and exposure duration (EF). Among these, IR is assumed to follow a normal distribution, BW a log-normal distribution, and EF a trigonometric distribution. The distributions of pollutant concentrations and exposure parameter values are presented in Table S5.
Formulas (8)-(12) were used to calculate the non-carcinogenic risks for infants, children, adolescents, and adults in the Daihai Lake region, with the distribution results illustrated in Fig. 6. In the figure, red bars denote risk samples below the assessment threshold of 1.0, whereas blue bars indicate samples exceeding this threshold, signifying potential hazards to human health. Based on the uncertainty analysis of the simulation, the overall risk of exposure across all groups in the study area was found to follow a log-normal distribution. Using the average fluoride concentration in groundwater within the study area, the mean non-carcinogenic health risk hazard quotients for infants, children, adolescents, and adults were 1.39, 0.83, 0.79, and 0.95, respectively. From the perspective of non-carcinogenic risk probability, the uncertainty model indicated that the probabilities of exceeding health risk thresholds for infants, children, adolescents, and adults were 74.25?%, 28.18?%, 24.82?%, and 42.47?%, respectively. These findings are largely consistent with those of deterministic studies, confirming that infants face the highest non-carcinogenic fluoride risk.
3.4.3. Sensitivity analysis of risk assessment
The primary objective of sensitivity analysis is to assess the responsiveness and sensitivity of various indicators to anticipated outcomes (Saltelli et al., 2019). This process helps identify, among numerous unknown factors, the elements that significantly influence assessment results, thereby providing valuable guidance for risk management (Kaur et al., 2020). In sensitivity analysis, indicators with high sensitivity exert a proportionally greater influence on assessment outcomes. Improving the accuracy of data for highly sensitive indicators enhances the overall precision of assessment results. By contrast, less sensitive parameters contribute relatively little to the overall assessment outcomes. When environmental conditions do not permit high-precision data collection, the requirements for such data may be appropriately relaxed. In health hazard assessments, it is often not possible to precisely determine the extent to which parameter settings influence outcomes. Therefore, sensitivity studies serve to identify key factors, thereby providing a stronger basis for guiding and optimizing groundwater quality management measures around Lake Daihai.
Using Crystal Ball to construct a Monte Carlo uncertainty model, a sensitivity analysis was performed to investigate the influencing factors, with the corresponding results presented in Fig.10S. The absolute sensitivity value derived from the analysis indicates the relative magnitude of each parameter’s impact on risk assessment, while positive and negative values denote positive and negative correlations with the results, respectively. As shown in Fig.10S, F? contributes most significantly to non-carcinogenic risk in infants and children, with sensitivities of 73.9?% and 30.6?%, respectively. Exposure duration and body weight are the primary contributors to non-carcinogenic risk in adolescents and adults. Among exposure parameters, body weight shows relatively low sensitivity compared to other variables, whereas exposure duration (EF) contributes more substantially. For infants, exposure duration shows a positive correlation, as their incomplete blood-brain barrier and underdeveloped renal function reduce fluoride clearance capacity to only 30–50?% of that in adults. Breast milk or formula constitutes the primary intake route; if water fluoride exceeds standards, long-term exposure through biomagnification leads to sustained effects(Till et al., 2020). For children, exposure duration shows a significant negative correlation. Between the ages of 6 and 12, rapid increases in glomerular filtration mitigate short-term exposure risks through efficient excretion (Wang et al., 2021). However, long-term exposure upregulates fluoride tolerance genes, paradoxically reducing apparent toxicity. As children age, expanded outdoor activity and reduced reliance on single water sources diminish the cumulative effects of prolonged exposure. For adolescents, exposure duration shows a significant positive correlation. Hormonal changes during puberty enhance fluoride release from bones, while long-term exposure increases thyroid and nervous system doses (Ferreira et al., 2024). Adolescents are additionally exposed through pathways such as sports drinks, tea, and cosmetics, amplifying cumulative effects over time. For adults, exposure duration shows a negative correlation. Adults possess well-developed hepatic enzyme systems and renal excretion functions; long-term exposure upregulates detoxification enzymes, thereby enhancing fluoride metabolism. In addition, adults exhibit reduced skin permeability and stable drinking habits, lowering actual intake from long-term exposure (Nizam et al., 2022). Therefore, safeguarding the health of residents in the Daihai Lake region requires effective strategies to reduce groundwater pollution and prevent excessive exposure to harmful substances. Particular emphasis should be placed on the treatment of fluoride-containing wastewater, which is critical for maintaining groundwater quality in arid and semi-arid regions.
4. Conclusion
(1) The fluoride concentration in Lake Daihai’s water body substantially exceeds the Class V threshold (1.5?mg/L) specified in the Surface Water Environmental Quality Standard (GB 3838-2002). Meanwhile, fluoride levels in groundwater from surrounding rural drinking wells are significantly higher than the national average of 0.66?mg/L and also exceed the 1.0?mg/L guideline value specified in China’s Sanitary Standards for Drinking Water (GB 5749-2022). Accordingly, the region can be classified as a high-fluoride pollution zone in China.
(2) The distribution and temporal dynamics of fluoride concentrations in surface water and sediments are shaped by the combined effects of external inputs, groundwater migration, sediment regulation, and environmental conditions. Spatially, fluoride concentrations display distinct regional heterogeneity, with higher levels in western estuaries and northern surface waters, and pronounced accumulation in sediments of the northwest. Temporally, surface water fluoride concentrations demonstrate an initial rise followed by a decline, whereas fluoride in sediments exhibits relatively minor fluctuations, reflecting dynamic exchange processes at the sediment–water interface. Short-term external inputs (e.g., spring snowmelt runoff) can directly elevate fluoride concentrations in surface water. Sediments, functioning as long-term internal sources, release or adsorb fluoride under the regulation of hydrological factors (e.g., flow velocity, disturbance) and chemical properties (e.g., pH, salinity).
(3) Fluoride levels in the surface water of Lake Daihai have reached a moderate ecological risk level, posing potential threats to the lake’s ecological integrity and endangering the overall health of its ecosystem. Fluoride concentrations in local drinking water wells pose non-carcinogenic health risks to residents, with infants facing the highest risks and being particularly vulnerable to health complications from consuming high-fluoride groundwater. Sediments in Lake Daihai exhibit mild to moderate levels of contamination risk. Fluoride stability is generally poor, with sediment-bound species existing in a highly unstable state. These species possess potential mobility and bioavailability, presenting a notable risk of desorption and release.
(4) By integrating the EPA health risk model with Monte Carlo–Crystal Ball uncertainty analysis, risk values are expressed in probabilistic terms, thereby significantly enhancing the scientific rigor and practical applicability of the assessment outcomes for decision-making. The composite risk assessment model, based on morphological bioavailability, enables the dynamic quantification of coupled ecological and health risks through entropy-weighted methods and Crystal Ball probabilistic simulations. In terms of health risk severity, infants are at the greatest risk, followed by adults, children, and, lastly, adolescents.
CRediT authorship contribution statement
Xizheng Wang: Validation, Methodology, Investigation. Zhuo Li: Validation, Methodology, Conceptualization. Jiahui Mi: Visualization, Methodology, Investigation. Junping Lu: Supervision, Project administration, Funding acquisition, Data curation, Conceptualization. Jiale Sun: Writing – review & editing, Writing – original draft, Validation, Software, Methodology, Formal analysis, Conceptualization. Tingxi Liu: Visualization, Supervision, Resources, Funding acquisition, Conceptualization. Yinghui Liu: Supervision, Methodology, Formal analysis, Conceptualization.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This research was funded by the National Natural Science Foundation of China (U24A20633;52260029); Inner Mongolia Autonomous Region science and technology plan project (2023YFHH0060;2025YFHH0188); Inner Mongolia Autonomous Region Department of Education Science and Technology Talent Project (NJYT22040); Inner Mongolia Agricultural University Young Teacher Research Ability Enhancement Project (BR220102); National Key Research and Development Program project (2019YFC0409204); Inner Mongolia Natural Science Foundation project (2022MS05053); First-class Academic Subjects Special Research Project of the Education Department of Inner Mongolia Autonomous Region (No.YLXKZX-NND-010); Inner Mongolia Autonomous Region Science and Technology Leading Talent Team (2022LJRC0007).
Appendix A. Supplementary material
Supplementary material
Data availability
Data will be made available on request.
References
Amri et al., 2023. S. Amri, L. Trabelsi, M.A. Dahim, B. Brik, W. Oueslati, M.V.A. Martins, L. Aleya, N. Zaaboub. Assessment of trace elements and fluoride originating from phosphogypsum in the sediment of Gulf of Gabes (southeastern Tunisia): what are the potential sources of accumulation and bioavailability?. Environ. Sci. Pollut. Res., 30 (29) (2023), pp. 73405-73424, 10.1007/s11356-023-27518-9
Badeenezhad et al., 2023. A. Badeenezhad, H. Soleimani, S. Shahsavani, I. Parseh, A. Mohammadpour, O. Azadbakht, P. Javanmardi, H. Faraji, K.B. Nalosi. Comprehensive health risk analysis of heavy metal pollution using water quality indices and Monte Carlo simulation in R software. Sci. Rep., 13 (1) (2023), p. 15817, 10.1038/s41598-023-43161-3
Basha and Sujitha, 2012. P.M. Basha, N.S. Sujitha. Combined influence of intermittent exercise and temperature stress on the modulation of fluoride toxicity. Biol. Trace Elem. Res., 148 (1) (2012), pp. 69-75, 10.1007/s12011-012-9338-4
Botle et al., 2020. A. Botle, R.K. Singhal, H. Basu, M. V, J. Masih. Health risk assessment of heavy metals associated with Coarse and Quasi-accumulative airborne particulate matter in Mumbai City situated on the Western Coast of India. Environ. Technol. Innov., 19 (2020), Article 100857, 10.1016/j.eti.2020.100857
Bu et al., 2020. M. Bu, Y. Li, J. Wei, C. Tang. The influence of ship waves on sediment resuspension in the large shallow Lake Taihu, China. Int. J. Environ. Res. Public Health, 17 (19) (2020), p. 7055, 10.3390/ijerph17197055
Cao et al., 2024. L. Cao, Z. Nie, J. Shen, Z. Wang, Z. Cheng, W. Liu. Enrichment mechanism of groundwater fluoride and its hydrogeological indications in the Badain Jaran Desert, Northwest China. Appl. Geochem., 175 (2024), Article 106176, 10.1016/j.apgeochem.2024.106176
Chen et al., 2020. J. Chen, J. Lv, N. Li, Q. Wang, J. Wang. External groundwater alleviates the degradation of closed lakes in semi-arid regions of China. Remote Sens., 12 (2020), p. 45, 10.3390/RS12010045
Chen et al., 1991. J. Chen, F. Wei, C. Zheng, Y. Wu, D.C. Adriano. Background concentrations of elements in soils of China. Water Air Soil Pollut., 57 (58) (1991), pp. 699-712, 10.1007/BF00282934
Del Signore et al., 2016. A. Del Signore, A.J. Hendriks, H.R. Lenders, R.S. Leuven, A. Breure. Development and application of the SSD approach in scientific case studies for ecological risk assessment. Environ. Toxicol. Chem., 35 (9) (2016), pp. 2149-2161, 10.1002/etc.3474
Ding et al., 2018. S. Ding, M. Chen, M. Gong, X. Fan, B. Qin, H. Xu, S.S. Gao, Z. Jin, D.C.W. Tsang, C. Zhang. Internal phosphorus loading from sediments causes seasonal nitrogen limitation for harmful algal blooms. Sci. Total Environ., 625 (2018), pp. 872-884, 10.1016/j.scitotenv.2017.12.348
Du et al., 2022. Y. Du, W. Wan, Q. Li, H. Zhang, H. Qian, J. Cai, J. Wang, X. Zheng. Internal phosphorus loading from sediments causes seasonal nitrogen limitation for harmful algal blooms. Sci. Total Environ., 625 (2022), pp. 872-884, 10.1371/journal.pone.0266049
Edmunds and Smedley, 2012. W.M. Edmunds, P.L. Smedley. Fluoride in natural waters[M]//Essentials of medical geology: Revised Edition. Dordrecht, Springer Netherlands (2012), pp. 311-336, 10.1007/978-94-007-4375-5_13
Ferreira et al., 2024. M.K.M. Ferreira, P.C. Nascimento, L.O. Bittencourt, G.H.N. Miranda, N.C.F. Fagundes, F.V. Zahoori, E.A. Martinez-Mier, M.A.R. Buzalaf, R.R. Lima. Is there any association between fluoride exposure and thyroid function modulation? A systematic review. Plos One, 19 (4) (2024), Article e0301911, 10.1371/journal.pone.0301911
Fowler et al., 2024. A. Fowler, N. Goulding, S. Law. Fluoride geochemistry in groundwater at regulated industrial sites. Water Environ. Res., 96 (8) (2024), Article e11105, 10.1002/wer.11105
Ganyaglo et al., 2019. S.Y. Ganyaglo, A. Gibrilla, E.M. Teye, E.D.J. Owusu-Ansah, S. Tettey, P.Y. Diabene, S. Asimah. Groundwater fluoride contamination and probabilistic health risk assessment in fluoride endemic areas of the Upper East Region, Ghana. Chemosphere, 233 (2019), pp. 862-872, 10.1016/j.chemosphere.2019.05.276
Horowitz, 1991. A.J. Horowitz. A primer on sediment-trace element chemistry (No. 91-76). US Geol. Surv. Books. OpenFile Rep. Sect. [Distrib.] (1991)
Hu et al., 2023. W. Hu, W. Hao, H. Liu, W. Wang, R. Jiao. Water quality evaluation of Dahai Lake in Inner Mongolia based on improved Nemero pollution index method and principal component analysis (EDP Sciences). E3S Web Conf., 393 (2023), p. 03040, 10.1051/e3sconf/202339303040
Jia et al., 2019. H. Jia, H. Qian, W. Qu, L. Zheng, W. Feng, W. Ren. Fluoride occurrence and human health risk in drinking water wells from southern edge of Chinese Loess Plateau. Int. J. Environ. Res. Public Health, 16 (10) (2019), p. 1683, 10.3390/ijerph16101683
Kaur et al., 2020. L. Kaur, M.S. Rishi, A.U. Siddiqui. Deterministic and probabilistic health risk assessment techniques to evaluate non-carcinogenic human health risk (NHHR) due to fluoride and nitrate in groundwater of Panipat, Haryana, India. Environ. Pollut., 259 (2020), Article 113711, 10.1016/j.envpol.2019.113711
Kitalika et al., 2018. A.J. Kitalika, R.L. Machunda, H.C. Komakech, K.N. Njau. Fluoride variations in rivers on the slopes of mount meru in Tanzania. J. Chem., 2018 (1) (2018), p. 7140902, 10.1155/2018/7140902
Kudo et al., 1987. A. Kudo, J. Garrec, R. Plebin. Fluoride distribution and transport along rivers in the French Alps. Ecotoxicol. Environ. Saf., 13 (3) (1987), pp. 263-273, 10.1016/0147-6513(87)90025-x
Lafayette et al., 2020. G.N. Lafayette, P.S.K. Knappett, Y. Li, I. Loza-Aguirre, M.L. Polizzotto. Geogenic sources and chemical controls on fluoride release to groundwater in the Independence Basin, Mexico. Appl. Geochem., 123 (2020), Article 104787, 10.1016/j.apgeochem.2020.104787
Liu et al., 2023. Y. Liu, J. Lu, T. Liu, Z. Shi, H. Ren, J. Mi. Analysis of the distribution across media, migration, and related driving factors of fluoride in cold and arid lakes during the freezing period. Environ. Res., 244 (2023), Article 117899, 10.1016/j.envres.2023.117899
Liu et al., 2022. X. Liu, M. Tu, S. Wang, Y. Wang, J. Wang, Y. Hou, X. Zheng, Z. Yan. Research on freshwater water quality criteria, sediment quality criteria and ecological risk assessment of triclosan in China. Sci. Total Environ., 816 (2022), Article 151616, 10.1016/j.scitotenv.2021.151616
Marselina et al., 2025. M. Marselina, N.A. Rahmi, S.A. Nurhayati. The water quality of the Upper Citarum: applying the overall index of pollution, Said-WQI, and pollution index method. Heliyon, 11 (2) (2025), Article e41690, 10.1016/j.heliyon.2025.e41690
Miao et al., 2006. S. Miao, R. delaune, A. Jugsujinda. Influence of sediment redox conditions on release/solubility of metals and nutrients in a Louisiana Mississippi River deltaic plain freshwater lake. 1–3. Sci. Total Environ., 371 (2006), pp. 334-343, 10.1016/j.scitotenv.2006.07.027
Moirana et al., 2021. R.L. Moirana, J. Mkunda, M.P. Perez, R. Machunda, K. Mtei. The influence of fertilizers on the behavior of fluoride fractions in the alkaline soil. J. Fluor. Chem., 250 (2021), Article 109883, 10.1016/j.jfluchem.2021.109883
Moller and Gudjonsson, 1932. P.F. Moller, S.V. Gudjonsson. Massive fluorosis of bones and ligaments. Acta Radiol. (3-4) (1932), pp. 269-294, 10.1177/028418513201300309
Nizam et al., 2022. S. Nizam, H.S. Virk, I.S. Sen. High levels of fluoride in groundwater from Northern parts of Indo-Gangetic plains reveals detrimental fluorosis health risks. Environ. Adv., 8 (2022), Article 100200, 10.1016/j.envadv.2022.100200
Nogara et al., 2025. E.I. Nogara, M. Anim-Gyampo, S.A. Ganiyu. Seasonal fluctuations in groundwater fluoride concentration and health risk assessment: Tolon District, Northern Ghana. J. Water Land Dev. (2025), 10.24425/jwld.2025.154250
Okkerman et al., 1991. P. Okkerman, E. Plassche, W. Slooff, C. Van Leeuwen, J. Canton. Ecotoxicological effects assessment: a comparison of several extrapolation procedures. Ecotoxicol. Environ. Saf., 21 (2) (1991), pp. 182-193, 10.1016/0147-6513(91)90020-p
Pan et al., 2018. Z. Pan, X. Liu, W. Meng, et al. Geochemical characteristics of fluorine in soils and its environmental quality in central district of Guiyang[J] . Res. Environ. Sci., 31 (1) (2018), pp. 87-94, 10.13198/j.issn.1001-6929.2017.03.36
View in ScopusGoogle ScholarQin et al., 2007
Qu et al., 2025. S. Qu, X. Ren, Y. Zhao, H. Mao, S. Dong, R. Yu. Unveiling origin and enrichment of fluoride in the Daihai lake basin, China, using a hybrid hydrochemical and multi-isotopic method. J. Hydrol., 656 (2025), Article 133030, 10.1016/j.jhydrol.2025.133030
View PDFView articleView in ScopusGoogle ScholarRen et al., 2022
Saltelli et al., 2019. A. Saltelli, K. Aleksankina, W. Becker, P. Fennell, F. Ferretti, N. Holst, S. Li, Q. Wu. Why so many published sensitivity analyses are false: a systematic review of sensitivity analysis practices. Environ. Model. Softw., 114 (2019), pp. 29-39, 10.1016/j.envsoft.2019.01.012
Schulz et al., 2020. S. Schulz, S. Darehshouri, E. Hassanzadeh, M. Tajrishy, C. Schüth. Climate change or irrigated agriculture – what drives the water level decline of Lake Urmia. Sci. Rep., 10 (1) (2020), p. 236, 10.1038/s41598-019-57150-y
Sharma et al., 2017. D. Sharma, A. Singh, K. Verma, S. Paliwal, S. Sharma, J. Dwivedi. Fluoride: A review of pre-clinical and clinical studies. Environ. Toxicol. Pharmacol., 56 (2017), pp. 297-313, 10.1016/j.etap.2017.10.008
Singh et al., 2005. K.P. Singh, D. Mohan, V.K. Singh, A. Malik. Studies on distribution and fractionation of heavy metals in Gomti river sediments – a tributary of the Ganges, India. 1–4. J. Hydrol., 312 (2005), pp. 14-27, 10.1016/j.jhydrol.2005.01.021
Soliman et al., 2025. N.F. Soliman, M.A. Fahmy, M.A. Abdelkhaleq, S.M. Nasr. Mapping fluoride footprints: fractionation, bioaccumulation, sources, and sophisticated risk assessment in El-Bardawil Lake, Sinai Peninsula, Egypt. Sci. Total Environ., 1002 (2025), Article 180581, 10.1016/j.scitotenv.2025.180581
2002. Standardization Administration of China GB 3838-2002: Environmental quality standard for surface water, Standards Press of China, Beijing (2002). Google Scholar
2022. Standardization Administration of China. GB 5749-2022: Hygienic standard for drinking water, Standards Press of China, Beijing (2022). Google Scholar
Su et al., 2023. C. Su, M. Wang, X. Xie, Z. Han, J. Jiang, Z. Wang, D. Xiao. Natural and anthropogenic factors regulating fluoride enrichment in groundwater of the Nansi Lake Basin, Northern China. Sci. Total Environ., 904 (2023), Article 166699, 10.1016/j.scitotenv.2023.166699
Tang et al., 2021. J. Tang, Y. Dai, J. Wang, Y. Qu, B. Liu, Y. Duan, Z. Li. Study on environmental factors of fluorine in chagan lake catchment, northeast China. Water (Switz.), 13 (5) (2021), p. 629, 10.3390/w13050629
Tian et al., 2017. Z. Tian, B. Zheng, L. Wang, L. Li, X. Wang, H. Li, S. Norra. Long term (1997-2014) spatial and temporal variations in nitrogen in Dongting Lake, China. Plos ONE, 12 (2) (2017), Article e0170993, 10.1371/journal.pone.0170993
Till et al., 2020. C. Till, R. Green, D. Flora, R. Hornung, E.A. Martinez-Mier, M. Blazer, L. Farmus, P. Ayotte, G. Muckle, B. Lanphear. Fluoride exposure from infant formula and child IQ in a Canadian birth cohort. Environ. Int., 134 (2020), Article 105315, 10.1016/j.envint.2019.105315
USEPA, 1989. U.S. Environmental Protection AgencyUSEPA).,1989. Risk Assessment Guidance for Superfund (RAGS), Volume I: Human Health Evaluation Manual (Part A). Washington, DC: USEPA.
USEPA, 2002. USEPA, 2002. Supplemental guidance for developing soil screening levels for superfund sites. United States Environmental Protection Agency, Washington, DC. Google Scholar
Wang et al., 2021. A. Wang, Q. Ma, B. Gong, L. Sun, F.K. Afrim, R. Sun, T. He, H. Huang, J. Zhu, G. Zhou, Y. Ba. DNA methylation and fluoride exposure in school-age children: epigenome-wide screening and population-based validation. Ecotoxicol. Environ. Saf., 223 (2021), Article 112612, 10.1016/j.ecoenv.2021.112612
Wang et al., 2020. T. Wang, Z. Shao, H. Yu, H. Bah. Distribution of fluoride in surface water and a health risk assessment in the upper reaches of the Yongding River. J. Geogr. Sci., 30 (6) (2020), pp. 908-920, 10.1007/s11442-020-1761-4
Wang et al., 2022. S. Wang, C. Xu, W. Zhang, H. Chen, B. Zhang. Human-induced water loss from closed inland Lakes: hydrological simulations in China’s Daihai lake. J. Hydrol., 607 (2022), Article 127552, 10.1016/j.jhydrol.2022.127552
Wen et al., 2013. D. Wen, F. Zhang, E. Zhang, C. Wang, S. Han, Y. Zheng. Arsenic, fluoride and iodine in groundwater of China. J. Geochem. Explor., 135 (2013), pp. 1-21, 10.1016/j.gexplo.2013.10.012
WHO, 2011. World Health OrganizationWHO), 2011. Guidelines for Drinking-water Quality, fourth ed. World Health Organization, Geneva, Switzerland.
Wu et al., 2018. T. Wu, B. Qin, J.D. Brookes, W. Yan, X. Ji, J. Feng. Spatial distribution of sediment nitrogen and phosphorus in Lake Taihu from a hydrodynamics-induced transport perspective. Sci. Total Environ., 650 (2018), pp. 1554-1565, 10.1016/j.scitotenv.2018.09.145
Wu et al., 2025. R. Wu, X. Shi, X. Kang, S. Zhang, S. Zhao, Y. Liu, B. Sun, J. Lu, H. Yu, S. Wang, X. Pan, K. Shen, L. Arvola, J. Yan, R. Hao, R. Shi. Characteristics of nitrogen and phosphorus migration at sediment-water interface in seasonal frozen lakes and the mechanism of microbial driven cycling: a case study of Lake Daihai. J. Contam. Hydrol. (2025), Article 104598, 10.1016/j.jconhyd.2025.104598
Xie et al., 2021. H. Xie, X. Jiang, S. Tan, L. Wan, X. Wang, S. Liang, Y. Zeng. Interaction of soil water and groundwater during the freezing–thawing cycle: field observations and numerical modeling. Hydrol. Earth Syst. Sci., 25 (8) (2021), pp. 4243-4257, 10.5194/hess-25-4243-2021
Zhu et al., 2025. J. Zhu, R. Liu, H. Guo, J. Chen, D. Ning, X. Zang. Distribution characteristics and enrichment mechanisms of fluoride in alluvial–lacustrine facies clayey sediments in the land subsidence area of Cangzhou Plain, China. Water, 17 (19) (2025), p. 2887, 10.3390/w17192887