Hydrochemical Characteristics and Dominant Controlling Factors of the Qujiang River Under Dual Natural–Anthropogenic Influences

Abstract

This study investigated the hydrochemical characteristics, solute sources, and controlling factors of the Qujiang River, a vital tributary of the Qiantangjiang River, on the basis of 61 surface water samples collected from July to September 2019. A multi-method framework integrating ArcGIS spatial analysis, a Piper trilinear diagram, a Gibbs diagram, ion ratio analysis, and principal component analysis (PCA) was systematically applied. The results demonstrate significant spatial heterogeneity in pH values (6.39–7.15 in the upper reaches, 6.31–8.83 in the middle reaches, and 6.85–8.1 in the lower reaches), with hydrochemical facies dominated by Ca-HCO₃ types (87% frequency) in the upper–middle reaches, transitioning to SO₄·Cl-Na and SO₄·Cl-Na·Ca mixed types downstream. Source apportionment indicates that carbonate weathering and atmospheric precipitation are the primary controls on hydrochemistry in the upper and middle reaches, whereas industrial effluents, evidenced by a 2.3-fold increase in SO₄²⁻ concentration, and domestic sewage, with Cl⁻ levels reaching 16.07 mg/L, are the dominant influences in the lower reaches. This study’s innovation lies in the quantitative separation of natural weathering (68% contribution) and anthropogenic activities (32%) through an integrated methodological approach, offering a basin-scale understanding of longitudinal hydrochemical evolution. These results provide valuable insights for the management and ecological conservation of medium- to small-sized basins.

1. Introduction

Rivers act as vital linkages between terrestrial ecosystems and marine environments, playing an indispensable role in sustaining ecological balance, ensuring water resource availability, and supporting regional economic development. The hydrochemical composition of rivers contains crucial information reflecting both natural geomorphic processes, such as rock weathering, atmospheric precipitation, and evaporative concentration, as well as anthropogenic disturbances, including industrial discharges, agricultural irrigation, and domestic wastewater [1,2,3,4]. Consequently, understanding the spatiotemporal evolution of river hydrochemistry and its controlling factors is fundamental for assessing the health of aquatic ecosystems and promoting sustainable water resource management.

Global-scale studies on large rivers have established systematic frameworks. For example, the Amazon River has high dissolved organic carbon (DOC) and low ionic concentrations due to intense weathering and biogeochemical cycling [5], whereas the Nile River has elevated salinity due to Aswan Dam regulation and agricultural practices [6,7]. In China, research on the Yangtze and Yellow Rivers reveals codominant controls by carbonate weathering (contributing 65% of the Ca²⁺ in the Yangtze River Basin) and anthropogenic inputs (e.g., 3.44 mg/L NO₃⁻ from fertilizer application in the Yellow River Basin) [8,9,10]. However, research on medium- and small-sized basins remains fragmented and is often confined to localized case studies. Notable instances include sulfate enrichment (18.3 mg/L SO₄²⁻) in the mid-reaches of the Qiantang River caused by textile effluents [11] and chloride accumulation (24.8 mg/L Cl⁻) in the Min River estuary resulting from seawater intrusion [12]. Existing studies primarily concentrate on individual pollution sources, such as agricultural non-point or industrial point sources, with limited emphasis on basin-wide, multi-factor interactions. Against this backdrop, this study conducts a comprehensive investigation of the Qujiang River Basin.

The Qujiang River, a significant tributary of the Qiantangjiang River with a basin area of 11,138 km², originates from Huangshan Mountain in Anhui Province and flows through the Jinqu Basin before merging into the Qiantang River, serving as the principal water source for Zhejiang’s “Jinqu-Li Ecological Economic Zone” [13]. However, rapid industrialization and intensive agricultural activities have led to progressive deterioration of water quality [14]. Previous research has primarily focused on hydrological analysis [15,16], with limited attention given to pollution dynamics. Three major knowledge gaps remain: (1) the spatial differentiation of hydrochemical facies across the entire basin is not well understood; (2) the respective contributions of natural weathering processes, such as red-bed clastic dissolution, and anthropogenic influences, including chemical wastewater discharge, have not been quantitatively assessed; and (3) effective methodologies for distinguishing the combined effects of industrial, agricultural, and domestic pollution are lacking.

This study addresses these gaps through a comprehensive investigation of the Qujiang River. We systematically analyze spatial hydrochemical variations through multiple methods: visualize the spatial distribution of ion concentrations by using the Inverse Distance Weighting (IDW) method in ArcGIS; interpret hydrochemical types with Piper trilinear diagrams; and comprehensively analyze the sources of ion components by means of Gibbs diagrams, ion ratio methods, and principal component analysis (PCA). By systematically analyzing spatial hydrochemical variations and employing multi-method diagnostics (Gibbs diagrams, ionic ratios, and principal component analysis), we aim to unravel source-to-sink dynamics and elucidate natural–anthropogenic interaction mechanisms. The results provide scientific foundations for water resource allocation, pollution control strategies, and ecological restoration in the Qujiang Basin while establishing a transferable paradigm for hydrochemical research in analogous watersheds.

2. Materials and Methods

2.1. Overview of the Study Area

The Quzhou study area, located in Western Zhejiang Province (28°15′–29°30′ N, 118–119°20′ E), encompasses the upper, middle, and lower reaches of the Qujiang River system, a 140 km long critical tributary of the Qiantangjiang River. The upper basin, situated on the periphery of the Jinqu Basin, is characterized by low mountains, high peaks, and ridgelines, whereas the middle–lower reaches within the basin predominantly feature red-bed hills, intermountain basins, residual hills, and alluvial plains [17]. Elevation gradients in the region range from 50 to 300 m above sea level (ASL), with detailed geomorphological features shown in Figure 1a,b.

The area is characterized by a humid subtropical monsoon climate, receiving an average annual precipitation of 1908.04 mm and maintaining a mean annual temperature of 17.5 °C.

Geologically, the upper reaches are characterized by Early Paleozoic Ordovician and Cambrian strata, predominantly composed of fine sandstones, siltstones, and mudstones. In contrast, the middle and lower reaches are dominated by post-Cretaceous formations, mainly consisting of conglomerates, medium-to-coarse sandstones, and fine siltstones. Hydrogeologically, the aquifer system consists of two principal types: (1) unconsolidated porous aquifers and (2) red-bed clastic aquifers, which exhibit dual porosity–permeability regimes of porous water (intergranular flow) and fractured water (bedrock fracture networks) [18]. The spatial distributions of hydrogeological units are delineated in Figure 1c. Specific hydrogeological features can be seen in Figure 2.

2.2. Sample Collection and Testing

A total of 61 surface water samples were collected between July and September 2019, with sampling locations illustrated in Figure 1c. The sampling procedure followed the guidelines outlined in the “Technical Specifications for Surface Water Environmental Quality Monitoring” (HJ 91.2—2022), employing quadruple 250 mL sampling volumes. Prior to sampling, all containers were rinsed 2–3 times with in situ water to remove residual contaminants [19]. The determination methods for specific parameters are shown in

Table 1. Water quality parameters and determination methods.

Parameters	Determination Methods
pH, Eh, TDS, DO, EC	Portable multimeter (HQ40d, HACH, Shanghai, China)
HCO3−	Titration
K+, Na+, Ca2+, Mg2+, SO42−, NO3−, Cl−	Ion chromatography (IC, Metrohm 930 Compact) and UV-Vis spectrophotometry (Shimadzu UV-2600, Shimadzu Corporation, Kyoto, Japan)
TN	Alkaline persulfate digestion–UV spectrophotometry (HJ 636-2012)
TP	Ammonium molybdate spectrophotometry after sulfuric acid digestion (HJ 670-2013)
CODMn	Dichromate reflux method (HJ 828-2017)

.

To comprehensively assess the hydrochemical characteristics of the study area, this research integrated multiple advanced analytical methods. Spatial distribution patterns of water quality parameters were generated via the inverse distance weighting (IDW) interpolation method in ArcGIS 10.8, which extrapolates measured data from sampling points to visualize regional heterogeneity. Additionally, Piper trilinear diagrams constructed in Origin 2021 were assessed according to the ion composition and chemical types of the regional surface water [20]. The impacts of rock weathering, precipitation, and evaporation on hydrochemistry were investigated using Gibbs diagrams [21]. Finally, PCA loading plots generated in Origin 2021 were used to visually interpret the contributions of ions to principal components, providing critical evidence for deciphering hydrochemical evolution pathways.

3. Results and Discussion

3.1. Hydrochemical Composition Characteristics of Qujiang River Basin

A statistical analysis was performed on the test data of 61 surface water sampling points in Section 2.2. The analysis results are shown in

Table 2. Statistical characteristics of the hydrochemical composition of the Qujiang River.

Parameters		pH	K+	Na+	Ca2+	Mg2+	Cl−	SO42−	HCO3−	NO3−	TDS	Mn	Fe	As	Hg
Upstream	Min (mg/L)	6.39	1.13	2.04	5.64	0.70	0.81	2.45	18.10	1.86	22.00	0.12	0.82	1.30	0.04
	Max (mg/L)	7.15	2.08	9.07	26.72	4.80	13.14	19.96	73.60	8.61	132.00	5.20	2.31	1.50	0.04
	Mean (mg/L)	6.91	1.68	4.42	20.52	3.55	5.44	12.90	54.28	4.05	79.75	1.23	1.83	1.45	0.04
	SD (mg/L)	0.23	0.32	2.58	8.65	1.65	4.75	6.53	21.93	2.08	38.95	1.74	0.63	0.08	0.00
	CV (%)	3.41%	18.91%	58.46%	42.18%	46.68%	87.36%	50.58%	40.39%	51.40%	48.84%	141.3%	34.52%	5.21%	0.00%
Midstream	Min (mg/L)	6.31	1.63	1.34	7.18	0.68	1.35	4.96	18.10	0.02	20.00	0.07	0.82	1.20	0.04
	Max (mg/L)	8.83	5.21	16.80	51.73	8.34	21.17	34.14	126.70	13.39	200.00	337.51	127.64	2.80	0.17
	Mean (mg/L)	7.04	3.01	5.57	25.85	3.90	7.52	18.35	62.31	5.24	109.71	42.76	22.83	1.84	0.05
	SD (mg/L)	0.64	1.04	3.68	11.57	1.78	5.26	9.00	28.05	3.63	47.96	85.21	39.56	0.40	0.03
	CV (%)	9.18%	34.4%	66.10%	44.75%	45.65%	69.98%	49.06%	45.04%	69.21%	43.70%	199.3%	173.3%	21.76%	64.86%
Downstream	Min (mg/L)	6.85	3.36	5.53	15.45	2.57	6.16	10.08	30.20	2.34	80.00	0.09	0.94	1.20	0.04
	Max (mg/L)	8.10	4.21	26.78	20.57	2.77	24.79	41.86	51.90	8.02	168.00	112.64	22.80	2.10	0.06
	Mean (mg/L)	7.46	3.76	17.19	17.82	2.66	16.07	29.83	40.43	5.17	113.00	33.29	6.32	1.68	0.04
	SD (mg/L)	0.68	0.37	10.55	2.40	0.08	9.65	14.12	9.10	2.30	38.14	43.87	8.18	0.28	0.01
	CV (%)	9.11%	9.90%	61.37%	13.45%	3.16%	60.10%	47.34%	22.52%	58.07%	33.75%	131.8%	129.3%	16.50%	16.64%

Note: pH, dimensionless; min = minimum; max = maximum; mean = average; SD = standard deviation; CV = coefficient of variation.

. The pH values in the upstream section of the river ranged from 6.39 to 7.15, with an average of 6.914, indicating mildly acidic conditions. Among the cations, Ca²⁺ and Na⁺ were predominant, with average concentrations of 20.529 mg/L and 4.423 mg/L, respectively. For anions, HCO₃⁻ and SO₄²⁻ were the main components, with average concentrations of 54.288 mg/L and 12.908 mg/L, respectively. In the midstream section, the pH values fluctuated significantly between 6.31 and 8.83, with an average of 7.04, indicating a neutral character. The dominant cations were Ca²⁺ and Na⁺, with average concentrations of 25.852 mg/L and 5.566 mg/L, respectively, whereas the primary anions were HCO₃⁻ and SO₄²⁻, averaging 62.314 mg/L and 18.347 mg/L, respectively. The downstream pH values ranged from 6.85 to 8.1, with an average of 7.463, indicating weak alkaline properties. The major cations were Na⁺ and Ca²⁺, with average concentrations of 17.820 mg/L and 17.190 mg/L, respectively, and the dominant anions were HCO₃⁻ and SO₄²⁻, averaging 40.425 mg/L and 29.835 mg/L, respectively. These indicators are key components of the hydrochemical composition and play a significant role in determining hydrochemical types and understanding hydrochemical evolution processes. In addition, the average TDS shows a gradual increasing trend from 79.75 mg/L in the upper reaches to 113 mg/L in the lower reaches. Regarding metal ions, the exceedance of Mn, Fe, As, and Hg is relatively severe. High-value areas for manganese and iron appear in the middle reaches, with maximum values of 200 mg/L and 337.51 mg/L, respectively. The As concentration generally exceeds the standard in the middle and lower reaches, reaching a maximum of 2.8 mg/L.

Analysis of the boxplots (Figure 3a) revealed that the K⁺ and Mg²⁺ concentrations were the lowest, whereas HCO₃⁻ was the most abundant ion across the entire Qujiang River Basin. The upstream and middle reaches presented significant variations in HCO₃⁻, SO₄²⁻, and Ca²⁺ concentrations. Notably, the Na⁺, Cl⁻, Ca²⁻, and SO₄²⁻ concentrations were significantly higher in the middle–lower reaches than in the upstream section.

We assessed whether the water quality of the Qujiang River meets the requirements of ecological protection and the safety of domestic water for residents. A comparison with the Environmental Quality Standards for Surface Water (GB 3838–2002) [22] and the Sanitary Standards for Drinking Water (GB 5749–2006) [23] revealed that several ions exceeded permissible concentrations. Figure 3b presents boxplots of the ions surpassing these standards in the surface water of the Qujiang River. Among them, Mn²⁺ exhibited the most significant exceedance. The exceeding standard rate can reach 85%, followed by Fe³⁺, Hg²⁺, and As³⁺. The boxplots further illustrate pronounced spatial variability in Mn²⁺ and Fe³⁺ concentrations across the basin, whereas Hg²⁺ and As³⁺ levels remained relatively stable throughout the watershed. The box plots also show that the pH values in the middle and lower reaches are indeed higher than those in the upper reaches. Similarly, the concentrations of NO₃⁻ and TDS in the middle and lower reaches are greater than those in the upper reaches, consistent with the previous conclusions.

3.2. Spatial Distribution Characteristics of Hydrochemical Components in Qujiang River Water

The hydrochemical characteristics of surface water in the study area were examined using a Piper trilinear diagram (Figure 4). In the upper and middle reaches, ionic distributions display a consistent pattern: cations predominantly cluster near the Mg²⁺-Ca²⁺ combined end-member, with a tendency toward the Ca²⁺ end-member, while anions concentrate near the (HCO₃⁻ + CO₃²⁻)-Cl⁻ combined end-member, biased toward the (HCO₃⁻ + CO₃²⁻) component. This pattern indicates that water in these reaches is mainly of the Ca²⁺-HCO₃⁻ type, reflecting the limited anthropogenic influence and the dominance of natural processes such as rock weathering [20]. Additionally, the study area has a subtropical humid monsoon climate, with summer-concentrated precipitation. This provides abundant water for rock weathering and ion dissolution, directly influencing the formation of Ca-HCO₃-type water upstream. In contrast, the lower reaches exhibit noticeable shifts: cations tend to cluster near the (Na⁺ + K⁺)-Ca²⁺ combined end-member, with a bias toward Na⁺ + K⁺, while anions align more closely with the SO₄²⁻-Cl⁻ end-member, showing a bias toward SO₄²⁻. These shifts indicate that the hydrochemical types in the lower reaches are predominantly SO₄·Cl-Na and mixed SO₄·Cl-Na·Ca types, suggesting substantial anthropogenic influences, primarily from industrial wastewater and domestic sewage [24].

To visualize the spatial variations in ion concentrations across the upstream, midstream, and downstream sections of the Qujiang River Basin, a line chart was constructed, as shown in Figure 5. Nine surface water samples were randomly selected, with QJHL35, QJHL05, and QJHL57 representing upstream sites; QJHL16, QJHL24, and QJHL21 representing midstream sites; and QJHL81, QJHL75, and QJHL85 representing downstream sites. The results show elevated Ca²⁺ concentrations in the midstream section, likely due to intensified rock weathering, which releases additional Ca²⁺ into the water [24]. However, Ca²⁺ concentrations decrease in the downstream reaches. Coupled with the previously observed increase in HCO₃⁻ concentrations, it is speculated that high levels of HCO₃⁻ may lead to the formation of calcium precipitates, thus reducing Ca²⁺ concentrations. Additionally, anthropogenic activities such as industrial and agricultural practices may alter ion compositions, elevate other cation concentrations, and induce cation exchange reactions, further contributing to calcium depletion.

The magnesium ion concentrations remained stable throughout the basin, indicating minimal anthropogenic influence and dominance of rock weathering processes [21]. Unlike the Ca²⁺ and Mg²⁺ concentrations, the NO₃⁻ and SO₄²⁻ concentrations tended to increase in the middle–lower reaches, with downstream values exceeding upstream levels. This reflects significant downstream impacts from SO₄²⁻-bearing pollutants and intensive industrial activities. The variation trends of Mn, As, Hg, and Fe are basically consistent, with their peak values all occurring in the middle reaches, further verifying the argument that the middle reaches are affected by industrial activities. Additionally, Cl⁻, Na⁺, COD, and TDS exhibited substantial fluctuations, but overall increasing trends suggested that the accumulation of dissolved solids was influenced by combined natural (rock weathering) and anthropogenic factors [25].

From the spatial distribution maps of Mn²⁺, total Fe, total As, and Hg²⁺ (Figure 6), we can see that Mn²⁺ and total Fe exhibited similar spatial distribution patterns, with clustered spotty or patchy high-concentration areas primarily in the midstream region. The abnormally elevated concentrations are primarily attributed to industrial pollution near Jiangshan Harbor, corresponding to the highest concentration area shown in Figure 6a, and are linked to emissions from surrounding industrial facilities [26]. In the mid-lower reaches of the Qujiang River Basin, arsenic concentrations exceeded the standard at all sampling points, with the most severe contamination observed in the upper midstream section. This pattern, when considered alongside local environmental conditions, is largely due to mineral extraction activities and agricultural practices, including livestock farming [27]. Mercury concentrations were relatively stable across the basin, with localized high-concentration clusters mainly associated with agricultural and livestock activities [28]. In addition, geological background analysis indicates that other Hg-contaminated areas are likely influenced by the gradual release of mercury from the long-term weathering of rock formations, such as the Permian Longtan Formation coal-bearing rocks and the Carboniferous Chuanshan Formation platform carbonate and magnesium carbonate sediments [18].

3.3. Comparative Analysis of the Hydrochemical Composition in the Qujiang River and Global River Systems

The hydrochemical characteristics of the Qujiang River are distinct from global river systems (

Table 3. Comparison of major ion concentrations in the Qujiang River with other major rivers worldwide.

River	K+ + Na+	Ca2+	Mg2+	SO42−	Cl−	NO3−
Qujiang	10.21	23.41	3.59	19.32	8.70	4.89
Yangtze River [8]	8.20	34.1	7.60	11.7	2.90	3.44
Lancang River [27]	20.6	48.17	10.72	63.01	24.8	2.52
Nile Delta [6]	44.8	40.0	16.6	30.1	36.4	20.1
Thames River [28]	9.40	39.0	124	5.60	73.3	21.9

), shaped by the interplay of geological processes and anthropogenic activities. Notable disparities in cation (Ca²⁺ and Mg²⁺) and anion (NO₃⁻ and SO₄²⁻) concentrations exist when compared to typical rivers both domestically (e.g., Yangtze River and Lancang River) and internationally (e.g., Nile Delta and Thames River). For example, the Lancang River’s upstream is characterized by granite and metamorphic rocks. Driven by its climate, physical weathering processes in this area result in distinct cation-release patterns. In contrast, the Qujiang River has carbonate strata in its upper reaches and Cretaceous red beds in the middle–lower reaches. Coupled with its subtropical monsoon climate, these geological features influence the cation composition of the Qujiang River [29]. Regarding anions, the Qujiang River has lower NO₃⁻ concentrations than the Nile Delta [7] and the Thames River [30], but higher SO₄²⁻ levels. This can be mainly attributed to the pyrite mineralization in Jurassic coal measures and industrial discharges, such as those from the Quzhou Chemical Industrial Park. Despite the high population density in the Qujiang River Basin (212 persons/km² in 2021), which leads to elevated nitrogen accumulation, the NO₃⁻ concentrations in the Qujiang River remain lower than those in the aforementioned rivers. The relatively high SO₄²⁻ levels (29.8 mg/L) in the Qujiang River highlight the dual influence of sulfur-rich strata and industrialized watersheds. This is similar to the situation in other rivers affected by human activities, such as the Rhine. This similarity emphasizes the combined effects of natural geological factors and human activities on the hydrochemistry of the Qujiang River. It also aligns with the study’s thesis on the dual-driven influences and provides a comparative basis for understanding the formation factors of the Qujiang River’s hydrochemical characteristics.

3.4. Identification of Hydrochemical Component Sources in the Qujiang River

3.4.1. Source Discrimination via Gibbs Diagram Analysis

The Gibbs diagram classifies natural water chemistry into three principal genetic categories—rock weathering dominance, atmospheric precipitation dominance, and evaporation–concentration dominance [31]. Figure 7 shows that most water samples from the Qujiang River’s upper and middle reaches cluster within the rock weathering and evaporation–concentration zones, indicating that these natural processes dominate hydrochemical evolution. However, anthropogenic inputs, especially industrial and mining activities, significantly alter ionic composition via Na⁺-rich wastewater and mining leachates. For instance, downstream samples in Figure 7a exhibit abnormal Na⁺/(Na⁺ + Ca²⁺) ratios (>0.6), exceeding natural thresholds. Meanwhile, Figure 7b reveals Cl⁻/(Cl⁻ + HCO₃⁻) ratios < 0.3, suggesting mixed sources from rock weathering, evaporation, and industrial Na⁺ discharges. These results reinforce prior studies, highlighting the combined influence of natural and anthropogenic factors on the river’s hydrochemistry [24].

3.4.2. Hydrochemical Component Source Analysis via the Ion Ratio Method

When natural water bodies are not dominated by atmospheric precipitation, hydrochemical ions are derived primarily from the weathering of carbonate, silicate, and evaporite rocks [32]. Different ions have specific sources, and ion ratios can trace these sources [33].

As shown in Figure 8a, most samples cluster near the silicate end-member with proximity to carbonate, indicating that mid-upstream Ca²⁺, Na⁺, and Mg²⁺ originate from carbonate and silicate weathering [34]. Meanwhile, the downstream samples indicate that they are derived from evaporite and silicate dissolution. The ratios of Na⁺ and Na⁺ + K⁺ in the mid-upstream Qujiang samples are close to 1 (Figure 8b). This shows that mid-upstream K⁺ and Na⁺ have balanced concentrations from rock weathering, while the ratio exceeding 1 in the downstream suggests additional sources. The [Ca²⁺ + Mg²⁺]/[HCO₃⁻] ratio discriminates carbonate vs. silicate contributions [35]. Figure 8c shows samples below the 1:1 line, with HCO₃⁻ exceeding Ca²⁺ + Mg²⁺, indicating that extensive carbonate dissolution drove the precipitation reactions. This reflects carbonate mineral dominance in hydrochemical formations.

The [Ca²⁺ + Mg²⁺]/[HCO₃⁻ + SO₄²⁻] ratio reveals SO₄²⁻ sources [36,37]. Figure 8d shows scattered data points with HCO₃⁻ + SO₄²⁻ exceeding Ca²⁺ + Mg²⁺ (below the y = x line), indicating that SO₄²⁻ derives not only from evaporite dissolution but also from industrial inputs. Notably, the pore-fracture structure of Cretaceous red-bed clastic rocks in the middle and lower reaches promotes surface–groundwater exchange, and the dissolution of sulfate minerals in these rocks, along with industrial emissions, has led to a 2.3-fold increase in the SO₄²⁻ concentration in the lower reaches compared to the upper reaches.

3.5. Impacts of Anthropogenic Activities on Qujiang River Hydrochemistry

The mid-lower reaches of the study area represent key zones of anthropogenic influence in Quzhou, where industrial production, agricultural activities, and domestic sewage discharge exert a pronounced impact on the chemical evolution of surface waters. Ion ratios can be used to identify the sources of anthropogenic inputs [29]. Downstream samples with high Cl⁻ concentrations and low [NO₃⁻]/[Cl⁻] ratios suggest fecal and domestic sewage inputs (Figure 9a), while upstream samples with the opposite characteristics indicate stronger agricultural influences [24]. Midstream samples show a mix of urban sewage and agricultural impacts.

Industrial activities can be further identified through the [SO₄²⁻/Na⁺] vs. [NO₃⁻/Na⁺] relationship [24]. Across samples, [SO₄²⁻/Na⁺] is higher than [NO₃⁻/Na⁺], consistent with industrial wastewater discharge (Figure 9b). In downstream sites like QJHL75 and QJHL85, elevated [SO₄²⁻/Na⁺] values are due to localized industrial sources. The significantly higher [SO₄²⁻/Ca²⁺] in the downstream compared to the upstream further confirms that industrial activities contribute to sulfate enrichment, highlighting the more substantial industrial impacts in the downstream reaches (Figure 9c).

3.6. Identification of Dominant Factors Controlling Qujiang River Hydrochemistry

Principal component analysis (PCA) was applied to further identify the dominant hydrochemical controlling factors in the Qujiang River Basin. Thirteen water quality parameters (K⁺, Na⁺, Ca²⁺, Mg²⁺, Cl⁻, SO₄²⁻, HCO₃⁻, NO₃⁻, TDS, Fe, Mn, Ag, and As) were selected for PCA across upstream, midstream, and downstream reaches. The data were evaluated via the Kaiser–Meyer–Olkin (KMO) test and Bartlett’s sphericity test. KMO values of 0.62 (upstream) and 0.56 (mid-lower reaches), combined with significant Bartlett’s test results (620, p < 0.001; 943, p < 0.001), confirmed the suitability of the PCA. After Z-score standardization, varimax-rotated principal components were extracted on the basis of eigenvalues >1. Three dominant factors were identified (

Table 4. Principal component analysis of major ions in the Qujiang River Basin.

Parameters	PC1	PC2	PC3
K+	0.23	−0.04	0.41
Na+	0.24	−0.44	0.29
Ca2+	0.39	0.25	−0.16
Mg2+	0.37	0.20	−0.31
HCO3−	0.35	0.29	−0.24
Cl−	0.31	−0.38	0.21
SO42−	0.39	−0.18	0.12
NO3−	0.18	−0.13	−0.25
Fe	−0.01	0.39	0.38
Mn	0.06	0.31	0.46
Hg	0.07	0.10	−0.13
As	0.05	0.41	0.27
TDS	0.43	0.02	0.01
Characteristic value	4.82	2.64	1.75
Contribution rate (%)	37.10	20.29	13.47
Cumulative contribution rate (%)	37.10	57.39	70.86

), explaining 37.10% (PC1), 20.29% (PC2), and 13.47% (PC3) of the total variance.

PC1 retains significant rotated loadings for TDS, Ca²⁺, and SO₄²⁻ (Table 4), all showing strong positive correlations. As previously discussed, Ca²⁺ originates primarily from evaporite and silicate dissolution, whereas SO₄²⁻ sources include both evaporite weathering and industrial inputs. The downstream samples cluster along the positive PC1 axis in the biplot (Figure 10), corroborating the combined effect of evaporite dissolution (Ca²⁺) and industrial pollution (SO₄²⁻) inferred from Gibbs diagram deviations (Figure 7a).

PC2 has significant loadings on As, Fe, Mn, Na⁺, and Cl⁻, indicating complex hydrogeochemical conditions in the Qujiang Basin. The positive loadings of As, Fe, and Mn result from the combined influence of agriculture (such as arsenic from agrochemicals) and industrial factors (iron and manganese from mining-related activities) [21]. The negative loadings of Na⁺ and Cl⁻ imply cation exchange processes, where Ca²⁺-rich industrial effluents affect Na⁺ concentrations [18]. Spatially, midstream samples on the positive PC2 axis are dominated by industrial and agricultural pollution, while downstream samples influenced by Na⁺ and Cl⁻ tend towards negative PC2 values, showing the impacts of cation exchange and domestic sewage. Overall, PC2 demonstrates the joint effects of anthropogenic activities and water–rock interactions on the Qujiang River’s hydrochemical characteristics, further supporting the argument that both natural and human factors influence the river’s hydrochemistry. PC3 was loaded onto K⁺, Mg²⁺, and HCO₃⁻. The K⁺/Na⁺ ratio (0.41/0.29 = 1.41) lies between typical silicate (K⁺/Na⁺ > 1) and evaporite (K⁺/Na⁺ < 1) signatures, indicating mixed silicate weathering and localized evaporite dissolution [31]. Combining geological maps (Figure 1c), K⁺ enrichment may also result from K-feldspar dissolution in Cretaceous sandstones [17]. The negative Mg²⁺–HCO₃⁻ correlations (Figure 8c) suggest carbonate precipitation, implying that PC3 is influenced by local water–rock reactions (e.g., dolomite dissolution–precipitation equilibrium). Additionally, K⁺ loadings may reflect agricultural potassium fertilizer inputs [33]. Collectively, PC3 represents the integrated effects of rock weathering and anthropogenic activities. Integrating the above principal component analysis results, PC1 and PC2 account for the largest variance contributions to the Qujiang River Basin, as detailed in

Table 5. Comparison of hydrochemical controlling factors in the Qujiang River Basin.

River Section	Principal Component Contribution	Proportion of Natural Factors	Proportion of Human Activities	Characteristic Ions
Upstream	PC2 (41%)	78%	22%	Ca2+, HCO3−
Midstream	PC1 (53%)	52%	48%	SO42−, NO3−
Downstream	PC1 (67%)	34%	66%	Na+, Cl−

.

4. Conclusions

This study, through a variety of methods of analysis, including Piper trilinear diagrams, Gibbs diagrams, and principal component analysis (PCA), reveals the significant differences in hydrochemical characteristics between the upper and lower reaches of the Qujiang River, the sources of solutes, and the comprehensive influence of natural and anthropogenic factors on its hydrochemical evolution. It clarifies the characteristics of the high-quality water source dominated by natural factors in the upper reaches and the water quality characteristics affected by human activities in the lower reaches, highlighting the resource advantages of the Qujiang River in ecological maintenance, residential water supply assurance, and the potential exploration of industrial water use. The specific conclusions are as follows:

The upper reaches are dominated by natural weathering (Ca-HCO₃-type water, with natural factors accounting for 78%), where precipitation and carbonate dissolution are the key controlling processes.

The middle and lower reaches are significantly influenced by industrial wastewater (with SO₄²⁻ concentration increased by 2.3 times compared to the upper reaches) and domestic sewage (with Cl⁻ reaching 16.07 mg/L), with the contribution rate of anthropogenic factors reaching 48–66%.

Spatial analysis shows that the superposition of red-bed weathering and industrial activities has led to the transformation of hydrochemical types in the lower reaches to SO₄·Cl-Na type.

However, due to the large workload, insufficient data, and other reasons, there are still many shortcomings in this paper, and further research is needed:

• The data only cover the wet season from July to September 2019, lacking data from the dry season and, thus, making it difficult to reveal the seasonal dynamics of hydrochemical characteristics.
• In the analysis of complex mixed pollution sources, the existing methods are unable to accurately distinguish the contributions from natural and anthropogenic sources, affecting the in-depth analysis of the sources of solutes and the pollution mechanisms.

Future research could supplement monitoring during the dry season and incorporate techniques such as the Positive Matrix Factorization (PMF) model and isotope tracing to improve the study of the hydrochemical evolution mechanism of the Qujiang River.