Hydrogeological Background

The Huaibei Coalfield is located in the northwestern part of the Huaibei Plain in Anhui Province, China, and is separated from the Yangzi Plate by the Tanlu Fault in the east; it is bordered by the Xiayi Fault and the Henan Sedimentary Belt in the west, the Fengpei Rise of the Fengpei Fault in the north, and the Bengbu Rise of the Taihe-Wuhe Fracture in the south. Due to the superposition of multiple tectonic movements, the Huaibei Coalfield and its neighboring areas are crisscrossed by near-EW and NNE faults, forming a network of fault structures; these include the Fengpei, Subei, Banqiao-guzhen, and Taihe-Wuhe Faults in the near-EW direction and the Xiayi, Fengguo, and Guzhen-Changfeng Faults in the NNE direction (Fig. 1).

The area can be divided into the north and south, with the Suixiao mine in the north and the Guoyang, Linhuan, and Suxian mines in the south (Fig. 1). The area has a length of 50 km from east to west and a width of 25-45 km from north to south, covering an area of about 1,750 km². The area has production mines, such as the Linhuan, Haizi, Xutuan, Renlou, Tongting, and Suntuan mines [12]. According to the regional geological characteristics and the spatial distribution of groundwater storage, the groundwater in the Huaibei Coalfield is divided into four aquifers: the Cenozoic loose layer pore aquifer, the Permian coal measure sandstone fissure aquifer, the Carboniferous Taiyuan formation limestone karst fissure aquifer, and the Ordovician limestone karst fissure aquifer [12, 19]. In the Huaibei Coalfield, the groundwater of the Taiyuan limestone aquifer, Ordovician limestone fissure aquifer, and coal measure sandstone fissure aquifer are communicated through water-conducting fault, karst collapse column and mining fissure to form different degrees of hydraulic connection. A thick loose layer covers the southern part, and the Taiyuan limestone aquifer is connected with the Ordovician limestone karst fissure aquifer and the Cenozoic loose layer pore aquifer, forming a hydraulic connection [20].

Fig. 1.

Geological structure and distribution of the mining area in Huaibei coalfield.

The mineral composition of typical coal mines in the Huaibei coalfield can be summarized as follows: The metamorphic rock series is the bedrock, and the upper part is covered with a Cenozoic loose layer. The main coal-bearing strata are from the Carboniferous and Permian, and the lithology includes limestone, aluminous mudstone, sandstone, mudstone, and coal seams, reflecting the characteristics of marine and continental interaction. Among them, the coal seam is sandwiched between rock layers, and the thickness is unstable. In addition, it also contains Ordovician limestone, rich in calcite, and Jurassic, Cretaceous, and other continental sedimentary rocks. It has a long history of exploitation, which appeared in the early Tang Dynasty, began large-scale exploitation in the middle of the Ming Dynasty, and was interrupted by wars repeatedly in the late Qing Dynasty and early Republic of China. After the founding of New China, large-scale exploration and development began in the 1950s, with multiple mines, accelerated infrastructure, and frequent technological innovation, which gave birth to the prosperity of Huaibei City. At present, in the face of resource depletion, Huaibei City is actively transforming, exploring a sustainable road, committing to ecological restoration, and revitalizing the regional economy.

Sample Collection and Testing

Based on the analysis of topographic and geomorphic features as well as the hydrogeological conditions of the study area (Fig. 1), nine groups of groundwater samples from different aquifers were taken in July 2015, including four, three, and two groups from the loose layer aquifer (mainly taken from the second and the third aquifers), Taiyuan, and the Ordovician limestone aquifer, respectively (Table 1). Water samples were stored in 0.5 L sterile polyethylene bottles, preserved on dry ice for 24 h before transportation to the laboratory, and filtered through disposable syringes and 0.45 μm microporous aqueous filtration membranes into 10 mL centrifuge tubes. Samples were adjusted to pH<2 by adding superior purity HNO₃ and were then refrigerated at 4ºC as backups.

Sample collection procedures were carried out in strict accordance with the Technical Code for Environmental Monitoring of Groundwater in China (HJ/T 164-2004). Physicochemical indicators such as pH, temperature, Total Dissolved Solids (TDS), and Electric Conductivity (EC) were measured on-site using a portable analyzer (OHAUS Shanghai, China, Suite310). HCO₃^- was determined by acid-based titration. Na⁺, K⁺, Mg²⁺, Ca²⁺, F^-, Cl^-, SO₄^2-, NH₄⁺, and NO₃^- were determined by ion chromatography (Thermo Fisher Scientific, ICS-600/ICS-900). Cr, Mn, Fe, Co, Ni, Cu, Zn, and Pb were determined by inductively coupled plasma mass spectrometry (Shimadzu, ICPMS-2030LF). The quantitative methods were based on the external standard method, and the linear correlation coefficients of the calibration curves were all higher than 0.999. The relative errors of the anion and cation balances of all water samples were less than 5%. The sample processing and analytical tests were done in the Key Laboratory of Mine Water Resource Utilization of Anhui Higher Education Institutes (Suzhou University, Anhui, P.R. China).

DNA Extraction, PCR Amplification, and Sequencing

Groundwater samples (2 L) were filtered through pre-autoclaved microporous filter membranes with a pore size of 0.22 μm to collect biomass, and the membranes were preserved on dry ice and sent to Shanghai Meiji Bio-pharmaceutical Science and Technology Co. for 16S rRNA high-throughput sequencing. Genomic DNA was extracted using a DNA isolation kit (OMEGA-soil DNA Kit, Omega Bio-tek, Norcross, GA, U.S.). After completion of genomic DNA extraction, its integrity and purity were detected using 1% agarose gel electrophoresis. The V4-V5 region of the bacterial 16S rRNA gene was amplified using the universal bacterial sequencing primers 338F (ACTCCTACGGGGAGGCAGCAG) and 806R (GGACTACHVGGGGTWTCTAAT) on an ABI Gene Amp® 9700, and all samples were run in accordance with the formal conditions. All samples were analyzed in accordance with the formal experimental conditions, with three replicates for each sample. PCR products of the same sample were mixed and detected by 2% agarose gel electrophoresis, gel purified using an AxyPrepDNA Gel Recovery Kit (AXYGEN). The products were eluted on Tris HCl. Referring to the preliminary quantitative electrophoresis results, the PCR products were detected and quantified using a QuantiFluorTM-ST Blue Fluorescence Quantification System (Promega); after this, they were mixed in the appropriate proportion according to the sequencing amount required for each sample and were sequenced on the Illumina Miseq sequencing platform.

Data Analysis

The processing and analysis of the data (groundwater hydrochemical data as well as factor and correlation analysis) of this experiment were completed using Excel 2016 (Microsoft) and IBM SPSS Statistics 26. CorelDRAW 2023 was used to draw the distribution map of the groundwater sampling sites. Piper trilinear, Gibbs, and ion-scaled plots were generated using Origin 2024.

Characterization of Water Chemical Composition

The hydrochemical and physicochemical statistics of groundwater in the study area are shown in Table 2. The pH of groundwater in the loose layer ranged from 7.45 to 7.88 (mean: 7.66) and was weakly alkaline; the ρ(TDS) ranged from 1260 to 1690 mg·L^-1 (mean: 1430 mg·L^-1). The pH of the groundwater in the Ordovician limestone aquifer ranged from 7.20 to 7.51 (mean: 7.31) and was weakly alkaline; the ρ(TDS) ranged from 1260 to 3400 mg·L^-1 (mean: 2136.67 mg·L^-1). The pH of the groundwater in the Taiyuan limestone aquifer ranged from 7.30 to 7.43 (mean: 7.37) and was weakly alkaline; the ρ(TDS) ranged from 711 to 2140 mg·L^-1 (mean: 1425.50 mg·L^-1). Overall, the groundwater in the study area was weakly alkaline and hard.

The fluctuation of water chemistry in the groundwater from different aquifers was large. The anion concentrations in the groundwater of the loose and Ordovician limestone layers were, in descending order, HCO₃^- >SO₄^2- >Cl^- >F^- >NO₃^-; in the Taiyuan limestone aquifer, they were SO₄^2- >HCO₃^- >Cl^- >F^- >NO₃^-. The cation concentrations in groundwater from the loose Taiyuan limestone and Ordovician limestone aquifers were Na⁺ > Ca²⁺ > Mg²⁺ > K⁺ > NH₄⁺. The coefficient of variation (CV) indicates the degree of dispersion of the data; in pollutant traceability studies, higher CV values of pollution factors usually represent greater anthropogenic perturbation. The chemical components of deep groundwater, in addition to anthropogenic perturbation factors, are more reflective of the differences in the type and degree of water-rock interactions [20, 21]. In this study, the coefficients of variation of anions and cations in different aquifers ranged from 0.00 to 1.73, indicating a high spatial variability of groundwater hydrochemistry in different aquifers. In addition, the coefficients of variation (CV) of K⁺ (loose layer, Taiyuan limestone, and Ordovician limestone aquifers), Na⁺ (Ordovician limestone aquifer), F^- (Taiyuan limestone and Ordovician limestone aquifers), NO₃^- (Ordovician limestone aquifer), and Cl^- (Taiyuan limestone and Ordovician limestone aquifers) in different aquifers were high (0.5<CV≤1) (Table 2), which was speculated to be related to the differences in the mineral composition of the rocks in the aquifers. In particular, the CV value of NH₄⁺ in the loose bed groundwater was greater than 1, indicating that this ion has obvious spatial variability; it was hypothesized that the loose bed groundwater may have been affected by human activities, resulting in a high degree of local NH₄⁺ enrichment [22], and disturbance from coal seam mining has enhanced the hydraulic connection between the aquifers [12].

Characterization of Groundwater Hydrochemical Types

The Piper three-line diagram is one of the most widely used graphics in hydrogeochemical analysis because it contains no interference from human factors; using this method, the type of groundwater hydrochemistry and its evolutionary characteristics can be visualized [23-25]. Cations were mainly distributed in the Na⁺ + K⁺ end member, and the anions were dominated by HCO₃^- and SO₄^2-. In the upper-middle rhombic region, the SO₄•Cl-Na type dominated, followed by the SO₄•Cl-Ca•Mg and HCO₃-Mg•Ca types (Fig. 2), indicating that the formation of groundwater hydrochemistry in the study area is mainly controlled by the weathering and dissolution of carbonate minerals [22]. Under the influence of factors such as mining activities and anthropogenic perturbations, groundwater chemistry types show diverse characteristics [26].

Factors Controlling Water Chemistry

Correlation Analysis

In the study of hydrogeochemistry, correlation analysis determines whether the ions in the water chemistry components result from the same origin; it is usually assumed that significantly correlated variables have a similar origin [22, 27]. The results obtained from the Pearson correlation coefficient matrix are shown in Fig. 3. TDS in the groundwater of the study area was mostly significantly positively correlated with Ca²⁺ and Mg²⁺ (r>0.6, P<0.05), and was also significantly positively correlated with Na⁺ + K⁺ (P<0.01; correlation coefficient: 0.980), which indicates that these ions had a greater contribution to TDS. The correlations with all anions were not significant. There was a significant negative correlation between Ca²⁺ and HCO₃^- in the groundwater of the study area (r = -0.687, P<0.05), indicating that dissolution and weathering of carbonate and silicate minerals may have occurred in the groundwater [22, 28]. There was a significant positive correlation between Mg²⁺ and Na⁺ + K⁺ in groundwater (P<0.05; correlation coefficient: 0.773), which indicates that they had similar origins and may have originated from the same rock weathering process [22]. In addition, there were significant positive and negative correlations between the above two different groups of ions, which may be related to different types of water-rock interactions under different hydrogeological conditions [29].

Fig. 2.

Piper diagram of groundwater in different aquifers within the study area.

Water-Rock Interaction Mechanisms

Groundwater chemical components change through water-rock interactions with the surrounding media; these interactions can reveal the mechanism of hydrochemical evolution, which can be classified into three types: rock weathering, atmospheric precipitation, and evaporation-concentration [23, 30-32]. As can be seen from the Gibbs diagram (Fig. 4), Gibbs Ⅰ (Na⁺/(Na⁺+Ca²⁺)) ranged from 0.46 to 1.00, while Gibbs Ⅱ (Cl^-/(Cl^-+HCO₃^-)) ranged from 0.00 to 0.67.

In Gibbs Ⅰ, the groundwater samples were mainly located in rock weathering and evaporation-concentration zones, which suggests that the hydrological and geochemical processes and the hydrochemical formation of the groundwater in the study area are synergistically controlled by water-rock interactions and evaporation-concentration. A small number of groundwater samples (loose and Taiyuan limestone aquifers) fall outside the control area; these sampling sites are located in the coal mining area, which is more disturbed by factors such as mining and other human activities.

In Gibbs II, most of the groundwater samples point to rock weathering end elements, indicating that water-rock interactions dominantly control the hydrochemical characteristics of groundwater in the study area; a small number of samples fall in the evaporation-concentration region, indicating that some of the groundwater sources in the study area are also affected by evaporation-concentration interactions. The groundwater samples in the study area are all far away from the precipitation control region, indicating that the contribution of atmospheric precipitation to the main ions is small. This conclusion is consistent with the findings of Liu et al. on the groundwater in the Huainan coal mine area, indicating that the ionic components of groundwater in the study area are mainly controlled by water-rock interactions [22].

Fig. 3.

Correlation of groundwater chemical components in the study area. Note: * indicates significant at the 0.05 level, ** indicates significant at the 0.01 level; the direction of the long axis of the ellipse indicates the positive and negative correlation coefficients, with positive values in the upper-right-bottom-left direction and negative values in the upper-left-bottom-right direction, and the flatter the ellipse, the greater the correlation coefficients, and vice versa, the smaller the correlation coefficients.

Fig. 4.

Gibbs diagram for groundwater of the study area.

Microbial Community Structure

Community Diversity Analysis

The Operational Taxonomic Units (OTUs) and bacterial diversity indices of the samples are shown in Table 3. Coverage values ranged from 0.9988 to 0.9995, all above 99%, indicating that the reliability of the sequencing data was high. Abundance-based Coverage Estimator (ACE) and Chao indices usually reflect the total number of species, while Shannon and Simpson’s indices are usually used to estimate community diversity, making up the alpha diversity index; larger Shannon values indicate higher diversity, whereas larger Simpsons’ values indicate lower diversity [12, 34]. From the mean values, the ACE, Shannon, and Chao values were loose layer>Ordovician limestone aquifer>Taiyuan limestone aquifer. The Simpsons’ index was loose layer <Taiyuan limestone aquifer<Ordovician limestone aquifer, indicating that the bacterial community of the loose aquifer was the most diverse, followed by the Ordovician limestone and Taiyuan limestone aquifers. The average values of the observed species index in the three aquifers were loose layer (224)>Ordovician limestone aquifer (220)>Taiyuan limestone aquifer (200). The results showed that bacterial community richness was higher in the loose aquifer. In addition, Simpson and Shannon indices of Taiyuan limestone and Ordovician limestone aquifers are close, which is due to the mutual hydraulic relationship between Taiyuan limestone and Ordovician limestone aquifers, which makes the microbial community diversity structure between the two aquifers similar. [12, 19].

Fig. 5.

Relationship between Mg²⁺/Na⁺, Ca²⁺/Na⁺, and HCO₃ ^-/Na⁺ of groundwater.

Table 3.

Bacterial OTUs and diversity indices of samples.

Sample		Coverage	Observed Species	ACE	Simpson	Shannon	Chao
loose layer aquifer	LE-1	0.9990	194	212.07	0.16	2.78	213.89
	LY-2	0.9995	226	234.19	0.18	2.65	232.65
	BS-1	0.9989	236	260.38	0.14	2.79	262.64
	RL-2	0.9988	240	261.01	0.09	3.37	261.57
	AVG	0.9991	224	241.91	0.14	2.90	242.69
Taiyuan limestone aquifer	LY-3	0.9991	212	225.40	0.15	2.83	223.50
	WG-1	0.9988	212	237.29	0.12	3.00	238.71
	RL-3	0.9990	175	198.42	0.19	2.30	195.22
	AVG	0.9990	200	220.37	0.16	2.71	219.14
Ordovician limestone aquifer	LY-1	0.9991	242	255.59	0.17	2.83	255.00
	RL-1	0.9993	198	207.88	0.17	2.69	204.33
	AVG	0.9992	220	231.74	0.17	2.76	229.67

Analysis of Microbial Community Composition

Differences in the family-level composition of microflora with <1% abundance in the combined aquifer are shown in Fig. 6. Bacillaceae was the largest family in the groundwater of the study area, with a relative abundance of 29.93%-50.03% and a mean abundance of 39.83%. The relative and mean abundances of other major families were Streptococcaceae (29.93%-50.03%, 39.83%), Pseudomonadaceae (5.97%-10.93%, 8.71%), Burkholderiaceae (0.97%-38.56%, 9.38%), Moraxellaceae (2.82%-13.20%, 5.26%), Halothiobacillaceae (0.00%-31.76%, 4.54%), Xanthomonadaceae (1.66%-3.42%, 2.58%), and Weeksellaceae (0.64-2.35%, 1.15%). The abundance of Burkholderiaceae in BS-1 and Halothiobacillaceae in RL-3 were>30%, 38.56%, and 31.76%, respectively. It was hypothesized that anthropogenic disturbances, mining activities, and hydraulic connections between aquifers may have replenished nutrients that favor the growth and metabolism of Burkholderiaceae and Halothiobacillaceae.

The dominant bacterial families varied between aquifers at different depths. Bacillaceae remained the dominant family in the loose-layer samples, followed by Streptococcaceae, Pseudomonadaceae, and Burkholderiaceae (39.9%, 14.7%, 8.5%, and 13.5%, respectively). Bacillaceae, Streptococcaceae, Halothiobacillaceae, and Pseudomonadaceae were the top four bacterial families in the Taiyuan limestone aquifer samples (39.8%, 15.0%, 10.6%, and 9.0%, respectively). The top four families in the Ordovician limestone aquifer samples were Bacillaceae, Streptococcaceae, Pseudomonadaceae, and Moraxellaceae (46.4%, 18.6%, 10.1%, and 4.5%, respectively). The sum of the relative abundance of Bacillaceae and Streptococcaceae in each aquifer was greater than 50% at different depths, indicating that these families were the most dominant microflora in the groundwater of the study area, the microbial community structure is more balanced. The differences in aquifers at different depths may be reflected in the diversity of environmental conditions and the ecological adaptability of microorganisms, indicating the complexity and uniqueness of different groundwater microbial ecosystems [35, 36]. The relative abundance of dominant bacterial families in the groundwater in the study area varied at different depths and was loose layer>Taiyuan limestone aquifer >Ordovician limestone aquifer. Deeper groundwater had lower bacterial abundance, consistent with the results of Chen et al. characterizing the bacterial community in the groundwater of the Huaibei Coalfield [12].

Fig. 6.

Bacterial communities in groundwater from different aquifers at the family level.

Correlation Analysis between Microbial Communities and Environmental Factors

Autocorrelation exists between different impact factors; the closer the relationship, the greater the interference, meaning that the higher the autocorrelation, the more unfavorable the analysis of the factors impacting microbial community structure is [37]. The influence factor is considered unimportant if the Variance Inflation Factor (VIF) value exceeds ten [38]. Environmental factors leading to strong covariance (VIF>10) [17] were eliminated (Table 4).

To analyze the correlation between environmental impact factors and bacterial communities, five environmental variables (pH, F^-, CI^-, SO₄^2-, HCO₃^-) were screened (Fig. 7). CCA analysis showed that the degree of impact of the factors on the bacterial communities was HCO₃^- > F^- > CI^- > pH > SO₄^2-. HCO₃^-, F^-, and CI^- were the three most important drivers of the groundwater bacterial community structure. The angles between HCO₃^- and pH and CI^- F^- and SO₄^2- were acute and positively correlated, indicating that the mass concentration of these parameters increased together. HCO₃^- was positively correlated with BS-1 and RL-2, and pH and CI^- were positively correlated with LE-1 and LY-2, suggesting that HCO₃^-, pH, and CI^- had a greater effect on the loose layer. F^- and SO₄^2- were positively correlated with LY-1 and RL-1, as well as LY-3 and RL-3, suggesting that both F^- and SO₄^2- had a different degree of effect on both the Ordovician limestone and Taiyuan limestone aquifers. Studies have shown that environmental factors impact unique aquifers differently, causing environmental selection and adaptation in microbial communities [12]. The distance between the projection point and the origin on the influence factor vector indicates the strength of the correlation between the bacterial family and the influence factor (Fig. 7). The dominant bacterial families were mainly distributed along the pH vector, primarily because the pH of the groundwater in the region is weakly alkaline, which is favorable for the growth of bacteria. Zhang et al. [39] also showed that in a higher pH environment, the diversity of the microbial community would increase, which was conducive to the growth of the bacteria. Bacillaceae, Streptococcaceae, Pseudomonadaceae, Moraxellaceae, Weeksellaceae, and Xanthomonadaceae showed positive correlations with HCO₃^-, F^-, and Cl^-, suggesting that a certain range of these ions favored their growth and that these bacteria may use HCO₃^- as a carbon source or participate in other metabolic processes. F^- and Cl^- may act as regulatory ions or participate in the stability of cell membranes to make them adaptable in environments rich in these elements. Previous studies have shown that Firmicutes and Proteobacteria are tolerant and adaptable in environments rich in F^- and Cl^- [39, 40].

Fig. 7.

Canonical correspondence analysis of samples.

The pollutants HCO₃^- and F^- were the most important factors influencing the composition of the groundwater bacterial community in the study area. Mining activities promote the hydraulic connection between different aquifers and accelerate the dissolution and weathering of carbonate and silicate minerals, which gradually increases HCO₃^-; their influence on the groundwater bacterial community is also prominent. Huaibei Coalfield has strong groundwater cation exchange due to the high F^- content, which decreases Ca²⁺ and increases Na⁺. Lower Ca²⁺ concentration can promote the dissolution of fluorine-containing minerals such as fluorite, accelerating the release of F^- into the water column [12, 41].

Changes in contaminant concentrations lead to changes in the composition of microbial communities as microorganisms adapt to the environment by regulating their abundance and diversity [42, 43].

Different aquifers have different stratigraphic characteristics, hydraulic connections, and storage conditions that lead to water chemistry differences, affecting microbial communities’ composition and metabolism. The microbial community achieves a dynamic balance through regulating nutrient transport, geochemical cycling, and pollutant degradation [44]. Examining the groundwater of different aquifers in the Huaibei Coalfield, the differences in bacterial community structure and controlling factors in different aquifers could provide a potential method for identifying water sources during mine water emergencies. However, it should be emphasized that this study has certain limitations. The research results have certain reference significance for mining areas with similar hydrogeological conditions. When there are significant differences in hydrogeological conditions in mining areas, it is necessary to reanalyze different aquifers’ hydrochemical characteristics and microbial diversity.