Integrated Chemical and Ecotoxicological Assessment of Metal Contamination in the Andong Watershed: Identifying Key Toxicants and Ecological Risks

Abstract

This study employed an integrated field monitoring approach, combining chemical analysis and ecotoxicity testing of multiple environmental matrices—water, sediment, and sediment elutriates—to comprehensively assess the environmental health of the Andong watershed, located near a Zn smelter and mining area. The primary objectives were to evaluate the extent of metal contamination, identify key toxicants contributing to ecological degradation, and trace the sources of these pollutants. Our findings revealed severe metal contamination and significant ecotoxicological effects both in proximity to and downstream from industrial sites. Specifically, Cd, Zn, and Pb were strongly linked to the smelter, while Hg, Ni, Cu, and As were predominantly associated with mining activities in the tributaries. To further assess toxicity of field-collected sediment and their elutriates, a logistic regression analysis was employed to estimate benchmark values for distinguishing between toxic and non-toxic samples, using the sum of toxic units for sediment elutriates and the mean probable effect level (PEL) quotient for sediment toxicity. These models demonstrated greater predictive accuracy than conventional benchmarks for determining toxicity thresholds. Our results highlight that integrating chemical and ecotoxicological monitoring with site-specific concentration–response relationships enhances the precision of ecological risk assessments, facilitating more accurate identification of key toxicants driving mixture toxicity in complex, pollution-impacted aquatic ecosystems.

1. Introduction

The assessment of ecological impacts from metal contamination in aquatic ecosystems has become increasingly complex, necessitating an integrated approach that combines multiple lines of evidence. The Weight-of-Evidence (WoE) approach has emerged as a robust methodology for diagnosing ecological impacts in such environments [1,2,3]. This approach typically incorporates three main lines of evidence: chemical contamination, toxicity, and ecological effects at the community level. By integrating these diverse data types, the WoE method provides a more comprehensive understanding of environmental health and associated risks [4,5].

In this study, we specifically focus on two lines of evidence: chemical contamination and ecotoxicological monitoring [5]. Chemical analysis quantifies the concentrations of pollutants across environmental matrices, while ecotoxicological monitoring assesses the biological impacts of these contaminants through laboratory and field-based toxicity tests. This combined approach enables a more nuanced evaluation of ecosystem health and helps identify the key toxicants driving observed toxicity in environmental samples—a critical component of environmental risk assessment and management [6,7,8]. Central to this analysis is the development of site-specific concentration–response relationships using mixture toxicity models, such as the toxic unit (TU) model [9,10] and the mean probable effect level quotient (PELq) [11,12]. Logistic regression models have proven to be particularly valuable for establishing relationships between contaminant concentrations and toxicity probabilities, offering crucial insights for risk assessment and management decisions [13,14,15].

The Andong watershed, located near a zinc smelter and mining area, presents an ideal case study for applying this integrated approach. The industrial history of this region has raised concerns about metal accumulation in water bodies and sediments, necessitating a comprehensive evaluation of its ecological status. Previous studies have reported significant metal contamination in Andong Lake sediments, primarily from emissions by the smelter and the surrounding mining activities [16,17,18]. Metal sources were confirmed through historical sediment core analyses with age dating [19], Pb and Zn isotope analyses [20], and receptor modeling [21]. While these studies have focused on assessing contamination status, potential risks, and source identification [22,23,24,25], few have explored the ecotoxicological effects or identified the key toxicants in the Andong watershed’s contaminated sediments. This study addresses these gaps by conducting ecotoxicity tests on water, sediment, and sediment elutriates to assess site-specific ecological risks and identify the primary toxicants.

Based on the dynamics of metal contamination in aquatic ecosystems, we propose several hypotheses for this study. First, we hypothesize that metal contamination and ecotoxicity will display distinct spatial patterns related to pollution sources, with varying metals contributing to toxicity in different environmental compartments due to their physicochemical properties and bioavailability. Second, we anticipate that site-specific concentration–response relationships derived from integrated chemical and ecotoxicological data will yield more accurate toxicity predictions than generic environmental quality guidelines, facilitating the identification of key toxicants. Finally, we hypothesize that contaminated sediments may act as secondary pollution sources, potentially releasing metals back into the water column under certain environmental conditions, thereby posing long-term ecological risks.

The primary objectives of this study are to assess the extent of metal contamination in water, sediments, and sediment elutriates; evaluate associated ecotoxicological risks; identify key toxicants contributing to observed toxicity; and establish site-specific concentration–response relationships. By integrating chemical and ecotoxicological monitoring, we aim to provide a robust framework for understanding the direct impacts of metal contamination in the Andong watershed, including the potential role of sediments as secondary pollution sources.

2. Materials and Methods

2.1. Study Area

The Andong watershed, which includes Andong Lake, its upstream rivers, and tributaries, is located in the northern part of the Nakdong River in Korea (Figure 1). Andong Lake, an artificial reservoir constructed in 1977, serves both agricultural and hydroelectric purposes, with a surface area of 52 km² and a basin area of 1584 km². Approximately 90 km upstream from the lake, the Sukpo Zn smelter—the largest in Korea—has been in operation since 1973, producing zinc products, copper sulfate, and sulfuric acid. Furthermore, approximately 100 abandoned mines, mostly active since the 1960s, are located within the upstream tributaries.

2.2. Chemical Monitoring

Chemical monitoring of water and sediment samples from the Upper Nakdong River, its tributaries, and Andong Lake was performed to assess metal contamination patterns and identify potential sources. Samples were collected from Andong Lake (located 90~138 km downstream from the smelter) and the Upper Nakdong River (spanning the 90 km distance between the smelter and Andong Lake) (Figure 1).

2.2.1. Surface Water and Sediment Sampling

Surface water samples were collected seasonally, four times in 2020 (February, May, August, and October), while sediment samples were collected biannually (April and September). In the upstream region between the smelter and Andong Lake, surface water was collected from 29 sites (main stream: NMw01–NMw17; tributaries: NTw01–NTw12), and sediment samples were taken from 58 sites (N01–N58). In Andong Lake, surface water was collected from 21 sites (Aw01–Aw21), and sediment samples were collected from 109 sites (A001–A109). Therefore, a total of 116 water samples (4 times from 29 sites) and a total of 218 sediment samples (2 times from 109 sites) were obtained.

Water samples were filtered on-site using 0.45 μm syringe filters (ADVANTEC^® Syringe filter, PTFE, Japan). The filtered water was then stored in acid-washed 10 mL plastic bottles for transport. Upon arrival at the laboratory, the water samples were immediately acidified with purified concentrated nitric acid and stored at 4 °C in the dark until further analysis. Sediment samples were collected from submerged areas (2–3 cm depth) using a plastic scoop, sieved through a 2 mm mesh sieve in the field, and subsequently freeze-dried in the laboratory. The dried sediment samples were ground and homogenized using an agate mortar prior to chemical analysis.

The sediment samples were collected and processed using two distinct sieving methods to suit the different analyses. For the chemical analysis, the sediment samples were wet-sieved using a 0.125 mm mesh to isolate fine particulates, which are more representative of pollutant concentrations due to their higher surface area and binding capacity [26]. For the sediment toxicity tests, the samples were press-sieved using a 1 mm mesh to retain larger particles, ensuring that the sediment fractions used for biological testing reflected natural conditions relevant to benthic organisms [27]. All the water and sediment samples were stored in a refrigerator at a temperature of 4 °C until they were ready for analysis and analyzed within one month after collection.

2.2.2. Chemical Analysis

In the surface water samples, dissolved concentrations of Cu, Ni, Zn, Cd, Pb, As, and major cations (Ca, Mg, Na, and K) were analyzed using inductively coupled plasma-mass spectrometry (ICP-MS, Agilent 7500a; Agilent Technologies, Santa Clara, CA, USA) following filtration through 0.45 μm filters. The dissolved organic carbon (DOC) concentrations were measured using a total organic carbon analyzer (O.I. Aurora 1030C; Aurora Biotechnologies, Carlsbad, CA, USA) with combustion and infrared detection after filtration. The alkalinity was determined using the titration method 2320 from Standard Methods for the Examination of Water and Wastewater (https://www.standardmethods.org/doi/10.2105/SMWW.2882.023, accessed on 31 October 2024). The pH measurements were conducted in the field using a portable pH meter (Orion 710a; Thermo Scientific, Waltham, MA, USA).

For the sediment samples, the total concentrations of Cu, Ni, Zn, Cd, Pb, As, and Cr were determined using the total digestion method. This method was the official testing standard of environmental pollution in Korea (Ministry of Environment, Chapter 2, Article 6) and is described [21]. Briefly, 0.2 g of powdered sediment was digested with a mixed acid solution (nitric acid, perchloric acid, and hydrofluoric acid). The digest was treated with 5 mL of saturated boric acid and 1 mL of perchloric acid, and the solution was evaporated. The remaining residues were dissolved in 20 mL of 1% nitric acid. The dissolved metals were filtered through a 0.45 μm membrane filter and analyzed by ICP-MS, using the same protocol as for water samples.

For quality control of the chemical analysis, material data from the SRM 1640a of National Institute of Standards and Technology (NIST) for a natural water sample and the NRCC PACS-2 of National Research Council of Canada (NRC CNRC) for a sediment sample were combined and compared with the reference values. The recoveries were estimated to be 83–122%. The detailed QA/QC data are provided in Table S3.

2.3. Ecotoxicological Monitoring

The ecotoxicological status of the Andong watershed was assessed through a comprehensive evaluation using multiple ecological receptors representing different environmental compartments. Whole sediment toxicity was assessed using Hyalella azteca and Chironomus riparius, while sediment elutriate toxicity was evaluated using Daphnia magna and Oncorhynchus mykiss. Acute mortality tests were conducted on field-collected sediments using H. azteca and C. riparius, whereas reproduction and growth inhibition tests were performed on sediment elutriates using D. magna and O. mykiss, following standardized protocols.

2.3.1. Sediment Toxicity Test

Sediment toxicity tests were conducted using the amphipod H. azteca and the chironomid C. riparius. The tests with H. azteca followed the United States Environmental Protection Agency (US EPA) method (EPA 600/R-99/064) [28]. Approximately 100 mL of sediment was placed in 300 mL glass beakers, and 175 mL of reconstituted water was carefully added without disturbing the sediment. Four replicates were prepared for each sediment sample, with 10 test organisms introduced into each beaker. The organisms, ranging in size from 350 to 500 μm and aged between 7 and 14 days, were exposed for 10 days in an incubator at 23 °C under a 16:8 h light photoperiod.

The C. riparius sediment toxicity tests followed the OECD Test Guideline No. 218 [29]. The sediment (200 mL) was placed in 1 L glass beakers, and 800 mL of reconstituted water was carefully added to maintain a 1:4 sediment-to-water ratio. Four replicates were prepared for each sediment sample, and 10 first-instar larvae were added to each beaker. The beakers were incubated at 20 ± 2 °C under a 16:8 h light photoperiod for 20 days. The L/EC50 values for both test species were calculated using probit regression analysis with TOXCALC version 2.0 (Tidepool Scientific Software).

2.3.2. Sediment Elutriate Toxicity Test

The sediment elutriate samples were prepared following the guidelines provided by the American Society for Testing and Materials (ASTM Guide E 1391) [30]. A 1:4 (v/v) sediment-to-water ratio was used, and the mixture was agitated on a rotary shaker at 100 rpm for 1 h. After agitation, the samples were centrifuged at 3000 rpm for 20 min to separate the aqueous phase (elutriate) from the sediment. The elutriate was stored at 4 °C in cubitainers and used for acute toxicity testing within 24 h. The remaining sediment was stored at 4 °C for use in bulk sediment toxicity tests.

Reproduction tests with D. magna were conducted according to OECD Test Guideline No. 211 [31], while juvenile growth inhibition tests with O. mykiss followed OECD Test Guideline No. 215 [32]. Both tests were conducted at 20 ± 2 °C under a 16:8 h light photoperiod. The test media were renewed three times per week. The chemical parameters, including pH, competing cations, dissolved organic carbon, and alkalinity, were measured at the start of the test (0 h), during media renewals, and at the conclusion of the test (21 days for D. magna and 28 days for O. mykiss) for each test treatment and control sample.

2.4. Bioavailability-Based Ecotoxicity Prediction Model for Metals in Surface Water

The predicted no-effect concentrations (PNECs) for heavy metals in freshwater were calculated to account for the variability in ecotoxicity arising from different water quality conditions. The bioavailability models specific to each heavy metal were applied. For Cu, Ni, and Zn, the biotic ligand model (BLM) was employed to derive the PNECs, following the methodology used to establish environmental quality standards (EQSs) in Europe [33,34,35]. The BLM allows for site-specific predictions of toxicity by incorporating species sensitivity distributions (SSDs) for aquatic organisms.

A general formula for the BLM is illustrated in the equation for the D. magna BLM for Cu [36]:

(1)$E{C}_{x,M{e}^{2+}}={\displaystyle \frac{f_{MeBL}^{x\%}}{(1-f_{MeBL}^{x\%})K_{MeBL}}}\left({\displaystyle \frac{1+K_{CaBL}\left[C{a}^{2+}\right]+K_{MgBL}\left[M{g}^{2+}\right]+K_{NaBL}\left[N{a}^{+}\right]+K_{HBL}\left[H^{+}\right]}{1+R_{MeOHBL}{K}_{MeOH}\left[O{H}^{-}\right]+R_{MeC{O}_{3}BL}{K}_{MeC{O}_{3}BL}{K}_{MeC{O}_3}\left[C{O}_3^{2-}\right]}}\right)$

where $f_{MeBL}^{x\%}$ represents the fraction of metal binding sites occupied at a given toxic effect level x%, and K refers to biotic ligand constants for different model species, such as $K_{CaBL}$, $K_{MgBL}$, $K_{NaBL}$, $K_{HBL}$, and R referes to ratios of constants representing metal binding to organic and inorganic ligands such as $R_{MeOHBL}$ (=$K_{MeOHBL}/K_{MeBL}$) and $R_{MeC{O}_{3}BL}$ (=$K_{MeC{O}_{3}BL}/K_{MeBL}$). The biotic ligand and inorganic stability constants for the Cu, Ni, and Zn BLMs, applied to taxonomic groups, are presented in Table S1.

To derive the PNECs for these three metals, the BLMs for three taxonomic groups (microalgae, invertebrates, and vertebrates) were utilized to calculate the HC5 (5th percentile value of the SSD) in accordance with the Technical Guidance Document No. 27 for environmental quality standards derivation [35]. The influence of water quality variables such as pH, dissolved organic carbon (DOC), and calcium concentrations on the PNECs varies due to differences in the model structure and biotic ligand constants for Cu, Ni, and Zn BLMs.

The Windermere Humic Aqueous Model 7 (WHAM7) was employed to estimate the free ion activities of Cu and competing cations in the surface water samples [37]. The water quality variables from the surface water samples, including the DOC and estimated fulvic acid concentrations, served as the input parameters for WHAM7. The fraction of DOC assumed to be fulvic acid was 50%, 61%, and 40% for Cu, Ni, and Zn, respectively [33,34,38].

For Cd and Pb, PNECs were derived using empirical linear regression models based on water hardness, as applied in the United States to derive water quality criteria (WQC). The general formula for the linear regression model for Cd and Pb is shown as follows:

(2)$PNE{C}_{CdorPb}=e^{(b+SlnHardness)}CF$

where b and S are the slope and intercept values for Cd (−3.909 and 0.7977, respectively) [39] and Pb (−4.705 and 1.273, respectively) [40].

Conversion factors ($CFs$) were applied to convert total metal concentrations to dissolved forms, calculated as follows:

(3)$CF=b_{CF}-S_{CF}(lnHardness)$

with constants $b_{CF}$ and $S_{CF}$ being 1.101672 and 0.041838 for Cd and 0.145712 and 1.46203 for Pb.

This bioavailability-based approach was applied to five metals (Cd, Ni, Zn, Cu, and Pb) to account for the metal bioavailability in the water column. However, arsenic (As) bioavailability was not considered in this study.

2.5. Data Analysis

To assess the concentration–response relationships in the field-collected samples, a mixture toxicity model was applied. This model incorporated the sum of toxic units (ΣTU) derived from metal concentrations in sediment elutriates and the mean values of the probable effect level (PEL) quotient for sediment metal concentrations as explanatory variables in a logistic regression analysis to predict the probability of toxicity.

2.5.1. Ecological Risk Assessment of Metals in Water and Sediment

The ecological risk of metals in water samples was evaluated using the sum of hazard quotients (HQs) for individual metals. HQs were calculated by dividing the measured concentration of each metal by its PNEC. For Cu, Ni, and Zn, BLM-based PNECs were derived using predicted site-specific toxicity values for whole organisms in a species sensitivity distribution (SSD) analysis. The HC5 values (hazardous concentration affecting 5% of species) were calculated using ETX 2.0 software [41], assuming the log-normal distribution of the SSDs. The normality of species sensitivity data was verified using Anderson–Darling, Kolmogorov–Smirnov, and Cramér–von Mises tests. For Cd and Pb, hardness-based PNECs were calculated directly using Equation (2), negating the need for an SSD analysis. The PNEC for arsenic (147.9 $\mathsf{\mu }$g/L) was adopted from the US EPA water quality criteria for arsenic (III) [40].

The sediment ecological risk was assessed using the mean probable effect level (PEL) quotient. The PEL values for each metal in the freshwater sediment were obtained from [42] (Table A2). The PEL quotient was calculated by dividing the measured concentration of each metal by its PEL value and then averaging these quotients across all the metals present in the sediment sample. This approach provided a comprehensive evaluation of ecological risk for both the water and sediment samples, incorporating metal-specific toxicity thresholds and accounting for site-specific conditions where applicable.

2.5.2. Mixture Toxicity Models for Prediction of the Probability of Toxicity

To predict the probability of toxicity in field samples, mixture toxicity models were employed. For sediment elutriates, the sum of toxic units ($\mathsf{\Sigma }TU$) was calculated as the sum of the toxic units of each metal, where each metal’s $\mathsf{\Sigma }TU$ was determined by dividing its concentration by its toxicity threshold. Similarly, for sediment samples, the mean PEL quotient (mPELq) was calculated and used to assess the combined toxicity of multiple metals. These mixture toxicity models were applied to predict the probability of toxicity using logistic regression analysis, with $\mathsf{\Sigma }TU$ and mPELq as the independent variables.

2.5.3. Logistic Regression Analysis

A logistic regression (LR) analysis was performed to evaluate the relationships between the metal mixtures and toxicity in the sediment and water samples. Separate LR analyses were conducted for mixtures of metals and for individual metal elements. In the case of metal mixtures, $\mathsf{\Sigma }TU$ (for sediment elutriates) and the mean PEL quotient (for sediments) served as predictors of the probability of toxicity such as acute mortality or reproductive inhibition in test organisms. For individual metals, the LR analysis focused on the PEL quotient or toxic unit of each metal as the explanatory variable, allowing for the identification of key toxicants driving the observed toxicity in the sediment and water samples. The LR analysis was performed using the following equation [13,14,15]:

(4)$p={\displaystyle \frac{exp({\beta }_0+{\beta }_1\left(x\right))}{1+exp({\beta }_0+{\beta }_1\left(x\right))}}$

where the p is the probability of observing a toxic effect, the ${\beta }_0$ is an intercept parameter, the ${\beta }_1$ is a slope parameter, and the x is a log-transformed explanatory variable, such as TUs and PEL quotients.

3. Results

3.1. Chemical Status and Distribution

The metal concentrations and their corresponding bioavailability model-derived predicted no-effect concentrations (PNECs) in surface water, along with the sediment metal concentrations, are presented in Figures S3 and S4 of the Supplementary Material. The spatial distribution patterns of contamination in both the water and sediment matrices are also illustrated in Figure S5.

Hazard quotient (HQ) values exceeding 1 indicated severe metal contamination near and downstream of the Zn smelter and mining sites, extending up to 93 km downstream, including the entrance to Andong Lake (Figure 2). The main stream HQ values for Cd and Zn consistently surpassed 1, reaching levels up to 10 times higher than those observed in the tributaries. In contrast, the HQ values for Pb, Ni, and As in some tributaries were greater than in the main stream but remained below 1. While individual metals in tributaries showed HQ values less than 1, cumulative HQ values exceeded 1 in areas near the smelter and in three downstream tributaries, where Zn, Pb, and Ni were the primary contributors (Figure 3).

Sediment metal contamination was evaluated using the mean probable effect level quotient (mPELq), where values exceeding 0.34 indicate an increased likelihood of sediment toxicity in Korea [42]. Most of the study area exhibited elevated mPELq values (>0.34), with these values decreasing as the distance from the smelter increased (Figure 4). In the riverine area, the mPELq values in September, immediately following the rainy season, were lower than those recorded in April during the dry season. However, in the Andong Lake area, no significant seasonal variation in the mPELq values was observed.

The PEL quotient (PELq) values for Cd and As frequently exceeded 1 in both the Upper Nakdong River and Andong Lake. Near the smelter, Zn, Pb, Cu, and Hg also had PELq values greater than 1. Across the Upper Nakdong River, the PELq values for Cd, Zn, Pb, Cu, and Hg were consistently higher than those in Andong Lake. In contrast, there were no significant differences in the PELq values for As, Ni, and Cr between the river and lake regions (Figure 5).

A further analysis revealed that the Cd, Zn, Hg, and Pb concentrations in the main stream sediments were higher than those in the tributary sediments, with Pb, Cu, and Hg showing particularly elevated levels near the smelter. In contrast, the As concentrations peaked in the tributary sediments, suggesting a different contamination source. These concentration gradients indicate that the Zn smelter is the primary source of Cd, Zn, Pb, and Hg contamination, while As contamination likely originates from mining activities in various tributaries.

3.2. Ecotoxicological Status and Distribution

The ecotoxicological monitoring results indicated acute sediment toxicity in the samples collected upstream, near the Zn smelter, and within Andong Lake. Chronic and subchronic elutriate toxicity were primarily observed in Andong Lake and at three sites downstream of the Zn smelter (Figure 6).

Most samples from Andong Lake exhibited $\mathsf{\Sigma }TU$ values exceeding 1. Similarly, $\mathsf{\Sigma }TU$ values greater than 1 were also detected in the main Nakdong River near the smelter, as well as in some tributary sediments downstream (Figure 7).

3.3. Logistic Regression Analysis

3.3.1. Sediment Toxicity

The logistic regression analysis revealed a significant correlation between increasing m-PELq values and the probability of sediment toxicity in H. azteca and C. riparius (Figure 8). The cutoff values for sediment toxicity were 0.83 mPELq for H. azteca and 0.84 mPELq for C. riparius, with model reliabilities of 71% and 65%, respectively. These cutoff values are more than double the existing benchmark value of 0.34 mPELq, which had an overall reliability of 66% [42].

Except for Pb in C. riparius, all of the metals were significantly associated with the probability of toxicity with an accuracy over 70% for H. azteca and 63% for C. riparius, although only Zn and As exceeded a PELq of 1, with a cutoff value of 1.79 and 0.80 PELq, respectively (Figure A2 and Figure A3).

3.3.2. Sediment Elutriate Toxicity

For sediment elutriate toxicity, the probability of toxicity increased significantly with a $\mathsf{\Sigma }TU$ (p < 0.005 for D. magna and p < 0.01 for O. mykiss) (Figure 9). The cutoff values for sediment elutriate toxicity were 6.17 $\mathsf{\Sigma }TU$ for the Daphnia reproduction test and 1.42 $\mathsf{\Sigma }TU$ for the fish juvenile growth test, with overall model reliabilities of 67% and 75%, respectively.

For D. magna, Zn, As, and Cu were significantly associated with the probability of toxicity, although only Zn exceeded the toxic unit threshold with a cutoff value of 0.69 TU (Figure A4). In O. mykiss, Zn, As, Ni, and Pb were significantly related to toxicity probability, with Zn being the only metal to surpass the toxic unit threshold, having a cutoff value of 0.89 TU.

4. Discussion

4.1. Chemical and Ecotoxicological Status of Water, Sediment, and Sediment Elutriate

The ecotoxicological risks associated with individual metals in field-collected water and sediment samples were quantified using additive mixture models. For the water samples, $\mathsf{\Sigma }HQ$ was employed, while mPELq was used for the sediments. Our analysis revealed distinct spatial patterns in metal contamination and their contributions to ecotoxicological risks. In the water samples, Cd and Zn were the dominant contributors to HQ values within 60 km downstream of the smelter. However, beyond this distance, particularly in lake water, the contribution of Cd decreased, while Zn, Ni, and Cu became more significant contributors to the total HQ. Sediment elutriate toxicity tests on D. magna showed a similar trend, with Zn as the primary contributor, followed by Ni and Cu.

Interestingly, the sediments exhibited a different contamination profile. Cd and As, with PELq values exceeding 1, were the primary contributors to the mPELq in the Upper Nakdong River and Andong Lake areas. Pb, Zn, Cu, and Hg also showed a PELq exceeding 1 near the smelter. This observed divergence between metals contributing to the contamination in the surface water and sediment underscores the importance of assessing both contaminated sediments and the ecotoxicological status of sediment elutriates for a comprehensive understanding of pollution dynamics in the watershed.

Our investigation traced the transport of metals (Cd and Zn in water; Cd and As in sediments) to pollution sources approximately 130 km upstream in the Nakdong River. As a result, extensive areas in the upper zone of Andong Lake and the lower zone near Andong Dam were classified as second-level contamination zones, indicating a high probability of acute sediment toxicity.

To evaluate the potential of contaminated sediments acting as secondary sources of metal pollution, we conducted sediment elutriate toxicity tests. These tests confirmed that most sediments in Andong Lake exhibited elutriate toxicity, with Zn identified as the primary contributor to this secondary pollution. This finding suggests that the resuspension of contaminated sediments could indeed exacerbate ecotoxicity in aquatic ecosystems. These results highlight the necessity for separate evaluations of individual metals contributing to HQs for water, TUs for sediment elutriates, and the mPELq for sediments to provide a comprehensive assessment of the chemical and ecotoxicological status [42,43]. This multi-compartmental approach aligns with the recommendations of [44] for integrated sediment quality assessments for metals. Furthermore, our findings emphasize the critical importance of continued monitoring and management of both water and sediment quality in metal-contaminated aquatic ecosystems.

4.2. Identification of Key Toxicants and Their Sources

The identification of key toxicants within complex environmental matrices necessitates robust statistical approaches to differentiate among multiple hazardous substances. Tarnawski et al. [45] applied principal component analysis (PCA) and Pearson correlation coefficients to identify the main toxic substances among various polycyclic aromatic hydrocarbons (PAHs) and metals in sediments collected from a dam reservoir in Poland. Nevertheless, multivariate analytical techniques, including PCA, demonstrate inherent limitations when establishing concentration–response relationships, particularly for substances exhibiting elevated probable effect level quotients (PELqs) and probable effect concentration quotients (PECqs) in environmental matrices.

Moran et al. [46] identified a major toxicity contributor from 118 biocides in stream sediments in agricultural and urban areas of the Midwestern United States. Among various biocides with high detection rates and high PECq values, bifenthrin was identified as the major toxic contributor, showing a strong correlation (p < 0.01) with a biomass reduction in Hyalella azteca through a linear regression analysis. These findings demonstrate that a univariate analysis of concentration–response relationships represents a crucial methodological approach for identifying statistically significant associations between individual toxicants and observed effects.

Furthermore, even when a metal’s toxic unit (TU) or PELq approaches or exceeds the threshold value of 1, it cannot be definitively classified as a key toxicant without establishing a statistically significant association between its concentration and the observed toxic response. This approach can contribute to more accurate and reliable results in the toxicity assessment of complex pollutants in environmental samples.

To address this challenge, we employed logistic regression analysis to examine the relationship between individual metal concentrations and the probability of ecotoxicity in sediments and elutriates. Logistic regression analysis is frequently used to characterize the effects of single chemicals in water or sediment and can also be used to characterize toxicity along a gradient of metal contamination in field-collected sediment [47]. In this study, toxic units were utilized for evaluating elutriate toxicity, while PEL quotients were applied to sediment toxicity.

Our sediment tests revealed that all the metals, except Pb in C. riparius, were significantly associated with toxicity (p < 0.05). However, only Cd and As exhibited PEL values exceeding 1. This pattern likely results from high correlations among metal concentrations in Andong Lake sediments (Figure S1), suggesting that dissolved phase metals from the Upper Nakdong River contribute significantly to sediment contamination in Andong Lake. Consequently, Cd and As were identified as the key toxicants in the sediments.

The results of the elutriate toxicity analysis varied depending on the test organism. In the Daphnia reproduction test, Zn, As, and Cu showed significant associations with toxicity probability, with TU values exceeding 1 for Zn and Ni. Notably, Zn was the only metal that both had a TU value greater than 1 and exhibited a significant association with toxicity probability. In the fish juvenile growth inhibition test, Zn, Ni, Pb, Cu, and As were significantly associated with toxicity, although only Zn had a TU value above 1. These findings suggest that Zn is the primary toxicant in sediment elutriate toxicity. The high correlations among the dissolved metal concentrations in the elutriates (Figure S2) explain why several metals were significantly associated with fish juvenile toxicity, despite most having TU values below 1.

The application of additive mixture toxicity models, such as the $\mathsf{\Sigma }TU$ and mean PELq, in combination with ecotoxicological monitoring of field-collected samples, provided crucial insights into site-specific concentration–response relationships [42,48,49]. This approach was instrumental in evaluating the watershed’s ecotoxicological status and identifying key toxicants driving toxicity in this complex environmental system.

4.3. Site-Specific Concentration–Response Analysis

The analysis of concentration–response relationships for metal mixtures in field samples presents unique challenges compared to traditional approaches used for single compounds or constant-composition mixtures [50]. In our study, the variable metal composition in sediments and elutriates complicated the application of mixture models such as the $\mathsf{\Sigma }TU$ model, particularly when concentration–response slopes for individual metals varied [44]. Field-collected samples contain unknown compounds in addition to target metals, introducing inherent uncertainty in the concentration–response relationships for these complex mixtures [50]. To address this, we adopted a more robust approach by analyzing the concentration and probability of toxicity while accounting for this uncertainty, as suggested by [51].

This study emphasized the importance of statistically evaluating the relationship between toxicity probability and both the individual metal concentrations and $\mathsf{\Sigma }TU$ at a threshold point to differentiate toxic from non-toxic samples. This approach revealed that metals with TUs below 1 may still show a statistical relationship with toxicity due to strong correlations with key toxicants, while site-specific factors like metal bioavailability may result in metals with TUs above 1 not displaying significant relationships with toxicity probability. Consequently, we selected the probability of toxicity (in terms of survival, growth, or reproduction) as the dependent variable for concentration–response analyses, considering both the $\mathsf{\Sigma }TU$ and individual metal TUs. This approach allowed us to identify potential unmeasured toxicants or confirm specific metals as key toxicants based on the significance of relationships in logistic regression analyses.

Our findings establish a probabilistic concentration–response relationship for sediment metal toxicity, accounting for inherent uncertainties in mixture toxicity models [52]. These uncertainties primarily stem from variations in concentration–response slopes for individual metals, the presence of unknown toxicants, and chemical interactions within test samples. The high explanatory power of metal concentrations accounting for up to 71% of sediment toxicity and 75% of elutriate toxicity in fish juveniles demonstrates the robustness of our integrated field monitoring approach. This reinforces its value for identifying key toxicants in mixed-contaminant environments and highlights the critical need for site-specific assessments in complex contamination scenarios [53,54].

Our study’s integration of chemical monitoring with ecotoxicity testing enabled a comprehensive assessment of site-specific ecotoxicological risks. This approach allowed us to verify whether predicted risks represent actual problems under real-world conditions rather than merely potential hazards [44]. Furthermore, it facilitated the identification of key toxicants contributing to observed ecotoxicity and aided in tracing contaminant sources and pathways.

The effectiveness of combining chemical-ecotoxicological monitoring with site-specific concentration–response modeling for accurately diagnosing environmental risks in complex, multi-source contamination scenarios is a key finding of this study, supporting the conclusions of Cedergreen et al. [55]. This integrated methodology proves particularly valuable in watersheds impacted by multiple pollution sources, enabling more precise ecological impact assessments and informing targeted remediation and source management strategies [53]. Our research underscores the importance of site-specific, integrated approaches in assessing and managing complex environmental contamination scenarios, providing a robust framework for future studies and environmental management practices.

5. Conclusions

This study employed an integrated approach combining chemical and ecotoxicological monitoring to evaluate the environmental health of the Andong watershed, which is significantly impacted by a Zn smelter and adjacent mining operations. Chemical analyses indicated substantial contamination, particularly with Cd and Zn in downstream waters and Cd and As in sediments near both the Zn smelter and Andong Lake. These findings reveal clear spatial patterns of metal contamination linked to specific pollution sources. Ecotoxicological assessments corroborated these results, confirming the ecological risks associated with metal contamination in the watershed.

A logistic regression analysis further identified Cd and Zn as the primary toxicants contributing to water toxicity, while Cd and As were the dominant drivers of sediment toxicity. For sediment elutriates, Zn was identified as the key contributor. Strong site-specific concentration–response relationships for these metals, with hazard quotients (HQs) and probable effect level quotients (PELqs) consistently exceeding 1, underscore the importance of site-specific assessments over generic environmental quality benchmarks. Additionally, the detection of elutriate toxicity indicates that contaminated sediments could act as secondary sources of pollution, releasing metals back into the water column under certain conditions, which may pose long-term ecological risks for the watershed.

In conclusion, this study underscores the critical need for integrating chemical and ecotoxicological data to improve the accuracy of risk assessments in complex contamination scenarios. The findings emphasize the value of site-specific assessments in effectively predicting ecological impacts and managing metal pollution. Future studies focusing on community-level ecological responses will provide further insight into the long-term resilience and health of the ecosystem.

Author Contributions

J.C. conceived the study, developed the methodology, performed the formal analysis, wrote the original draft, and created the visualizations. S.-H.K. contributed to the resources, data curation, and investigation. D.-s.H. contributed to the formal analysis and resources. C.-G.S. contributed to the resources, data curation, and investigation. S.-D.M. contributed to the resources, data curation, and investigation. C.K. contributed to the resources, data curation, and investigation. M.C. contributed to the resources, data curation, and investigation. J.-H.L. contributed to the methodology, conducted the formal analysis, and reviewed and edited the manuscript. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The data and source code for the statistical analysis that support the findings of this study are available from the corresponding authors upon reasonable request.

Acknowledgments

This research was administratively and technically supported by the Ministry of Environment, Korean government.

Conflicts of Interest

Author Jiwoong Chung, Su-Hyun Kim, Dae-Sik Hwang, Chan-Gyoung Sung, Chankook Kim and Jonghyeon Lee were employed by the company EH Research & Consulting Co., Ltd. Author Seong-Dae Moon was employed by the company Institute of Environmental Safety and Prodection, NEB Co. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures

Figure 1. Location map of sampling stations. The left panels show the location of the Nakdong River, situated upstream of Andong Lake. The right panels depict the location of Andong Lake. Panels (A,B) indicate the sampling stations for water bodies: gray circles represent surface water sampling sites in the main stream, and white circles represent surface water sampling sites in tributaries. Panels (C,D) show the sampling stations for surface sediments.

Figure 2. Variation in the sum of the hazard quotient (HQ) values for eight metals in the surface water as a function of distance from the zinc smelter, along with the contribution of each metal to the total HQ. Andong Lake is located approximately 90 km downstream from the smelter, with the Nakdong River flowing into the lake from upstream.

Figure 3. Variation in the HQ (hazard quotient) for the individual metals in the surface water according to the distance from the Zn smelter.

Figure 4. Variation in the mean probable effect level quotient (PELq) for heavy metals in sediments as a function of the distance from the zinc smelter, along with the contribution of each metal to the mean PELq. Andong Lake is located approximately 90 km downstream from the zinc smelter, with the Nakdong River flowing into the lake from upstream.

Figure 5. Variation in the probable effect level quotient (PELq) of individual metals in sediments as a function of the distance from the Zn smelter.

Figure 6. The location map showing the sites of the samples with confirmed ecotoxic effects. The left panels (A,C) are the location map of the Nakdong River located upstream of Andong Lake. The right panels (B,D) are the location map of Andong Lake. The upper panels (A,B) show the acute toxic effects of sediments on the amphipod (Hyalella azteca) and the midge (Chironomus riparius). The lower panels (C,D) show the chronic toxic effects of sediment elutriates on the crustacean (Daphnia magna) and fish (Oncorhynchus mykiss). The red semicircles are sediment samples showing toxic effects (toxic effects exceeding 20%). The green semicircles are sediment samples that show no toxic effects (toxic effects less than 20%).

Figure 7. Variation in the sum of toxic units ([Forumla omitted. See PDF.]) for heavy metals in sediment elutriates as a function of the distance from the zinc smelter, along with the contribution of each metal to the total [Forumla omitted. See PDF.]. Andong Lake is located approximately 90 km downstream from the zinc smelter, with the Nakdong River situated between the smelter and the lake.

Figure 8. Results of logistic regression analysis for sediment toxicity tests using field-collected samples and the contribution of individual heavy metals to the mean probable effect level quotient (PELq). The upper panels illustrate the comparison between the probability of sediment toxicity and the mean PELq. The lower panels show the contribution of each metal to the mean PELq in sediment samples where toxicity was confirmed for test species Hyalella azteca and Chironomus riparius.

Figure 9. Results of logistic regression analysis for sediment elutriate toxicity tests using field-collected samples and the contribution of individual heavy metals to the sum of toxic units ([Forumla omitted. See PDF.]). The upper panels present the comparison between the probability of sediment elutriate toxicity and the [Forumla omitted. See PDF.]. The lower panels illustrate the contribution of each metal to the [Forumla omitted. See PDF.] in sediment elutriate samples where ecotoxicity was confirmed for test species Daphnia magna and Oncorhynchus mykiss.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/w16223176/s1, Tablel S1: Species- and element-specific parameters of zinc, nickel, and copper BLMs; Table S2: Sediment quality assessment guidelines for eight metals in Korean freshwater sediments; Figure S1: Correlation among individual metal concentrations in sediment toxicity test samples collected from the Nakdong River and Andong Lake. Below the diagonal are all the pairwise scatter plots comparing each pair of metals. Above the diagonal is the coefficient of determination; Figure S2: Correlation among individual metal concentrations in sediment elutriate toxicity test samples from the Nakdong River and Andong Lake. Below the diagonal are all the pairwise scatter plots comparing each pair of metals. Above the diagonal is coefficient of determination.

References

1. Burton, G.A.; Chapman, P.M.; Smith, E.P. Weight-of-Evidence Approaches for Assessing Ecosystem Impairment. Hum. Ecol. Risk Assess. Int. J.; 2002; 8, pp. 1657-1673. [DOI: https://dx.doi.org/10.1080/20028091057547]

2. Chapman, P.M. Determining when contamination is pollution — Weight of evidence determinations for sediments and effluents. Environ. Int.; 2007; 33, pp. 492-501. [DOI: https://dx.doi.org/10.1016/j.envint.2006.09.001] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17027966]

3. Suter, G.W.; Cormier, S.M. Why and how to combine evidence in environmental assessments: Weighing evidence and building cases. Sci. Total Environ.; 2011; 409, pp. 1406-1417. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2010.12.029] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21277006]

4. Luoma, S.N.; Rainbow, P.S. Metal Contamination in Aquatic Environments: Science and Lateral Management; Cambridge University Press: Cambridge, UK, 2008; [DOI: https://dx.doi.org/10.1017/CBO9780511490531]

5. Suter, G.W.; Norton, S.B.; Cormier, S.M. The science and philosophy of a method for assessing environmental causes. Hum. Ecol. Risk Assess. Int. J.; 2017; 23, pp. 1359-1367. [DOI: https://dx.doi.org/10.1080/10807030903459254]

6. Burgess, R.M.; Berry, W.J.; Mount, D.R.; Di Toro, D.M. An integrated contaminant-sediment exposure approach for the assessment of sediment hazard. Environ. Toxicol. Chem.; 2013; 32, pp. 102-114. [DOI: https://dx.doi.org/10.1002/etc.2025] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23060276]

7. Chapman, P.M.; McDonald, B.G.; Lawrence, G.S. A decision-making framework for sediment contamination. Hum. Ecol. Risk Assess.; 2002; 8, pp. 1489-1507. [DOI: https://dx.doi.org/10.1080/20028091057457]

8. Kwok, K.W.; Batley, G.E.; Wenning, R.J.; Zhu, L.; Vangheluwe, M.; Lee, S.Y. Sediment quality guidelines: Challenges and opportunities for improving sediment management. Environ. Sci. Pollut. Res.; 2014; 21, pp. 17-27. [DOI: https://dx.doi.org/10.1007/s11356-013-1778-7] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23673922]

9. Ankley, G.T.; Di Toro, D.M.; Hansen, D.J.; Berry, W.J. Technical basis and proposal for deriving sediment quality criteria for metals. Environ. Toxicol. Chem.; 1996; 15, pp. 2056-2066. [DOI: https://dx.doi.org/10.1002/etc.5620151202]

10. Di Toro, D.M.; Allen, H.E.; Bergman, H.L.; Meyer, J.S.; Paquin, P.R.; Santore, R.C. Biotic ligand model of the acute toxicity of metals. 1. Technical basis. Environ. Toxicol. Chem.; 2001; 20, pp. 2383-2396. [DOI: https://dx.doi.org/10.1002/etc.5620201034] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11596774]

11. Long, E.R.; MacDonald, D.D. Recommended uses of empirically derived, sediment quality guidelines for marine and estuarine ecosystems. Hum. Ecol. Risk Assess.; 1998; 4, pp. 1019-1039. [DOI: https://dx.doi.org/10.1080/10807039891284956]

12. MacDonald, D.D.; Ingersoll, C.; Berger, T. Development and evaluation of consensus-based sediment quality guidelines for freshwater ecosystems. Arch. Environ. Contam. Toxicol.; 2000; 39, pp. 20-31. [DOI: https://dx.doi.org/10.1007/s002440010075] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10790498]

13. Field, L.J.; MacDonald, D.D.; Norton, S.B.; Severn, C.G.; Ingersoll, C.G. Evaluating sediment chemistry and toxicity data using logistic regression modeling. Environ. Toxicol. Chem. Int. J.; 1999; 18, pp. 1311-1322. [DOI: https://dx.doi.org/10.1002/etc.5620180634]

14. Field, L.J.; MacDonald, D.D.; Norton, S.B.; Ingersoll, C.G.; Severn, C.G.; Smorong, D.; Lindskoog, R. Predicting amphipod toxicity from sediment chemistry using logistic regression models. Environ. Toxicol. Chem.; 2002; 21, pp. 1993-2005. [DOI: https://dx.doi.org/10.1002/etc.5620210929]

15. Smith, E.P.; Robinson, T.; Field, L.J.; Norton, S.B. Predicting sediment toxicity using logistic regression: A concentration-addition approach. Environ. Toxicol. Chem. Int. J.; 2003; 22, pp. 565-575.

16. Park, J.J.; Kim, K.J.; Yoo, S.M.; Kim, E.H.; Seok, K.S.; Huh, I.A.; Kim, Y.H. Distribution of heavy metals and mercury in sediment from the lake An-dong. Anal. Sci. Technol.; 2012; 25, pp. 441-446. [DOI: https://dx.doi.org/10.5806/AST.2012.25.6.441]

17. Kim, S.; Kim, J.; Lee, K.; Lee, K.; Jeon, H.; Yu, J.; Lee, I.; Ahn, J. Distribution and pollution of heavy metals in surface sediments from Nakdong river. J. Environ. Sci. Int.; 2015; 24, pp. 969-980. [DOI: https://dx.doi.org/10.5322/JESI.2015.24.8.969]

18. Kim, J.; Choi, U.; Baek, S.H.; Choi, H.B.; Lee, J. A study on chemical compositions of sediment and surface water in Nakdong river for tracing contaminants from mining activities. J. Korean Earth Sci. Soc.; 2016; 37, pp. 211-217. [DOI: https://dx.doi.org/10.5467/JKESS.2016.37.4.211]

19. Choi, M.; Park, J.; Cho, D.; Jang, D.; Kim, M.; Choi, J. Tracing metal sources in core sediments of the artificial lake An-Dong, Korea: Concentration and metal association. Sci. Total Environ.; 2015; 527, pp. 384-392. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2015.05.013] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25981937]

20. Joe, D.J.; Choi, M.S.; Lee, J.H.; Kim, C.K.; Choi, M.S.; Shin, H.S. Discrimination of metal contaminant sources in river sediments influenced by mining and smelting activities using stable Pb and Zn isotopes. Environ. Sci. Pollut. Res.; 2024; 31, pp. 20521-20533. [DOI: https://dx.doi.org/10.1007/s11356-024-32508-6]

21. Jung, J.; Park, J.; Choi, Y.; Choe, J.K.; An, J.; Nam, K. Environmental forensic approach towards unraveling contamination sources with receptor models: A case study in Nakdong River, South Korea. Sci. Total Environ.; 2023; 892, 164554. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2023.164554] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37270013]

22. Farhat, H.I.; Gad, A.; Saleh, A.; Abd El Bakey, S.M. Risks assessment of potentially toxic elements’ contamination in the Egyptian Red sea surficial sediments. Land; 2022; 11, 1560. [DOI: https://dx.doi.org/10.3390/land11091560]

23. Yongo, E.; Jin, F.; Mutethya, E.; Wu, D.; Zhang, P.; Guo, Z. Sediment heavy metal pollution assessment in Changwang and Wuyuan Rivers in Hainan Island, China. Water; 2023; 15, 1580. [DOI: https://dx.doi.org/10.3390/w15081580]

24. Ganea, I.V.; Bălc, R.; Begy, R.C.; Tanțău, I.; Gligor, D.M. Combining Contamination Indices and Multivariate Statistical Analysis for Metal Pollution Evaluation during the Last Century in Lacustrine Sediments of Lacu Sărat Lake, Romania. Int. J. Environ. Res. Public Health; 2023; 20, 1342. [DOI: https://dx.doi.org/10.3390/ijerph20021342] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36674098]

25. Wei, H.; Wang, Y.; Liu, J.; Zeng, R. Heavy metal in river sediments of Huanghua City in water diversion area from Yellow River, China: Contamination, ecological risks, and sources. Water; 2022; 15, 58. [DOI: https://dx.doi.org/10.3390/w15010058]

26. Horowitz, A.J. A Primer on Trace Metal-Sediment Chemistry; United States Geological Survey Water-Supply Paper US Geological Survey: Reston, VA, USA, 1985; Volume 2277, 67.

27. U.S. Environmental Protection Agency. Methods for Assessing the Toxicity of Sediment-Associated Contaminants with Estuarine and Marine Amphipods; Technical Report EPA/600/R-94/025 Office of Research and Development, U.S. Environmental Protection Agency: Washington, DC, USA, 1994.

28. U.S. Environmental Protection Agency. Methods for Measuring the Toxicity and Bioaccumulation of Sediment-Associated Contaminants with Freshwater Invertebrates; 2nd ed. EPA/600/R-99/064 Office of Water: Washington, DC, USA, 2000.

29. OECD. Test No. 218: Sediment-Water Chironomid Toxicity Using Spiked Sediment; OECD Publishing: Paris, France, 2004.

30. ASTM International. Standard Guide for Collection, Storage, Characterization, and Manipulation of Sediments for Toxicological Testing and for Selection of Samplers Used to Collect Benthic Invertebrates; ASTM International: West Conshohocken, PA, USA, 2000.

31. OECD. Test no. 211: Daphnia Magna Reproduction Test; OECD Publishing: Paris, France, 2012.

32. OECD. Test no. 215: Fish, juvenile growth test. OECD Guid. Test Chem; 2000; 1, pp. 1-16.

33. ECB. European Union Risk Assessment Report, NICKEL; Office for Official Publications of the European Communities: Luxembourg, 2008.

34. ECB. European Union Risk Assessment Report, ZINC; Office for Official Publications of the European Communities: Luxembourg, 2010.

35. EUW Directors. Guidance document No. 27: Technical Guidance for deriving environmental quality standards. Common Implementation Strategy for the Water Framework Directive (2000/60/EC). Updated Version; European Commission: Brussels, Belgium, 2018.

36. De Schamphelaere, K.A.; Janssen, C.R. A biotic ligand model predicting acute copper toxicity for Daphnia magna: The effects of calcium, magnesium, sodium, potassium, and pH. Environ. Sci. Technol.; 2002; 36, pp. 48-54. [DOI: https://dx.doi.org/10.1021/es000253s] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11817370]

37. Lofts, S. User’s Guide to WHAM7; NERC Centre for Ecology and Hydrology: Lancaster, UK, 2012; pp. 221-230.

38. Van Sprang, P.; Verdonck, F.; Van Assche, F.; Regoli, L.; De Schamphelaere, K. Environmental risk assessment of zinc in European freshwaters: A critical appraisal. Sci. Total Environ.; 2009; 407, pp. 5373-5391. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2009.06.029]

39. U.S. Environmental Protection Agency. AQUATIC LIFE AMBIENT WATER QUALITY CRITERIA CADMIUM-2016; U.S. Environmental Protection Agency: Washington, DC, USA, 2016.

40. U.S. Environmental Protection Agency. 1995 Updates: Water Quality Criteria Documents for the Protection of Aquatic Life in Ambient Water; Office of Water: Washington, DC, USA, 1996.

41. Van Vlaardingen, P.; Traas, T.; Wintersen, A.; Aldenberg, T. A Program to Calculate Hazardous Concentrations and Fraction Affected, Based on Normally Distributed Toxicity Data; National Institute for Public Health and the Environment (RIVM): Bilthoven, The Netherlands, 2004.

42. Jin, D.R.; Lee, C.H.; Kim, C.K.; Moon, S.D.; Huh, I.A.; Kim, S.; Park, M.S.; Lee, J.H. Development and validation of freshwater sediment quality assessment guidelines for trace elements in Korea. Environ. Eng. Res.; 2022; 27, 200643. [DOI: https://dx.doi.org/10.4491/eer.2020.643]

43. Long, E.R.; Ingersoll, C.G.; MacDonald, D.D. Calculating revised screening-level PELs for sediment contamination by mixtures of chemicals. Environ. Toxicol. Chem.; 2006; 25, pp. 2620-2636.

44. Fairbrother, A.; Wenstel, R.; Sappington, K.; Wood, W. A framework for metals risk assessment. Ecotoxicol. Environ. Saf.; 2007; 68, pp. 145-227. [DOI: https://dx.doi.org/10.1016/j.ecoenv.2007.03.015]

45. Tarnawski, M.; Baran, A. Use of chemical indicators and bioassays in bottom sediment ecological risk assessment. Arch. Environ. Contam. Toxicol.; 2018; 74, pp. 395-407. [DOI: https://dx.doi.org/10.1007/s00244-018-0513-2] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29487958]

46. Moran, P.W.; Nowell, L.H.; Kemble, N.E.; Mahler, B.J.; Waite, I.R.; Van Metre, P.C. Influence of sediment chemistry and sediment toxicity on macroinvertebrate communities across 99 wadable streams of the Midwestern USA. Sci. Total Environ.; 2017; 599, pp. 1469-1478. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2017.05.035] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28531955]

47. Besser, J.M.; Brumbaugh, W.G.; Ingersoll, C.G. Characterizing toxicity of metal-contaminated sediments from mining areas. Appl. Geochem.; 2015; 57, pp. 73-84. [DOI: https://dx.doi.org/10.1016/j.apgeochem.2014.05.021]

48. Chapman, P.M.; Wang, F.; Caeiro, S.S. Integrative sediment quality assessment in estuarine-coastal ecosystems. Treatise Estuar. Coast. Sci.; 2013; 5, pp. 125-162.

49. Backhaus, T.; Faust, M. Identifying the environmental risk of chemical mixtures. Environ. Sci. Technol.; 2012; 46, pp. 2564-2573. [DOI: https://dx.doi.org/10.1021/es2034125]

50. Jonker, M.J.; Svendsen, C.; Bedaux, J.J.; Bongers, M.; Kammenga, J.E. Significance testing of synergistic/antagonistic, dose level-dependent, or dose ratio-dependent effects in mixture dose-response analysis. Environ. Toxicol. Chem.; 2005; 24, pp. 2701-2713. [DOI: https://dx.doi.org/10.1897/04-431R.1] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16268173]

51. Posthuma, L.; Dyer, S.D.; de Zwart, D.; Kapo, K.; Holmes, C.M.; Burton, G.A., Jr. Eco-epidemiology of aquatic ecosystems: Separating chemicals from multiple stressors. Sci. Total Environ.; 2016; 573, pp. 1303-1319. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2016.06.242]

52. Backhaus, T.; Faust, M. Predictive Environmental Risk Assessment of Chemical Mixtures: A Conceptual Framework. Environ. Sci. Technol.; 2012; 46, pp. 2564-2573. [DOI: https://dx.doi.org/10.1021/es2034125]

53. Beyer, J.; Petersen, K.; Song, Y.; Ruus, A.; Grung, M.; Bakke, T.; Tollefsen, K.E. Environmental risk assessment of combined effects in aquatic ecotoxicology: A discussion paper. Mar. Environ. Res.; 2014; 96, pp. 81-91. [DOI: https://dx.doi.org/10.1016/j.marenvres.2013.10.008]

54. Simpson, S.L.; Batley, G.E. Sediment Quality Assessment: A Practical Guide; CSIRO Publishing: Clayton, VIC, Australia, 2016.

55. Cedergreen, N. Quantifying synergy: A systematic review of mixture toxicity studies within environmental toxicology. PLoS ONE; 2014; 9, e96580. [DOI: https://dx.doi.org/10.1371/journal.pone.0096580]