Integrated Analysis of Heavy-Metal Pollution in Three Gorges Reservoir Sediments: Spatial Distribution, Source Apportionment, and Ecological Risk Assessment

Abstract

The Three Gorges Reservoir, serving as a crucial ecological barrier for the middle-lower Yangtze River basin, faces substantial threats to watershed ecosystems from sediment-associated heavy metal, threatening aquatic ecosystems and human health via bioaccumulation. Leveraging the legislative framework of the Yangtze River Protection Law, this study analyzed sediment cores (0–65 cm) collected from 12 representative sites in the Three Gorges Reservoir using 2020 Air–Space–Ground integrated monitoring data from the Changjiang Water Resources Commission. Concentrations of nine heavy metals (Cd, Cu, Pb, Fe, Mn, Cr, As, Hg, and Zn) were quantified to characterize spatial and vertical distribution patterns. Source apportionment was conducted through correlation analysis and principal component analysis (PCA). Contamination severity and ecological risks were assessed via geo-accumulation index (I_geo), potential ecological risk index (RI), and acute toxicity metrics. The findings indicated substantial spatial heterogeneity in sediment heavy-metal concentrations, with the coefficients of variation (CV) for Hg and Cd reaching 214.46% and 116.76%, respectively. Cu and Pb showed surface enrichment, while Cd exhibited distinct vertical accumulation. Source apportionment indicated geogenic dominance for most metals, with anthropogenic contributions specifically linked to Cd and Hg enrichment. Among the metals assessed, Cd emerged as the primary ecological risk driver, with localized strong risk levels (E_i > 320), particularly at FP and SS sites. These findings establish a scientific foundation for precision pollution control and ecological restoration strategies targeting reservoir sediments.

1. Introduction

The Three Gorges Reservoir serves as both a pivotal ecological barrier [1,2], abundant in natural resources and species diversity within China’s southwestern region [3], and a critical ecologically sensitive area safeguarding flood control security, water quality, and ecological equilibrium in the middle-lower Yangtze River [4]. Its prominent ecological functional status has consequently heightened widespread concern regarding the ecological integrity of the reservoir environment [2,5,6]. However, the reservoir’s distinctive cyclical and substantial water-level fluctuations induce markedly diminished flow velocity, protracted hydraulic retention time, and substantially enhanced sedimentation of suspended solids and particulate matter [5,7]. This process culminates in the formation of extensive sediment accumulation areas within reservoir bays, low-velocity regions, and drawdown zones [8].

The drawdown zone plays a dual role during cyclical water-level fluctuations [7,9]. During low-water phases, contaminants including heavy metals—from mining, industrial/urban discharges, agricultural runoff, intensive navigation, and sediment/bound pollutants from upstream inflows—undergo sedimentation and enrichment, establishing the zone as a pollution sink [10]. Conversely, during high-water phases, inundation exposes sediments to alternating desiccation–wetting cycles and dramatic redox fluctuations, which facilitate contaminants’ liberation back into the water column via desorption and dissolution, thereby transforming the sediments into a potential pollution source [7]. Furthermore, high-intensity seasonal dry–wet cycles and summer rainfall erosion markedly alter the physicochemical properties of drawdown zone sediments [5,11]. These alterations profoundly influence pollutant bioavailability, migration, and release behaviors. Consequently, sediments serve as critical pollutant sinks and sources, exerting a governing influence on the migration and transformation dynamics of contaminants [12,13].

With the accelerated development of the Yangtze River Economic Belt [14], heavy-metal contamination in aquatic environments has escalated significantly [15], particularly concerning heavy metals in sediments, which have emerged as a prominent research focus in environmental protection studies [5]. Heavy metals enriched in sediments pose severe threats to the reservoir’s ecological integrity due to their persistence, bioaccumulation potential, and inherent toxicity. Through biomagnification within food chains, they not only exert direct toxicity on benthic organisms and destabilize aquatic ecosystem structure but may also jeopardize aquatic biodiversity both within the reservoir and downstream regions [16]. Crucially, the mechanisms of desorption, dissolution, and rainfall erosion within the drawdown zone can directly facilitate the transport of mobilized heavy metals into the water column, thereby posing latent risks to regional ecological security and human health, while creating persistent and direct hazards for the security of downstream drinking water sources and water resource utilization [5,17,18].

In recent years, several approaches have been developed for assessing the environmental impact of heavy metals in sediments, including the geo-accumulation index (I_geo), potential ecological risk index (RI), and sediment quality guidelines (SQGs) [19,20]. Driven by evolving contemporary water governance philosophies and Yangtze River Economic Belt’s development, substantial research efforts have yielded significant advances into ecological risk assessment of heavy metals in the Three Gorges Reservoir. Liang et al. identified Hg, Cd, and As in Changshou Lake as moderately to heavily polluted [21]. Herath et al. demonstrated that Pb, Cd, and Zn pose significant threats to the ecosystem of Three Gorges Reservoir [22]. Recent studies confirm Cd-dominated ecological risks around coal gangue dumps in Three Gorges Reservoir despite all coal mines being closed in 2020 [23]. Furthermore, integrated assessments—incorporating species sensitivity distribution models, entropy water quality indices, and human health risk evaluation—have quantified the latent ecological threats posed by heavy metals [24,25]. Collectively, these studies provide a theoretical foundation for environmental management in reservoir areas.

Notably, nearly half of China’s priority control zones for heavy metals reside within the Yangtze River Economic Belt, according to environmental protection plan covering the Yangtze River Economic Belt. In 2020, the Yangtze River Protection Law was formally enacted [26,27,28], underscoring the critical importance and urgency of basin-wide ecological conservation and restoration, encompassing the Three Gorges Reservoir [28]. However, a critical review of recent literature reveals a scarcity of comprehensive ecological risk assessments for the Three Gorges Reservoir, particularly studies providing comparative data before and after the Yangtze River Protection Law was enacted, which precludes a robust evaluation of the law’s efficacy and a mechanistic understanding of recent pollution dynamics.

To precisely quantify contamination and ecological risks of heavy metals in Three Gorges Reservoir sediments by (1) analyzing their spatial distribution characteristics, (2) apportioning their potential sources, and (3) conducting a comprehensive ecological risk assessment, this study was initiated during the inaugural year of the Yangtze River Protection Law’s enactment. Utilizing nine-metal (Cd, Cu, Pb, Fe, Mn, Cr, As, Hg, Zn) concentration data from shallow surficial sediments (0–65 cm depth) within the Air–Space-Ground integrated project conducted by Changjiang Water Resources Commission, spatial distribution characteristics of sediment-bound metals were analyzed, and principal component analysis (PCA) and correlation analysis were employed for preliminary source apportionment, while geo-accumulation index (I_geo), potential ecological risk index (RI), and sediment quality guidelines (SQGs) assessments were applied to quantified ecological risks. These integrated approaches provide reference frameworks and empirical support for sediment heavy-metal pollution mitigation, establishing scientific foundations for precision pollution control, formulation of ecological restoration strategies, and sustainable environmental management under long-term operational safety of the Three Gorges Project.

2. Materials and Methods

2.1. Study Area

As the largest artificial reservoir in the world [29], the Three Gorges Reservoir extends 660 km longitudinally across 26 counties/cities in Chongqing and Hubei Province [30,31], approximately between 29°16′ N to 31°25′ N, 106°50′ E to 110°51′ E. The reservoir area encompasses approximately 57,600 km², and experiences a subtropical monsoon climate with abundant precipitation [32,33]. Conducive thermal-humid conditions foster prolific algal and aquatic vegetation growth. Since impoundment, hydrological alterations—extended hydraulic retention and attenuated flow velocities—have thereby elevated eutrophication risk [34], while the water quality directly governs aquatic environmental security of the middle-lower Yangtze River reaches [18]. Functioning as a distinctive semi-natural–semi-artificial water body, the Three Gorges Reservoir differs markedly from natural lakes and marine systems in water depth, hydrological characteristics, and ecosystem structure, serving as a critical platform for studying eco-environmental impacts of large-scale hydraulic engineering [35].

Field sampling for this investigation was conducted during autumn 2020, focused on the Yangtze mainstem and tributaries from Badong County to the Three Gorges Dam. Twelve monitoring stations were established at the following locations: Miao River (MR), Xiangxi River (XR), Zhaxi River (ZR), Quyuan Town (QT), Xietan (XT), Shennong Stream (SS), Puzhuang River (PR), Badong (BD), Fanjiaping (FP), Guandukou (GK), Guizhou (GZ), and Qinggang River (QR). The site configurations are detailed in Figure 1.

2.2. Sample Collection and Analytical Procedures

The selection of the 12 representative sampling sites was informed by the ‘Air-Space-Ground integrated monitoring’ program conducted by the Changjiang Water Resources Commission in 2020. This program employed a multi-platform approach, encompassing satellite remote sensing, unmanned aerial vehicle photography, and in situ multi-spectral water quality monitoring buoys, to establish a three-dimensional environmental monitoring network across the Three Gorges Reservoir. Concurrently with sediment sampling, surface water samples were collected at each site for laboratory analysis of parameters including temperature (Temp.), electrical conductivity (EC), dissolved oxygen (DO), pH, turbidity (Turb.), total organic carbon (TOC), total nitrogen (TN), and total phosphorus (TP). These laboratory measurements were used to calibrate and validate the data obtained from the in-situ sensors, ensuring data accuracy and reliability for subsequent correlation analysis with sediment properties.

Sediment cores were collected at each transect using a column sample deployed via winch from the hydrological survey vessel. At each sampling transect, 65-cm-deep sediment cores were retrieved, with stratified subsamples collected at 5 cm intervals. Samples were sealed in polyethylene sample bags to exclude air and stored at 4 °C immediately upon collection.

Collected sediments were air dried, subsequently pulverized, sieved through a 100-mesh nylon sieve, and stored in polyethylene bags. Prior to analysis, the samples were microwave digested. Heavy metal concentrations were quantified using inductively coupled plasma optical emission spectrometry (iCAP6300 Duo, Thermo Fisher Scientific, Waltham, MA, USA) and atomic fluorescence spectrometry (AFS-930, Jitian Instrument Co., Beijing, China). The entire process, from sampling preservation, transportation to laboratory testing, strictly implemented the quality control measures stipulated in the Specifications for Water Environmental Monitoring (SL 219-2013) [36] to ensure data quality and reliability. Corresponding detection methodologies and limits are detailed in

Table 1. Analytical methods and standards for target heavy metals.

Methods and Standards	Analytes Determination	Detection Limits (mg·kg−1)
Soil and sediment—Determination of aqua regia extracts of 12 metal elements-Inductively coupled plasma mass spectrometry (HJ 803-2016) [37]	Cd	0.09
	Cu	0.6
	Pb	2
	Fe	68 [38]
	Mn	0.4
	Cr	2
	Zn	1
Soil and sediment—Determination of mercury, arsenic, selenium, bismuth, antimony-Microwave dissolution/Atomic Fluorescence Spectrometry (HJ 680-2013) [39]	As	0.01
	Hg	0.002

.

2.3. Heavy Metal Contamination Assessment Methods

2.3.1. Geo-Accumulation Index (I_geo)

The Geo-accumulation Index, initially proposed by the Müller in 1969 [19], has been extensively employed for assessing heavy-metal contamination in sediments [40,41,42,43,44]. In this study, background values of heavy metals in Three Gorges Reservoir soils served as the reference values [45] (

Table 2. Background Values of Soil Heavy Metals in the Three Gorges Reservoir.

Heavy Metals	Cd	Cu	Pb	Cr	As	Hg	Zn
Background Values (mg·kg−1)	0.134	25.00	23.88	78.03	5.835	0.046	69.88

). The I_geo is calculated using the Formula (1), where C_i is the measured concentration of the target metal, C_B is the reference value, and K is a correction factor (K = 1.5). The resulting I_geo values were classified into seven contamination grades, corresponding to uncontaminated to extremely contaminated (

Table 3. Classification of pollution level based on I_geo.

Igeo	Classification
(−∞, 0]	Uncontaminated
(0, 1]	Uncontaminated to moderately contaminated
(1, 2]	Moderately contaminated
(2, 3]	Moderately to heavily contaminated
(3, 4]	Heavily contaminated
(4, 5]	Heavily to extremely contaminated
(5, +∞)	Extremely contaminated

).

$${\mathit{I}}_{\mathit{g}\mathit{e}\mathit{o}}={\mathit{l}\mathit{o}\mathit{g}}_2{\displaystyle \frac{{\mathit{C}}_{\mathit{i}}}{{\mathit{K}\times \mathit{C}}_{\mathit{B}}}}$$

(1)

2.3.2. Potential Ecological Risk Index (RI)

The Potential Ecological Risk Index (RI) proposed by Håkanson is extensively employed to quantify the composite ecological risk of heavy metals in aquatic ecosystems [20,46,47,48,49]. The calculation formula refers to Formula (2), where E_i is the potential ecological risk index for a specific metal, T_i is the response toxicity coefficient (

Table 4. Toxicity Coefficients, and Ecotoxicity Thresholds of Soil Heavy Metals.

Heavy Metals	Toxicity Coefficients (mg·kg−1)	TEL (mg·kg−1)	PEL (mg·kg−1)
Cd	30	0.68	4.21
Cu	5	18.7	108
Pb	5	30.2	112
Cr	2	52.3	160
As	10	7.24	41.6
Hg	40	0.13	0.7
Zn	1	124	271

), C_i is the measured concentration of the target metal, and C_B is the reference value required for calculation (Table 2).

$${\mathit{E}}_{\mathit{i}}={\mathit{T}}_{\mathit{i}}\times {\displaystyle \frac{{\mathit{C}}_{\mathit{i}}}{{\mathit{C}}_{\mathit{B}}}}$$

(2)

$$\mathit{R}\mathit{I}=\mathsf{\sum }_{\mathit{i}=1}^{\mathit{n}}{\mathit{E}}_{\mathit{i}}$$

(3)

The Risk Index (RI) quantifies the cumulative potential ecological risk of heavy metals, calculated as the sum of individual ecological risk factors (Formula (3)). Following Håkanson’s methodology [20], the cumulative toxicity coefficient for the eight target metals in this study was 93. Consequently, the threshold value for the low ecological risk category was determined to be 100 following rounding, with subsequent risk level threshold doubling relative to the preceding category incrementally. Corresponding E_i and RI values are presented in

Table 5. Classification of pollution level based on potential ecological risk.

Ei Range	RI Range	Classification
[0, 40]	[0, 100]	Low ecological risk
(40, 80]	(100, 200]	Moderate ecological risk
(80, 160]	(200, 400]	Considerable ecological risk
(160, 320]	(400, 800]	High ecological risk
(320, +∞)	(800, +∞)	Very High ecological risk

.

2.3.3. Sediment Quality Guidelines (SQGs)

Sediment quality guidelines (SQGs) enable rapid identification of priority areas through potential acute toxicity assessment [50,51]. When contaminant concentrations are below Threshold Effect Level (TEL), they are considered unlikely to have adverse effects on sediment, and when concentrations exceed Probable Effect Level (PEL), adverse biological effects may occur frequently [52,53]. Concentration values that lie between the TEL and the PEL indicate that sediment organisms suffer from occasional adverse effects [52,53]. The PEL serves as the threshold for high-risk contamination requiring urgent remediation [54]. Corresponding TEL and PEL values are provided in Table 4. The sum of the toxic units (∑TUs) could be used to estimate the potential acute toxicity of the contaminants in a sample (Formula (4)), where TUs is the toxic unit for individual metals, C_i is the measured heavy-metal concentration, and PEL_i is the probable effect concentration. Heavy-metal toxicity classification criteria are detailed in

Table 6. Classification of pollution level based on sediment quality guidelines.

∑TUs	Classification
(−∞, 4]	Uncontaminated
(4, 6]	Moderately Contaminated
(6, +∞)	Severely Contaminated

.

$$\mathsf{\sum }\mathbf{TUs}=\mathsf{\sum }_{\mathit{i}=1}^{\mathit{n}}{\displaystyle \frac{{\mathit{C}}_{\mathit{i}}}{{\mathit{P}\mathit{E}\mathit{L}}_{\mathit{i}}}}$$

(4)

2.4. Sample Analysis and Data Processing

Descriptive statistical analyses of sediment heavy metal concentrations and mass fractions were conducted using average, max, min function in Microsoft Excel 365. Graphical plotting was performed using OriginPro 2024. Multivariate statistical analyses were performed in RStudio 4.3.2. Specifically, correlation analysis was conducted using the cor and pheatmap package, and principal component analysis (PCA) was performed using the pca package after Z-score normalized. Spatial distributions were visualized using ArcGIS 10.8.

3. Results

3.1. Hydrochemical Characteristics and Spatial Distribution of Heavy Metals in the Study Area

3.1.1. Hydrochemical Characteristics of Study Area

Water physicochemical parameters in the study area are summarized in

Table 7. Physicochemical parameters of water samples at sampling sites.

Sampling Site	Temp. (°C)	EC (μS·cm−1)	DO (mg·L−1)	pH	Turb. (NTU)	TOC (mg·L−1)	TN (mg·L−1)	TP (mg·L−1)
MR	24.5	334.7	7.41	7.8	110	2.69	1.95	0.10
XR	24.8	327	7.24	8.14	10.3	2.94	1.57	0.05
ZR	24.5	332.9	7.89	7.9	16	3.02	1.56	0.07
QT	24.5	322.6	7.39	7.75	43.2	2.46	1.84	0.09
XT	23.8	343.9	7.07	7.8	112.9	2.36	1.55	0.12
SS	23.8	341.3	5.96	7.79	73.2	2.68	1.78	0.11
PR	25	326.3	7.52	8.19	16	3.20	1.86	0.07
BD	23.8	338.6	6.1	7.78	96	2.38	1.59	0.11
FP	23.8	343.9	7.78	7.69	76.2	2.46	1.63	0.08
GK	24	340.4	6.17	7.82	10.9	2.36	1.71	0.09
GZ	24	340.3	6.09	7.8	87.7	2.60	1.75	0.09
QR	25	307.3	8.27	8.84	16	3.34	1.37	0.07

. Monitoring data from 12 sites indicate consistently neutral to weakly alkaline conditions, with pH ranging from 7.69 at FP to 8.84 at QR, a median of 7.8, and SD of 0.31. Limited spatial variability suggests dominance of natural buffering processes, with no localized acidification. EC ranges from 307.3 to 343.9 μS·cm⁻¹ (mean: 333.7 μS·cm⁻¹), reflecting moderate ionic concentrations with minimal longitudinal variation. No significant downstream conductivity decrease was observed, indicating negligible dilution effects. These patterns imply hydrogeochemical processes governed by carbonate weathering [55], with ionic composition controlled by lithological and climatic factors.

DO levels were generally favorable (5.96–8.27 mg·L⁻¹), though slightly reduced at SS (5.96 mg·L⁻¹), BD (6.10 mg·L⁻¹), and GK (6.17 mg·L⁻¹), potentially associated with localized increases in water depth or accumulation of oxygen-consuming organic matter. Significant Turb. heterogeneity was observed: MR (110 NTU), XT (112.9 NTU), and GZ (87.7 NTU) constituted high-turbidity zones, likely influenced by nearshore sediment input or runoff disturbances. Conversely, lower turbidity at XR (10.3 NTU) and ZR (16 NTU) indicated superior water clarity. Nutrient concentrations remained stable for TOC (2.36–3.34 mg·L⁻¹), TN (1.37–1.95 mg·L⁻¹), and TP (0.05–0.12 mg·L⁻¹). QR exhibited distinctively low TN (1.37 mg·L⁻¹) and TP (0.07 mg·L⁻¹), potentially due to isolation from agricultural non-point source pollution. Notably, elevated turbidity sites such as MR and XT showed no anomalous TOC/TN ratios, suggesting that turbidity predominantly originated from inorganic suspended particles rather than organic detritus.

Based on the above analysis, the physicochemical characteristics of water bodies in the study area demonstrate pronounced homogeneity, featuring weak alkalinity and moderate ionic concentrations characteristic of natural river systems. Limited indications of mining activities or industrial point-source pollution was detected. Water quality variations primarily arise from sediment inputs and localized oxygen-consuming processes. Although limited non-point source pollution exists within the watershed, anthropogenic inputs remain insufficient to trigger eutrophication.

3.1.2. Statistical Characteristics of Sedimentary Metal Concentrations

Statistical analysis of eight heavy metal concentrations in 0–65 cm sediment horizons across 12 sampling sites is presented in

Table 8. Statistical Characteristics of Sediment Heavy-Metal Concentrations.

Heavy Metal	Min (mg·kg−1)	Max (mg·kg−1)	Median (mg·kg−1)	Mean (mg·kg−1)	SD (mg·kg−1)	CV (%)
Cd	0.18	5.76	0.74	1.33	1.55	116.76
Cu	10.29	242.78	78.12	78.78	35.65	45.25
Pb	8.14	61.01	30.66	29.19	12.23	41.90
Fe	5747.60	32,596.11	20,116.48	20,362.06	5686.36	27.93
Mn	162.11	1209.76	754.03	751.60	234.43	31.19
Cr	16.56	98.48	60.36	60.49	17.01	28.12
As	0.96	7.30	3.82	3.98	1.25	31.35
Hg	<0.002	0.26	<0.002	0.03	0.06	214.46
Zn	30.72	279.56	174.19	163.82	52.91	32.17

. The results indicate that Cd and Hg exhibit the most pronounced anthropogenic contamination. Cd concentrations ranged from 0.18 to 5.76 mg·kg⁻¹, with a max-min ratio of 32.0 and coefficient of variation (CV) of 116.76%. Hg concentrations varied between 0 and 0.26 mg·kg⁻¹, demonstrating the highest spatial heterogeneity (CV = 214.46%) across the study area, suggesting intensive point-source inputs. Significant spatial fluctuations were observed for Cu, Pb, and As, with CV values of 45.25% and 41.90%, respectively, indicating moderate anthropogenic perturbation. Fe, Cr, Mn, and Zn exhibited relative homogeneity (CV ≈ 30%), implying predominant natural origins. The mean Fe concentration (20,362.06 ± 5686.36 mg·kg⁻¹) aligns with fluvial sediment geochemical backgrounds. Mn (751.60 ± 234.43 mg·kg⁻¹) and Cr (60.49 ± 17.01 mg·kg⁻¹) concentrations corresponded to regional sedimentary rock weathering baselines [56], showing no marked anthropogenic enrichment.

3.1.3. Vertical Distribution Patterns of Heavy Metals in Sediments

To characterize depth-dependent distribution patterns of nine heavy metals, sediment profiles (0–65 cm) were stratified into the following three horizons: surface (0–20 cm), intermediate (25–40 cm), and deep layers (45–65 cm) [57]. Concentration frequency distributions across strata were statistically analyzed and visualized via violin plots (Figure 2). Results revealed that mean concentrations of Cd and Hg substantially exceeded their interquartile ranges (IQR). Outliers (defined as >Q3 + 1.5IQR or <Q1 − 1.5IQR) constituted 13.9% and 17.5% of observations, respectively, exclusively above the IQR, exhibiting markedly right-skewness indicative of frequent extreme values from point-source pollution. Conversely, Pb, Fe, Mn, Cr, and As displayed quasi-normal distributions characterized by high congruence between medians and means, with outlier proportions <1%, reflecting stable diffusion governed by geological processes.

Vertical maximum-concentration frequencies quantified aggregation effects and historical contamination. As in Figure 3, Vertical profiling revealed the maxima occurred predominantly in surface sediments for Cu, Pb, Fe, Mn, and Zn, suggesting recent anthropogenic influence or inherent surface enrichment. Conversely, Cd exhibited maximal peak deep-layer accumulation frequency, suggesting historical contamination or downward migration via soluble complexation and porewater transport, demonstrating pronounced vertical accumulation effects, which were consistent with a recent study [58]. Spatially, MR and GK showed intense surface enrichment, signifying active contemporary inputs. XR and ZR exhibited marked deep-layer accumulation, potentially influenced by legacy pollution or deep groundwater infiltration. Given differential ecological risks and toxicological effects among contaminants, further investigation integrating source apportionment and ecological risk assessment is warranted.

3.2. Source Apportionment of Heavy Metals in Sediments

3.2.1. Correlation Analysis Between Sediment Metals and Water Physicochemical Properties

Correlation analysis between sediment heavy metals and physicochemical factors (Figure 4) revealed similar environmental response for Fe, Mn, and Zn. These elements showed significant positive correlations with DO and Temp., but negative correlations with EC. This pattern indicates the following redox conditions as a primary control mechanism: elevated DO promotes precipitation of Fe/Mn/Zn oxides or insoluble salts, while increased EC potentially inhibits dissolved-phase migration through competitive ion adsorption or soluble complex formation. TOC correlated positively exclusively with Fe, Mn, and Zn. Conversely, Cd and Hg exhibited inverse Temp. dependencies, implying that warming potentially enhances their release via accelerated desorption kinetics or organic matter mineralization. The positive Cd-EC correlation reflects saline inputs from sources like industrial effluents [59]. Most metals (excluding Cd, Hg, and Zn) correlated negative with turbidity, indicating adsorption onto suspended particulates in turbid waters [60]. Weak correlations for Cu, Pb, Cr, and As with physicochemical parameters suggest governance by unmonitored factors, requiring further investigation through isotopic tracing or speciation analysis. No significant TN/TP-metal correlations were observed, implying limited nutrient cycling influence on metal transport.

3.2.2. Inter-Element Correlation Analysis

Inter-element correlation analysis of sediment heavy metals (Figure 5) revealed distinct source groups and coupled environmental behaviors. A robust positive correlation (r = 0.95) between Cd and Hg indicated high homology from anthropogenic sources such as industrial metallurgy or e-waste dismantling [59,61]. Strong geochemical paragenesis among Fe, Mn, Cr, and Zn reflected migration governed by coprecipitation with Fe-Mn oxides and surface adsorption, confirming dominant natural diagenetic contributions [62,63]. Significant Cu-As covariation suggested mining or industrial inputs. Notably, weak negative correlations between Cd/Hg and Fe/Mn pairs revealed antagonistic effects between anthropogenic contamination and natural geochemical processes. Minimal correlations among other elements demonstrated clear source differentiation without substantial cross-influence mechanisms.

3.2.3. Principal Component Analysis and Source Identification

Principal Component Analysis (PCA) was employed to identify potential heavy metal sources in sediments [46], with loadings detailed in

Table 9. Principal component analysis results (rotated component matrix and total variance interpretation).

Parameter	PC1	PC2	Contribution
Cd	−0.095	0.929	33
Cu	0.672	0.12	11.3
Pb	0.14	0.791	24.2
Fe	0.929	−0.037	20.6
Mn	0.9	0.017	19.2
Cr	0.792	0.421	21.6
As	0.77	−0.299	17.5
Hg	−0.232	0.927	33.9
Zn	0.885	0.031	18.6
Eigenvalues	4.21	2.63
Percentage of Variances	46.76	29.24
Cumulative Variance	76.00

and Figure 6. The suitability of the data for PCA was statistically validated. The Kaiser-Meyer-Olkin measure yielded a value of 0.708, indicating adequate sampling adequacy, and Bartlett’s test of sphericity was highly significant (p = 5.06 × 10⁻¹⁴⁵), confirming sufficient correlation among variables. The analysis was performed on standardized data to account for scale differences. Two principal components (PC1 = 4.21, PC2 = 2.63) with eigenvalues >1 cumulatively explained 76% of variance (PC1 = 46.76%, PC2 = 29.24%), indicating robust source identification. Source apportionment revealed the following: PC1 was dominated by Fe (0.929), Mn (0.90), Zn (0.885), Cr (0.792), As (0.77), and Cu (0.672). The Fe/Mn loading ratio (1.03) approached natural sedimentary values, indicating control by Fe-Mn oxide coprecipitation—a signature process of natural weathering governed by geological backgrounds [62,63]. High As loading (0.77) suggested adsorption coupling with iron oxides. PC2 exhibited dominance by Cd (0.929), Hg (0.927), and Pb (0.791). The minimal Cd-Hg loading differential (0.002) confirmed homologous origins from anthropogenic sources like electronic industry emissions.

3.3. Ecological Risk Assessment of Heavy Metal in Sediments

3.3.1. Geo-Accumulation Index (I_geo) Assessment

Analysis of I_geo values for seven heavy metals (Cd, Cu, Pb, Cr, As, Hg, Zn) across 12 sampling sites revealed significant spatial heterogeneity (Figure 7 and Figure 8). Cd exhibited the most severe pollution, with nearly 92% of samples reaching moderate contamination or higher, demonstrating basin-wide pollution. Excluding GZ (mean I_geo = 0.50), all sites showed mean values corresponding to moderate contamination. Peak values were observed at FP (4.71 ± 0.09) and SS (3.80 ± 1.18), classified as heavily-to-extremely contaminated and heavily contaminated, respectively. Hg primarily reflected natural background levels due to frequent non-detects. However, consecutive I_geo > 1.2 values at 15–25 cm in SS and I_geo = 1.91 at 10 cm in FP indicated extreme spatial heterogeneity, which is consistent with buried contaminants from historical point sources. Mean I_geo ranges for Cu (−1.15 to 1.81) and Zn (−1.06 to 1.05) indicated predominantly low contamination. Pb showed 36.5% low contamination data with remaining values uncontaminated, while Cr and As consistently exhibited I_geo < 0, confirming natural origins. Spatially, uncontaminated data were clustered in GZ, reflecting minimal anthropogenic disturbance. Conversely, SS and FP exhibited significantly elevated Cd and Hg contamination, requiring priority remediation. Vertically, Cd contamination generally intensified with depth, indicating cumulative distribution in sediment columns. Other metals showed lower vertical vertical variability in I_geo, suggesting homogeneous natural sourcing.

3.3.2. Potential Ecological Risk Index (RI) Assessment

Comprehensive E_i data across 12 sediment sampling sites (Figure 9 and

Table 10. Statistical results of E_i of each heavy metal and RI of each site.

Range	Percentage of Data for Ei (%)							Percentage of Datas for RI (%)
	Cd	Cu	Pb	Cr	As	Hg	Zn
Ei < 40	0	98.54	100	100	100	81.02	100	RI < 100	8.03
40 ≤ Ei < 80	7.30	1.46	0	0	0	4.38	0	100 ≤ RI < 200	37.96
80 ≤ Ei < 160	37.96	0	0	0	0	13.14	0	200 ≤ RI < 400	39.42
160 ≤ Ei < 320	40.15	0	0	0	0	1.46	0	400 ≤ RI < 800	0
Ei ≥ 320	14.60	0	0	0	0	0	0	RI ≥ 800	14.60

) revealed extreme heterogeneity in RI value (range: 0.00773–1289.55; CV = 217%), indicating pronounced spatial contamination disparities. The risk contribution hierarchy was as follows: Cd (mean 337.6) ≫ Hg (mean 24.3) > As (mean 6.8) > Cu (mean 6.2) > Pb (mean 5.9) > Zn (mean 2.4) > Cr (mean 1.6). Risk gradation distribution indicated the following: low ecological risk constituted 61.4% (94/153) of data, predominantly from Cu, Pb, Cr, Zn; moderate ecological risk accounted for 10.5% (16/153), primarily As and Hg; considerable ecological risk comprised 4.6% (7/153), concentrated in As and Hg; very high ecological risk represented 23.5% (36/153), exclusively contributed by Cd and Hg (Cd accounting for 97.2%).

Cd served as the primary basin-wide risk driver, with contamination detected across all sites. Overall, 98.7% of sampling locations exceeded considerable risk levels (RI > 80), and 45.1% reached very ecological high risk. FP (mean RI = 1187.2) exhibited higher contamination than SS (mean RI = 928.6), reflecting both extensive distribution and localized anthropogenic inputs. Hg displayed the most pronounced point-source characteristics; only 17.6% of samples showed ecological risk, yet 59.3% (16/27) of these exceeded considerable or higher ecological risk, exemplified by RI = 147.8 at 20-cm depth in SS and RI = 226.1 at 30-cm in FP. Spatially, elevated Cd and Hg concentrations coincided with historical industrial zones and navigation hubs, indicating point-source inputs from industrial emissions, coal combustion, or Hg-containing waste disposal. Cu presented a single moderate ecological risk point at 5–10 cm depth in BD. Pb, Cr, As, and Zn consistently demonstrated low ecological risk across all samples.

Vertical distributions showed increasing RI values of Cd with depth in surface sediments at SS and FP, persistently exceeding 900 within the 20–55 cm stratum. Similarly, positive RI-depth correlations occurred for Cd at XR and XT, indicating historical industrial emissions or burial diagenesis as contamination sources. Conversely, Cu contamination exhibited a decreasing trend with depth at MR, QT, PR, BD, and GZ; Pb showed analogous declines at MR, ZR, XT, BD, and QR; Zn followed this vertical attenuation pattern at MR, QT, BD, GK, GZ, and QR. This consistent surface enrichment implicates atmospheric deposition and surface runoff as primary input pathways.

Spatial distribution of comprehensive potential ecological risk revealed that 10 of 11 depth-stratified samples at GZ exhibited RI < 100, whereas SS and FP demonstrated elevated risks with localized high ecological risk zones. These areas constitute both pollution accumulation hotspots and focal points for Cd-associated ecological risks, demonstrating anthropogenic activities as the dominant cause of Cd enrichment, with significant contributions from the geological background as well.

3.3.3. Sediment Quality Guidelines (SQGs) Assessment

Assessment of vertical mean concentrations using sediment quality guidelines methodology (Figure 10) indicated generally low biological toxicity effects across 12 sampling sites. Mean toxic units (TUs) decreased in the following order: Cu (0.74) > Zn (0.60) > Cr (0.38) > Cd (0.31) > Pb (0.26) > As (0.10) > Hg (0.04). Comparative analysis with TEL revealed Cd concentration at FP, and Cu concentration at BD reached 1.25 times and 1.30 times TEL. Both metals fell within the TEL-PEL risk range, suggesting potential occasional toxic effects. Concentrations of Cd and Cu at all other sites remained below TEL, indicating no significant risk. Pb, Cr, As, Hg, and Zn consistently remained below TEL across all sites, demonstrating the unlikelihood of adverse effects. Among heavy metals, Cd contributed minimally to the TUs across sampling sites, despite elevated I_geo value. The higher PEL value than background value of Cd would lead to underestimation of its toxicity in TU assessment. Thus, an integrated framework should be established to accurately and comprehensively evaluate ecosystem risks from heavy metals.

Comprehensive biological toxicity assessment through summation of ΣTUs (Figure 11) revealed values ranging from 0.83 to 3.76. All sites exhibited ΣTUs < 4, classified as uncontaminated. GZ (ΣTUs = 0.83) represented the cleanest site within the study area, with all elements <0.2 × TEL. Although FP reached the maximum ΣTUs (3.76), natural geochemical buffering prevented synergistic effects from exceeding critical thresholds, maintaining low biological toxicity. In summary, localized Cd and Cu exceedances in sediments induced no systemic toxic effects. However, specific ecological risks at FP and SS necessitate prioritized monitoring.

4. Discussion

This study systematically investigated heavy-metal contamination characteristics and ecological risks in sediments of the Three Gorges Reservoir. Spatial and vertical distribution analyses revealed significant heterogeneity in metal concentrations, with CV exceeding 50% only for Cd (116.76%) and Hg (214.46%). Cd and Hg exhibited distinctly right-skewed distributions, reflecting frequent extreme values from point-source pollution. Surface enrichment was revealed for Cu and Pb, whereas Cd displayed significant cumulative effects along vertical profiles within specific depth interval, which is consistent with observations from Lake Jangseong, South Korea [64]. Intensive surface accumulation was identified at MR and GK, while deep-layer enrichment dominated at XR and ZR, reflecting the temporal differences in pollution sources.

The distinct correlation patterns reveal fundamentally different biogeochemical behaviors between the metal groups. Fe/Mn/Zn correlations with TOC/DO suggest redox-driven diagenesis [65], likely through Fe-Mn oxide coprecipitation [66]. Their negative EC correlation indicates reduced solubility at high ionic strength. Conversely, Cd and Hg’s positive EC correlation suggests formation of soluble complexes in high-ion environments, while their negative temperature dependence may indicate temperature-dependent adsorption processes. These contrasting behaviors highlight how different environmental parameters selectively control metal fate and transport in reservoir systems.

Integration of correlation analysis and PCA source apportionment revealed geogenic processes as the predominant origin for Fe, Mn, and Zn. Conversely, for Cd and Hg, the minimal loading differentials (0.002) in PCA, coupled with their elevated coefficients of variation (116.76% and 214.46%, respectively) and marked spatial heterogeneity, strongly support anthropogenic origins rather than natural weathering processes. The identified anthropogenic signatures align with regional industrial activities, specifically suggesting contributions from electronic manufacturing emissions, coal combustion, and mining operations.

Ecological risk assessment revealed the following I_geo accumulation hierarchy: Cd > Cu > Zn > Pb > Hg > Cr > As. Approximately 92% of Cd data indicated moderate or higher contamination, with FP and SS exhibiting heavily-to-extremely contaminated and heavily contamination, while GZ showed minimal anthropogenic disturbance. Parallel findings were revealed by the potential ecological risk index methodology as follows: Cd exceeded considerable risk levels at 98.7% of sampling locations, with the maximum at FP and SS, demonstrating pronounced spatial heterogeneity and point-source characteristics. Sediment quality guidelines assessment indicated localized risks of occasional toxic effects for Cd and Cu, though systemic biological toxicity effects were absent.

Our findings regarding Cd as the primary ecological risk driver are consistent with a growing body of global research on large reservoir sediments. For instance, studies on the Halda river in Bangladesh [67] and a review in India [68] have similarly identified Cd, often derived from industrial and mining activities, as the metal of utmost concern, with concentrations and risk indices closely aligned with those reported here. This consensus underscores Cd’s persistent bioavailability and potent toxicity in aquatic sediments worldwide. Cd poses long-term risks through sustained release at sediment-water interfaces, whereas Hg’s bio-accumulative potential and methylated toxicity facilitate food-chain magnification, threatening aquatic ecosystems and human health [69]. Considering the strong correlation between Cd and Hg and the characteristics of anthropogenic sources, this necessitates stringent controls on industrial wastewater discharge, coal combustion emissions, and mining effluents to mitigate Cd and Hg contamination.

A methodological consideration merits attention in our risk assessment. The use of regional soil background values, while a practice adopted by other studies in similar contexts [70,71,72], may introduce uncertainty when applied to aquatic sediments. To enhance the accuracy and regional specificity of future ecological risk evaluations, some researchers have proposed the use of local geochemical background values combined with the local enrichment factor method, utilizing Fe-normalized element concentrations as a global reference to achieve dual normalization and improve assessment accuracy [73]. However, this method depends on rigorous field sampling and the integration of multidisciplinary expertise in fluvial geomorphology, sedimentology, and geochemistry [73].

Lastly, while this study focused on spatial and vertical distribution, the question of lateral heterogeneity across the width of the reservoir remains an important aspect of the contaminant’s three-dimensional distribution. The concentration may differ substantially from mid-channel to bank areas due to variations in flow velocity, sediment composition, and proximity to pollution sources [74,75]. Additionally, numerous studies have highlighted the significant temporal variations across years [76,77]. Future studies should therefore combine lateral mapping with temporal monitoring to better understand how hydrodynamics, seasonal changes, and human activities jointly influence metal distribution in large reservoirs.

5. Conclusions

In summary, among nine heavy metals investigated, Cd exhibited basin-wide contamination, exceeding environmental benchmarks across all sampling sites. Both Cd and Hg demonstrated distinct point-source characteristics, indicating predominant anthropogenic origins linked to industrial activities. GZ maintained superior environmental integrity, while FP and SS presented significant ecological risks. Future environmental management requires targeted and efficient interventions in these high-risk zones to mitigate ecological threats from Cd and Hg, ensuring long-term regional ecological security.