Bacterial communities in the polluted area of pyrite tailings: From the upstream, pollutant source, and to the downstream

Pyrite tailings can cause serious pollution to the surface water as the strong acidity, high iron and sulfate concentration in the leachate. The bacterial communities of pyrite tailings polluted area were still unclear which could restrict the recognition of the pyrite tailings pollution effect and further impede the development of microbial or ecology treatment technologies. In this study, the bacterial communities in the polluted area of pyrite tailings, from the upstream, pollutant source, and to the downstream, were analyzed with Illumina HiSeq sequencing. Results showed that Acinetobacter and Flavobacterium were abundant in the water and sediment of upstream and downstream while Bacteroides, Lactobacillus, and Akkermansia were abundant in the pollutant source. Sulfur-metabolizing or iron-metabolizing bacteria extensively existed in the polluted area in which Acidiferrobacter, Ferrithrix, and Desulfovibrio played crucial roles on the whole communities. Sulfur-metabolizing bacteria (e.g. Thiomonas, Sulfurospirillum, and Desulfobulbus) and iron-metabolizing bacteria (e.g. Ferrimicrobium, Ferrithrix, and Ferrovum) were introduced to the river polluted by pyrite tailings. Pyrite tailings can remarkably change the physicochemical characteristics and bacterial communities of river water and sediment.


Introduction
Pyrites (FeS 2 ) were widespread and important minerals, mainly used in the manufacture of sulfuric acid, and were heavily mined in the last few decades globally which generated huge numbers of pyrite tailings.
In many developing countries and low income areas, pyrite tailings were discarded in the valley near the pyrite mine without further reasonable disposal as the reuse value of pyrite tailings was low. The FeS 2 in pyrite tailings can be oxidized to soluble iron (Fe 2+ and Fe 3+ ), sulfate, and hydrogen ion (H + ), present in the leachate, with the participation of oxygen, water, and certain microbes when the pyrite tailings are directly exposed to the air (Lowson, 1982;Edwards et al., 1998). Therefore, the pyrite tailings leachate, one kind of acid mine drainage (AMD), has strong acidity and contains high concentration of iron (Fe), sulfate (SO 4 2− ), and other heavy metals which can cause serious pollution to the downstream water (Li et al., 2019). The ecological system of the river polluted by pyrite is usually seriously destroyed by low pH, high heavy metal concentration, and high turbidity caused by hydrolysis of iron ion (Liu et al., 2015). To better understand the pollutant generating process and further reduce the production of pollution, plenty of research had revealed the special bacteria such as Thiobacillus, Leptospirillum, Sulfobacillus, and Thiomonas which could fasten the pyrite oxidization (Crundwell, 1996;Edwards et al., 1998 To increase pH and precipitate iron ion, alkali neutralization process was used to AMD treatment and To explore the bacterial communities in the polluted area of pyrite tailings, an actual river owing through the pyrite tailings polluted areas was selected and the bacteria in the water and sediment of upstream, pollutant source, and downstream were analyzed by Illumina Hiseq sequencing in this study. The ndings innovatively revealed the bacterial communities of pyrite tailings polluted area and further con rmed the effect of pyrite tailings pollution on natural water bodies.
2 Materials And Methods

Research spot and sampling method
The research spot (110°2'44" E, 32°38'48" N), polluted by open stacked pyrite tailings, was located on one river in a county of China (Fig. 1). From 1950s to 2000s, millions cubic meter of pyrite tailings generated and were discarded in the county and over 0.3 million cubic meter of pyrite tailings were stacked in the studied river region. One pyrite tailings leachate treatment project, mainly taking lime and PAM as the treatment agent, was established in 2021 and the iron ion and acidity of leachate were partially removed.
The position of treatment project was the pollutant source of this region. The water and sediment in pollutant source, upstream (not contaminated) and downstream (contaminated) were sampled ( Table 1).
The water and sediment were sampled thrice (named XX-1, XX-2, and XX-3) with sterile bottles and sterile medicine spoon, respectively, and stored at 4℃ (for conventional analysis) and -20℃ (for bacterial communities detection). In detailed, water samples used to bacterial communities detection were ltered with 0.22µm sterile micro ltration membrane after sampling immediately and stored at -20℃.

Illumina sequencing and data processing
The DNA from the water samples and sediment samples was extracted with the PowerSoil® DNA Isolation Kit (MOBIO, USA) for Illumina sequencing. Universal PCR bacterial primer sets 338F (5'-ACTCCTACGGGAGGCAGCA-3') and 806R (5'-GGACTACHVGGGTWTCTAAT-3') were used to amplify the 16S rRNA gene sequences and the Illumina HiSeq sequencing (HiSeq 2500, Illumina, USA) was conducted by Beijing Biomarker Technologies Co. Ltd., Beijing, China. The PCR procedure and sequencing data processing were described by previous study (Li et al., 2020a). Operational taxonomic units (OTUs) were regarded with 97% similarity threshold. Taxa were assigned to all OTUs by comparing them to SILVA databases using QIIME2. The network of total bacterial communities was operated by the CoNet module in Cytoscape (3.6.1) with ve methods (Pearson correlation, Spearman correlation, mutual information similarity, Bray Curtis dissimilarity distance, and Kullback-Leibler dissimilarity distance) (Li et al., 2020b) and the relationship of sulfur-related and iron-related bacteria was calculated with Spearman method (correlation index > 0.5 and P < 0.05) in SPSS (25.0). The network was nally visualized with Gephi (0.9.1).  3.2 Bacterial communities in pyrite tailings polluted area Based on Illumina HiSeq sequencing, 1,174,592 sequences were obtained from 21 samples in this study, assigned to 2,261 OTUs. The alpha diversity index of samples is shown in Fig. S1. The solid in pollutant source (i.e. PT and ST) had low bacteria richness but high bacteria diversity. The bacteria diversity of water and sediment samples in downstream (contaminated by pyrite tailings leachate) was much lower than that in upstream, indicating pyrite tailings leachate signi cantly changed bacterial communities of the river. Samples from different positions showed obvious distinction especially in upstream (WU and SU) and downstream (WD and SD) based on principal component analysis (Fig. S2). Samples obtained from the pollutant source (PT, PL, and ST) were relatively similar compared with other positions. Proteobacteria (mean at 47.0%) was the most abundant bacterial phylum in this study, followed by Firmicutes (mean at 16.4%) and Bacteroidetes (mean at 16.2%), shown in Fig. 2a. In pollutant source, the relative abundance of Cyanobacteria, Verrucomicrobia, and Nitrospirae was higher than upstream and downstream. Acinetobacter (mean at 17.4%) and Flavobacterium (mean at 11.1%) was the most abundant bacterial genus in upstream water and sediment, respectively, while the relative abundance of Novosphingobium (mean at 18.1%) increased in the downstream (Fig. 2b). The bacterial communities of pollutant source in the genus level had visible difference with the river where Bacteroides (mean at 3.2%), Lactobacillus (mean at 2.6%), and Akkermansia (mean at 2.4%) were abundant in the pollutant source. The bacterial communities in samples from the liquid phase or sediment phase showed distinction analyzed by ternary method (Fig. S3). Some genus belong to Chloro exi (such as Kouleothrix and Leptolinea) tended to inhabit in pyrite tailings leachate compared with water in the upstream or downstream and many iron-related genus such as Ferrimicrobium, Leptospirillum, and Ferrovum also tended to exist in pyrite tailings leachate. In sediment phases, many genus belong to Firmicutes (Butyrivibrio and Megasphaera, for instance) were more likely to live in leachate treatment sediment and sulfur-related genus, e.g. Thiobacillus, Sulfuricurvum, Desulfoprunum, were also likely to exist in leachate treatment sediment compared with sediment in the upstream or downstream.
Bacteria network was generated with 736 nodes and 1628 edges which copresence relationship was 58.1% and mutual exclusion relationship was 41.9% (Fig. 3). The average degree, average path length, network diameter, average clustering coe cient, and modularity index of this network was 4.57, 2.69, 9, 0.06, and 0.72, respectively. Bacterial relationship in the polluted area of pyrite tailings showed obvious distinction and aggregation which was divided into three main modules since the role of keystone species. Acidiferrobacter, an extreme acidophilic iron oxidizer (Issotta et al., 2017), played an important role in the bacterial communities whose species was the center of mutual exclusion with other species. Ferrithrix, an acidophilic iron oxidizer or reductor (Johnson et al., 2009), played similar mutual exclusion center roles with Acidiferrobacter. Bacteriovorax, a biphasic life cycle (predacious form or saprophytic form) bacteria (Baer et al., 2000), was another mutual exclusion center in the bacterial communities.
Different with the two mutual exclusion center, copresence relationship showed relatively uniform distribution. Alloprevotella, Desulfovibrio, and Lactobacillus had the highest degree in the copresence relationship and all these three genus were anaerobic bacteria (Devereux et al., 1990;Downes et al., 2013;Zheng et al., 2020). The bacterial communities showed great characteristics which could re ect the pyrite polluting area as the key effect of sulfate-reducing bacteria (i.e. Desulfovibrio) and iron-metabolizing bacteria (i.e. Acidiferrobacter and Ferrithrix).

Sulfur-metabolizing bacteria and iron-metabolizing bacteria distribution
Abundant sulfur-metabolizing bacteria and iron-metabolizing bacteria existed in pyrite tailings polluting area as the su cient sulfate and iron of the polluted area. The pollutant source (i.e. PT, PL, and ST) contained high amount of sulfur-metabolizing bacteria and iron-metabolizing bacteria and further in uenced the detailed distribution in the downstream (Fig. 4). Thiobacillus, Acidiferrobacter, and Cloacibacterium were the predominant sulfur-oxidizing bacteria and showed obvious different habitat characteristics (Fig. 4a). Thiobacillus (2.6%) was mainly existed in leachate treatment sediment while Acidiferrobacter (1.3%) and Cloacibacterium (1.1%) tended to live in pyrite tailings and downstream sediments, respectively. Desulfovibrio (0.3%) was the most abundant sulfate-reducing bacteria in this study and widely distributed in all samples. The total abundance of sulfur-metabolizing bacteria in the downstream (water and sediment) was higher than that in the upstream, caused by the input of pollutant sources. Flavobacterium had the highest relative abundance (2.9%) in sulfur-metabolizing bacteria but mainly existing in the upstream and downstream, indicating that Flavobacterium was the natural and background bacteria but not the bacteria introduced by pollutant. Compared with the upstream water, the pyrite tailings leachate markedly increased the relative abundance of many sulfur-metabolizing bacteria, such as Thiomonas (403 times), Sulfurospirillum (19 times), and Desulfobulbus (4 times), in the downstream water and the leachate was the source of Acidiferrobacter and Sulfurisoma in the downstream water (Fig. S4a).
Pseudomonas (1.3%), Sphingomonas (1.0%), and Thiobacillus (0.5%) were the predominant ironmetabolizing bacteria and mainly existed in pollutant sources (Fig. 4b). Thiobacillus, has the ability to form bio lms on pyrite and further oxide ferrous and sul de (Crundwell, 1996), mainly existed in solids (i.e. PT and ST) which had the potential to reinforce the dissolve of iron and sulfur from pyrite tailings and leachate treatment sediment. Therefore, the leachate treatment sediment (ST) should be disposed timely to avoid the redissolution of pollutant. Different with sulfur-metabolizing bacteria, the total abundance of iron-metabolizing bacteria in the downstream was lower than that in the upstream mainly resulted from the reduction of Pseudomonas. Previous research had demonstrated that Pseudomonas can be suppressed rapidly by pyrite leachate within ve days (Han et al., 2013). The pyrite tailings leachate obviously increased iron-metabolizing bacteria abundance in the downstream water, e.g. Ferribacterium (increased by 61%), Sediminibacterium (increased by 52%), and Hyphomicrobium (increased by 11%), and the leachate was the source of Ferrimicrobium, Ferrithrix, Ferrovum, and Curvibacter (Fig. S4b). Overall, the pyrite tailings leachate introduced abundant sulfur-metabolizing bacteria and iron-metabolizing bacteria into the river passing through the pollutant source and further changed the distribution of related bacteria.
Based on Spearman correlation analysis, sulfur-metabolizing bacteria and iron-metabolizing bacteria showed close connection (Fig. 4c). The genus of sulfur-metabolizing bacteria had an intimate connection with each other especially in the high positive relationship of Sulfurospirillum, Desulfobulbus, Paludibacter, Desulfomicrobium, Sulfurovum, and Arcobacter. However, the internal correlation of ironmetabolizing bacteria was relatively weak. Sphingomonas (1.0%) played an important role in the relationship of sulfur-metabolizing bacteria and iron-metabolizing bacteria which connected nine genus of sulfur-metabolizing bacteria mainly with negative relationship. The most strong negative relationship (Spearman correlation index = -0.826) in Fig. 4c occurred in the genus of Bosea and Desulfoprunum while the most strong positive relationship (Spearman correlation index = +0.811) occurred in Arcobacter and Sulfurospirillum. On the whole, positive relationship was the mainstream (65.4%) in the network of sulfurmetabolizing bacteria and iron-metabolizing bacteria. Regardless of the upstream, pollutant sources or downstream, sulfur metabolism functions was widespread based on PICRUST2 (Fig. 4d). Iron oxidation functions mainly existed in pyrite tailings leachate (PL) while sul de oxidation functions mainly occurred in leachate treatment sediment (ST) analyzed by FAPROTAX (Fig. 4d). Therefore, the leachate treatment sediment should be transported from the scene timely to avoid the sul de oxidization process to generate more amount of sulfate.

Conclusions
The bacterial communities in the polluted area of pyrite tailings, including the water and sediment of upstream, pollutant source, and downstream were analyzed by in this study. Pyrite tailings leachate had  Bacteria characteristics at the phylum (a) and genus (b) level. Figure 3