Nowadays, environmental contamination is one of the most serious problems confronting modern human society. Most near-shore marine environments around the world are under growing threat (David and Obura, 2000). Heavy metal (HM) contamination in aquatic environments is a global environmental issue that has garnered increasing attention over the last few decades because of its negative impacts (Iksandar and Keeney, 1974; Allen-Gil et al., 1997; Kim et al., 2003a; Ma et al., 2013; Wang et al., 2013). Heavy metals in the aquatic ecosystems originate from natural activities, i.e., geogenic (erosion and weathering) and anthropogenic sources (Bai et al., 2011; Wang et al., 2015b; Lee et al., 2017; Yazidi et al., 2017; Barut et al., 2018). The major sources of HMs are anthropogenic activities, such as smelting, mining, agriculture, aquaculture, and domestic or industrial discharge. (Vareda et al., 2019). Once HMs enter the aquatic ecosystem, they can be bioaccumulated by marine species and biomagnified through the food web, posing a serious menace to humans and animal hygiene (Tian et al., 2020) or adsorbed by suspended particles, then heavily accumulated in sediments (Gümgüm et al., 1994; Yi et al., 2011) while only ions persist soluble in the water column. As a result, sediments appear to be the main repository and sink for heavy metals in comparison to water, and may also serve as sources of contamination. Hence, sediments are commonly employed as geo-markers for the management and detection of likely sources of contamination in aquatic habitats (Hakanson, 1980; Ahmad et al., 2010; Bastami et al., 2012; Wang et al., 2012; Alves et al., 2014; Tavakoly Sany et al., 2014). Due to their biotoxicity, widespread origins, persistence, non-degradability, and subsequent bio-accumulation, heavy metal pollution, particularly in sediments, is one of the greatest risks to the aquatic environment (Forstner and Wittmann, 1983; Barlas et al., 2005; Gargouri et al., 2011; Fu et al., 2014).

Additionally, coastal sediments are important habitats for a wide range of microbiota such as bacteria and archaea that mediate the uptake, stashing, release, and transport of heavy metals and play a pivotal role in the biogeochemical transformations of nutrients or pollutants. (Ul-Hasan et al., 2019; Liu and Yang, 2020; Custodio et al., 2021). Since, the pollution by heavy metals not only jeopardizes the health of the native people and animals, but is also noxious to almost all the microbiomes by interfering with the growth, shape, and metabolism and inhabiting fundamental cellular activities including the synthesis of proteins and the cell membrane integrity (Kandeler et al., 2000; Walker et al., 2000) leading to shifts in structure, abundance, and distribution pattern of sediment microbiota community (Gillan et al., 2005; Deng et al., 2020; Di Cesare et al., 2020). Due to the extreme sensitivity of these communities to environmental disturbances in their natural habitats, the structure and diversity of sediment-associated microbiota communities are increasingly explored.

High-throughput sequencing (HTS) technology is improving in rate, efficiency, and cost (Di Bella et al., 2013). Consequently, it is becoming increasingly prevalent to investigate the entire population of prokaryotes in many habitats. Metagenomic approaches are culture-independent molecular tools used to explore ambient microbiota, including activity, function, diversity, and direct DNA sequencing extracted from the environment (Thies, 2015; Liao et al., 2020; Liu et al., 2020). Currently, using HTS technology to analyze microbiota in heavy metal-contaminated sediments has received more attention (Wang et al., 2010; Li et al., 2020a; Li et al., 2020b). Although the environmental DNA metagenomic approach is used to analyze microbial communities in heavy metal-contaminated sediments, this approach is rarely used in Asia, especially in Taiwan. Therefore, in our study, the microbiota communities in metal-contaminated sediments were analyzed using the eDNA technique for the first time at the Siangshan wetland in Taiwan.

In the current study, we collected the surface and sub-surface sediments from the Siangshan wetland in Hsinchu city of Taiwan. Environmental variables comprising sediment grain size, total organic matter, and heavy metals content such as Fe, Al, Mn, Zn, Cu, Ni, Co, Cr, Ga, In, Cd, and Pb were analyzed. In parallel, the composition and diversity of the microbiota community as well as how the microbiota community structure responds to environmental variables were investigated using high-throughput 16S rRNA gene (V3-V4) next-generation sequencing. The main aims of this work were to: (1) ascertain the sediment composition, total organic matter (TOM) content, and the levels of Fe, Al, Mn, Zn, Cu, Ni, Co, Cr, Ga, In, Cd, and Pb, (2) evaluate the pollution degree of HMs using stander pollution indices such as sediment quality guidelines (SQGs), geo-accumulation index (I_geo), enrichment factor (EF), pollution index (PI), and pollution load index (PLI) (Ali et al., 2015; Xu et al., 2017; Salam et al., 2019), (3) profile the diversity and structure of the sediment-associated microbiota in Siangshan wetland, and (4) explore the most important environmental variables that influence the sediment microbiota community structure in Siangshan wetland. This research can help us understand the role of heavy metals, as well as the responses and potential adaptability of microbiota communities to different metal-contaminated habitats. This will upgrade the existing understanding of the ecological risks posed by HMs deposition in the Siangshan wetland and provide a useful benchmark for future metal-microbial interaction investigations.

In this work, we explored the diversity and structural shifts in sediment-associated prokaryotic communities to elucidate the potential impacts of environmental variables including grain size, organic matter, and HMs in shaping microbiota community structure.

Generally, the Siangshan wetland was characterized by sandy or muddy sediments, but the gravel sediments recorded limited percentages at all studied sites. Based on the results of a grain size measurement, sand granules predominated in KY1, KW2, HS1, HS2, and YK1, whereas mud sediments were most plentiful in KW1, HS3, YK2, and MA3. Furthermore, the OB, DJ, HM, and FC areas’ surface sediments contained a high percentage of sand particles > 50%. The flow rate and velocity have a considerable effect on grain size (Kim et al., 2017). In fact, variations in the river channel direction and riverbed slope caused by manmade structures are related to changes in flow velocity and temporary stoppage of flowing streams (Kim et al., 2003b). The content of TOM was different between the sites we looked at in our research, where TOM was observed high percentage in the KY1 site’s surface sediments (10.76%). It could be ascribed to the presence of freshwater seepage from the Ke-Ya river, which has a high composition rate of organic elements. Furthermore, MA sites (MA1, MA2, MA3) had relatively significant concentrations at two sediment levels. Mangrove forests are regarded as a very productive ecosystem (Odum and Heald, 1975), with significant rates of organic carbon storage (Huc, 1980).

We then determined the heavy metal concentrations (Fe, Al, Mn, Zn, Cu, Ni, Cr, Co, Ga, In, Cd, and Pb) in Siangshan wetland’s sediments, and comprehensively assessed the pollution degree using different indices like; Enrichment factor (EF), Geo-accumulation index (Igeo), Pollution index (PI), and Pollution load index (PLI). The results indicated that the surface sediments of the KY1 site recorded the highest values of Zn, Cu, and Ni, and the maximum concentrations of Fe, Co, and In at MA2. Moreover, DJ1 showed the highest values of Al, Cr, and Ga, and OB3 for Mn. Inversely, YK1 recorded the lowest concentrations for most metals; Zn, Cu, Ni, Co, Cr, and In. While the DJ3 site’s sub-surface sediment recorded the highest concentrations for Zn, Cu, and Ni, as well as Fe, Mn, Co, Cr, and In recorded significant values in DJ1; Al and Ga in MA2. On the other hand, the YK1 site recorded the lowest concentrations for Fe, Zn, Cu, Ni, Co, Cr, and In; while Ga was observed at the KY1 site. The accumulation of heavy metals at these sites is due to the presence of a terrigenous freshwater supply, untreated residential sewage, and a large proportion of fine sediments that deposit with high organic matter content, which tend to be associated with heavy metals (Irvine and Birch, 1998; Yu et al., 2012; El-Metwally et al., 2017). Additionally, Siangshan sites are characterized by sandy and muddy sediments, hence the heavy metals were accumulated in high content at these sites due to the vast surface region and the availability of organic materials (Reddy et al., 2004).

In comparison with the previous studies (Supplementary Table 11), the total average concentrations of Fe (30582.95 mg/kg), Al (37036.41 mg/kg), Mn (501.18 mg/kg), and Cr (80.31 mg/kg) in the studied sites were more than those of the previous studies, but they were lower than the average concentration in shale (Turekian and Wedepohl, 1961). Contrary, Cd and Pb (BDL) were less than those in all previous studies, upper limits for natural background concentration in Taiwan (Lai et al., 2018), and average shale value (Turekian and Wedepohl, 1961). The overall mean of Zn (113.41 mg/kg) was close to the concentrations at the Nakdong River in Korea (Kim et al., 2020) and Main River & wetlands in Hsinchu of Taiwan (Ye and Young, 2008), while it lower than the concentrations of (Lai et al., 2018; Ma et al., 2018), but higher than those of other sites and the average shale value (Turekian and Wedepohl, 1961). However, the average concentration of Cu (40.18 mg/kg) was lower than those of Love River (Ma et al., 2018), the upper limit of natural background concentration in Taiwan (Lai et al., 2018), and Shale value, but higher than other previous studies. Additionally, Ni (24.70 mg/kg) was relatively similar to coastal areas of Bohai and Yellow sea, China (Tian et al., 2020), and the results of (Ye and Young, 2008), but lower than Lake Bafa in Turkey (Algül and Beyhan, 2020), Sfax solar saltern in Tunisia (Bahloul et al., 2018), Siangshan wetland, Taiwan (Young, 2019), the upper limit in Taiwan, and Shale value. Furthermore, Co (100.54 mg/kg), Ga (95.65 mg/kg), and In (36.85 mg/kg) were shown more than all the previous studies and average Shale values.

To evaluate sediment quality, we compared the overall average of HMs in sediments to sediment quality guidelines (SQGs). This is the most traditional method (El Nemr et al., 2016) and is particularly successful in assessing HM pollution in sediments (Lin et al., 2016; Duodu et al., 2017). In general, the SQGs have lower and upper limits for the frequency and severity of ecological effects, respectively. The reference values that were implemented by the Australian and New Zealand Environment and Conservation Council and the Agriculture and Resource Management Council of Australia and New Zealand (Anzecc, 2000) are analogous to the reference values that were established by the National Oceanic and Atmospheric Administration, also known as NOAA (Long et al., 1995). Those developed by the Canadian Council of Ministers of the Environment (CCME, 2002) and the Taiwan EPA (Taiwan, 2010) are comparable. Generally, the former two guidelines have relatively higher concentrations than the last two, implying that the last two have more rigorous limits for the SQGs strategy. According to our data, the total average of Zn and Ni concentrations surpassed the Taiwan EPA lower limit, NOAA ERL, and ANZECC & ARMCANZ low levels, although Cu exceeded NOAA ERL and CCME ISQG. Furthermore, Cr exceeded the CCME ISQG and Taiwan EPA’s low limit. on the other hand, Cd and Pb were below all of the different referenced values. In conclusion, except for Pb and Cd, the sediment averages of heavy metal concentrations in our work surpassed the low reference values in the various SQGs, indicating these metals have infrequent or occasional adverse effects on the aquatic population (Sarkar, 2018).

SQGs have been frequently employed in the evaluation of HMs contamination in sediments, although they do not include the background values. To more precisely assess HMs contamination in sediment, background values (Average Shale Values) of sediments must be considered. Hence, various quantitative indices (EF, I_geo, PI, and PLI) have been employed to identify the pollution degree of HMs in sediments. Based on EF values, Al, Mn, Ni, Cr, Cd, and Pb at all studied sites showed minimally enrichment (> 2) suggestive of no or minimal contamination, indicating that they are mostly originating from lithogenic or terrigenous processes. While EF values for Zn, Cu, Co, and Ga have minimal (< 2) to significant pollution (< 20).

The I_geo values for elements viz. Fe, Al, Mn, Zn, Cu, Ni, and Cr were marked as unpolluted, while Co and Ga range between class 0 and class 1 (unpolluted to very slightly polluted).

PI values for Fe, Al, and Mn were subjected to class 0 means that the studied sites were unpolluted with these metals, while PI values of Zn, Cu, Ni, Co, and Cr range from low to high polluted in different studied sites. Overall, EF, I_geo, and PI of Indium (In) recorded high values in all studied sites, indicating that the Siangshan sediments were highly or extremely polluted with this metal, this may be originating from an anthropogenic effluent like industrial seepage from Hsinchu Science Industrial Park. Gallium (Ga) and Indium (In) are critical transition metals utilized in liquid crystal displays (LCD), photovoltaic cells, semiconductors, light-emitting diodes (LED), and light extraction processes that cause deleterious effects on people (Ram et al., 2015; Rocchetti et al., 2015; Wang et al., 2015a; Lee et al., 2016; White et al., 2017). They are discharged into the intertidal zone during the production process such as epitaxy, cleaning activities, and chip preparation in the manufacturing of high-speed semiconductors and light-emitting diodes (Yang and Chen, 2018). Foregs Geochemical Atlas (2006) reported that Gallium content in stream sediments ranged from 1-36 mg/kg (Atlas, 2006). Also, Indium and Gallium have been discovered in the groundwater under the semiconductor fabrication areas of northern Taiwan (Chen, 2006; Hsu et al., 2011).

The surface sediments of KW1, DJ1, DJ2, DJ3, OB2, OB3, KY2, and MA2 were strongly polluted with studied metals based on PLI (> 2), while KY1 was very strongly polluted (PLI > 3) and the remaining sites were moderately polluted (PLI < 2). Inversely, most sub-surface sediment sites were moderately polluted with metals (PLI < 2) but KW2, and YK1 were unpolluted sites (PLI < 1).

Microbial communities play a vital role in the management of the marine environment, and the abundance or composition of these communities is largely influenced by environmental variables in the ecosystem. We determined 14 environmental variables influencing the microbiota communities in the Siangshan wetland.

High-throughput analysis of the 16S rRNA gene of microbial communities in all surface and sub-surface sites revealed that Proteobacteria (including Gammaproteobacteria and Alphaproteobacteria) are the most diverse and dominant that have a significant function in sediment decomposition and breakdown (Huang et al., 2019). Despite the studied sites being polluted with heavy metals, Proteobacteria were dominated. Compared to other phyla, Proteobacteria were highly abundant at the metal-contaminated sites may be attributed to the adaptation of its lower taxonomic groups to environmental changes (Bouskill et al., 2010). Moreover, Proteobacteria members are widely known for their capacity to tolerate high levels of metals (Guo et al., 2017; Chen et al., 2018). Proteobacteria are reported in both HMs contaminated and uncontaminated regions (Ma et al., 2015; Yan et al., 2016). Previous studies documented the response of Proteobacteria to HMs contamination in polluted habitats (Hu et al., 2021; Liu et al., 2021b; Serrana and Watanabe, 2022). Bacteroidota (class Bacteroidia) and Cyanobacteria (class Cyanobacteria) were more abundant phyla in surface sediments of MA1.F with relative abundances of 23.91% and 26.40%, respectively. Similarly, Asgardarchaeota (class Lokiarchaeia) were a more enriched phylum in surface sediments of OB1.F (27.84%) than in sub-surface sediments. Inversely, Desulfobacterota (represented in 3 classes Desulfobacteria, Desulfobulbia, Desulfuromonadia) and Chloroflexota (including class Anaerolineae) were more abundant phyla in sub-surface sediments, especially in MA3.S and MA2.S sites, respectively. Other important phyla observed in this study are, Actinobacterota, Planctomycetota, Acidobacteriota, Gemmatimonadota, Myxococcota, and Halobacteriota (Archaea) as the most abundant. All the abundant microbial phyla obtained in this study have been documented previously in other metal-contaminated environments, such as sediments (Lee et al., 2021; Rajeev et al., 2021; Yi et al., 2021; Custodio et al., 2022), iron tailing pond (Li et al., 2021), mining soils (Cao et al., 2017; Hemmat-Jou et al., 2018; Liu et al., 2021a), paddy soils (Tseng et al., 2021), arid loess region (Zeng et al., 2020), and electric waste soils (Jiang et al., 2019), this suggests that these phyla are associated with metal-polluted sediment.

Generally, the investigation of microbiota community composition revealed that there was variance among these communities, with ecological considerations responsible for a sizable proportion of the variation. This finding is corroborated by 16S rDNA gene sequencing, which observed that these environmental contaminations are also associated with substantial increases in ASVs within the phyla Proteobacteria, Bacteroidota, Actinobacterota, Chloroflexota, Planctomycetota, Acidobacteriota, and Verrucomicrobia (Beattie et al., 2018).

Human activities may cause shifts in microbiota diversity and structure (Wu et al., 2017). Based on the calculated pollution load index, most of the studied sites are polluted with heavy metals. Of these sites, DJ1.F, MA1.S, DJ2.F, and DJ3.F were moderate to strong polluted with heavy metals. Interestingly, these sites had more richness (S_obs and Chao1 indices) and diversity (Shannon diversity index) than other sites. The well-developed adaptation capacity of microbial communities and the effectiveness of metal detoxification by exploiting sequestering and precipitation processes may be the causes of higher richness in polluted sites (Rajeev et al., 2021). A pertinent study found that polluted sites have more biodiversity than unpolluted sites (Tao et al., 2017). Prolonged exposure of microorganisms to heavy metal contamination has different impacts on microbiota diversity, and over time, microorganisms progressively build up resistance to metals (Xu and Tang, 2018). The ability of the microbial community to tolerate contaminants through detoxifying mechanisms may enable it to survive in contaminated environments (Bourceret et al., 2016). Nevertheless, other studies have revealed that polluted regions by HMs had markedly lower levels of diversity and the structure or composition of sedimentary microbiota was shifted (Hemme et al., 2010; Frossard et al., 2017; Huaidong et al., 2017; Feng et al., 2018). Here, statistical analysis revealed there were no discernible variations in the microbial diversity and richness of the surface and sub-surface sediments (P > 0.05).

At the same time, the nMDS ordination exhibited short distances and similarities between the surface sites, except KY1.F, KY2.F, MA2.F, and MA3.F. The results illustrated that the sediment-associated microbiota community compositions are significantly dissimilar among surface and sub-surface sediment groups (ANOSIM with 999 iterations, R = 0.2526, p = 0.01).

RDA analysis was performed to further explore the environmental variables that may contribute to microbial community variances between sediment sites. In the present study, Mn and Cu were identified as the most crucial factors (F = 8.07 and p = 0.001; F = 6.04 and p = 0.002, respectively), followed by Al, Co, Ni, Sand, and Cr influencing the microbial community structure. This is consistent with previous studies in the Monkey Head iron tailing pond (Li et al., 2021), contaminated sediments in India (Rajeev et al., 2021), and the arid loess region (Zeng et al., 2020). Nevertheless, earlier studies contradicted our findings since heavy metals in aquatic sediment have minimal or no association with the composition of the microbial community (Gillan et al., 2005; Gough et al., 2008). Noteworthy it is critical to clarify that the majority of these works relied on traditional molecular methods like direct cell counts or denaturing gradient gel electrophoresis (DGGE) that primarily identify dominant bacterial communities, as result, have the potential to underestimate or misunderstand the structure of the community (Casamayor et al., 2000). Other studies have shown that the effluences of heavy metals on the abundance, structure, and composition of microbiota communities involve several fundamental mechanisms including (i) disrupting replication processes by destroying DNA and protein (Yan et al., 2020; Coclet et al., 2021) (ii) produce harmful reactive oxygen species by acting as redox catalysts, (iii) disrupting enzymatic processes, and (iv) disturbance of the membrane flexibility and ion regulation (Prabhakaran et al., 2016; Igiri et al., 2018). Indeed, the impacts of heavy metals on the communities of sedimentary microbiota are multifaceted. Initially, only bioavailable metals pose a threat to living things (Kunito et al., 1999). Secondly, the level of exposure determines how heavy metals affect microbial populations. Metals either cause a noticeable shift in microbial population and diversity in a short period of time (1 year) (Bouskill et al., 2010) or the diversity of the community remains steady after around a century of pollution, during which susceptible communities dwindle and the impervious population prevails (Reis et al., 2013).

It is worth noting that the current study only analyzed the structural diversity of the sediment-associated microbiota communities; thus, future research should identify the specific functional genes and associated biochemical functions in the communities of microbiota using PICRUSt or Tax4Fun functional predictions and Biolog analyses.

This is the first study that utilized eDNA metagenomic analysis to examine the microbial diversity in sediments of Taiwan’s Siangshan wetland. The obtained findings concluded that EF, Igeo, and PI of Indium (In) recorded high values in all studied sites, indicating that the Siangshan sediments were highly or extremely polluted with this metal, this may be originating from an anthropogenic effluent like industrial seepage of Hsinchu Science Industrial Park. A comparison of the metal concentrations with the SQGs of Taiwan, USA, Australia & New Zealand, and Canada revealed that pollution surpassed the low reference values indicating these metals have infrequent or occasional adverse effects on the aquatic population. Moreover, we applied the eDNA metagenomic approach to investigate the structure and composition of microbiota community. Study sites were dominated largely by Proteobacteria, followed by Bacteroidota, Desulfobacterota, Actinobacterota, Cyanobacteria, Chloroflexota, Planctomycetota, and Firmicutes. NGS sequencing results and alpha diversity indices demonstrated that the diversity of contaminated sites such as DJ1.F, MA1.S, DJ2.F, and DJ3.F was significantly higher than those of uncontaminated sites. The RDA analysis explored that Mn, Cu, Al, Co, Ni, Sand, and Cr had a sizable effect on the structure of microbiota communities (at the class level). Our findings provide a positive glimpse into the potential of eDNA metagenomic to define sediment-associated microbial communities and further the understanding of community structure and composition.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi.nlm.nih.gov/- PRJNA836729.

AS-T conceptualized the project, gathered samples, performed methodology, bioinformatics & statistical analyses, and manuscript writing. C-SC participated in the fieldwork, laboratory equipment setup, and sample sequencing. M-YL reviewed and revised the manuscript. S-SY conceptualized the study, sample gathering, review and revision of the article, supervision, and funding. All the authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.