Groundwater contamination, by persistent, bioaccumulative and bioactive compounds due to intensive industrialization is a critical environmental issue [1
]. The wide range of chemical compounds complicates the attempts to protect aquifers from their contamination [4
]. In addition to pollution from industry, agriculture, and the energy sector, personal care products, and pharmaceutical residues are emerging organic contaminants (EOC) in groundwater [6
]. In a comprehensive review by Lapworth et al. (2012), EOCs, such as detergents, antioxidants, fire retardants, plasticizers, as well as antibiotics, anti-inflammatories and barbiturates, were detected in substantial levels between 10–104
ng/L in groundwater globally [7
]. One of the largest groups of chemicals that cause environmental contamination are halogenated organic compounds [8
]. Due to their high common resistance to degradation, they often accumulate in the environment, and their fate depends highly upon microbial degradation [6
The microbial community structure in the ecosystem is critical for the biodegradation of pollutants to occur. In pristine aquifers, microbial diversity is explained by water chemistry, and therefore changes in geochemical conditions and contaminant content will affect the community structure [11
]. Bacteria adaptation accelerates the biodegradation of organic chemicals, where shifts in community composition or abundances are one of the adaptive processes. For example, in a pilotscale aquifer, Ma et al. [14
] presented clear shifts in the microbial community exposed to a chlorinated ethanol fuel plume. Interestingly, after community shifts led to biodegradation of the contaminant and a geochemical restoration of the aquifer, the microbial community structure was also restored. Long monitoring of a coal tar waste-contaminated groundwater ecosystem revealed that natural attenuation processes are accompanied by temporal changes in the microbial communities [15
Nonetheless, the existence of a microbial potential in the ecosystem does not always lead to in situ biodegradation [11
]. Multiple limitations, such as insufficient biomass, geochemical and hydraulic fluctuations, utilization of a wide range of substrates, competitive inhibition, or catabolite repression, can all inhibit biodegradation [16
]. Therefore, integrated approaches, such as geochemical-molecular or isotopic-molecular approaches are used for the actual monitoring of biodegradation in natural environments. For example, Tarnawski et al. [17
] followed the reductive dechlorination of chlorinated-ethenes in contaminated aquifers and presented an integrated long-term geochemical-microbial monitoring approach, in order to identify biogeochemical processes limiting or supporting biodegradation. Révész et al. [18
], on the other hand, quantified the relevant bacterial groups along with in situ trichloroethylene (TCE) anaerobic biodegradation, while monitoring degradation through the use of stable carbon isotopes.
The Neot-Hovav industrial complex accommodates numerous chemical industries in the pharmaceutical, bromine-based organic compounds and agrochemicals. The complex is situated above the fractured Eocene chalk formation [19
]. The fractures and joints in the chalk form an aquitard with preferable pathways for water flow and solute migration, thus transmitting contaminants from the surface to groundwater at relatively high fluxes in a short time [20
]. For this reason, and due to inadequate treatment of industrial waste, the groundwater underlying the industrial area has been contaminated by numerous organic compounds: Volatile and non-volatile halogenated organic compounds, heavy metals, and halogen-containing salts [4
In contrast to porous aquifers, the hydraulic heterogeneity in a fractured system is large, and strongly limits particle transport (like bacteria), solutes including electron donors and acceptors, as well as nutrients along the hydrologic gradients [24
]. Thus, the effect of hydrologic discontinuity on microbial community structure and catabolic potential is still an open question. Furthermore, in polluted groundwater like Neot-Hovav, where hundreds of different compounds are present, a unique approach to the aggregated effect of inorganic and organic compounds upon microbial community composition and function should be used. In turn, the diversity and richness of the groundwater microbiota, as well as their metabolic potential, can provide a broader perspective on the “health” of the groundwater and the possibility for natural recovery or the need for active remediation management.
In this study, we examine how the extent and type of contamination above ground determine and affect the microbial community structure in the groundwater. For this cause, we investigated spatial changes in the microbial community of the polluted groundwater that underlies the Neot-Hovav industrial complex in relation to the industrial sources of specific halogenated pollutants. Changes in communal structure as a factor of groundwater quality were examined while providing a more sustainable approach to the understanding of these multi-contaminant, complex environments. Microbial activity of the groundwater bacterial communities was also tested through a series of biodegradation microcosm experiments in the lab with various brominated organic compounds like TBNPA-tribromoneopentyl alcohol, DBNPG-dibromoneopentyl glycol, BCE-bromo-chloro-ethane, EDB ethylene dibromide, and EDC ethylene dichloride. These compounds were chosen, as they are part of the wide contaminant palette that exist in the groundwater, and are produced and used in the industrial area.
2. Materials and Methods
Tribromoneopentyl alcohol (TBNPA, >98% pure) was obtained from TCI chemicals, Japan; dibromoneopentyl glycol (DBNPG) from Israel Chemicals Ltd. (ICL Israel), 1-bromo-1-chloroethane (BCE), 1,2-Dibromoethane (EDB), and, 1,2-dichloroethane (EDC) (99.8% pure) were obtained from Sigma Aldrich (Rehovot, Israel). Methanol, acetone, di-propanol and acetonitrile were all HPLC grade (Supra-gradient, Bio-Lab, Israel).
2.2. Study Area and Groundwater Sampling
The Neot Hovav industrial site (180,063, 560,798 Israel Transverse Mercator), hosts various chemical industrial plants that overlie an extensively fractured chalk aquitard in which the fractures were shown to play a major role in infiltration, groundwater flow and solute transport [20
]. The groundwater underlying the industrial site is characterized by a slightly acidic to slightly alkaline pH range (6.0–8.5), high electrical conductivity (EC) (14.1–55.1 mS/cm), low dissolved oxygen (DO) concentrations (0.22–0.84 mg/L), and low metal concentrations (<5 µg/L).
In 2015, 17 boreholes were sampled for microbial biodiversity and dissolved organic carbon (DOC), with a spatial distribution from the production plant [20
], and up to 3.5 km downstream of the industrial area. In spring 2016 the RH49 pumping well (RH49P) was sampled for establishing microcosms for TBNPA and DBNPG. In February 2019, RH49P was sampled for establishing microcosms BCE, EDB, and EDC degradation.
Based on the land uses in the site, the industrial area was divided into three parts: (1) Adjacent to the brominated organic compounds (BOC’s) production plant (groundwater depth: 30–50 m below ground). (2) Along Wadi Hovav—a dry riverbed that drains the erratic runoff in the industrial area (groundwater depth: 2–10 m below ground). (3) Downstream to the production plant, up to ca. 3.5 km from the plant (groundwater depth: 30–40 m below ground) (Figure 1
). The area that is downstream to the production plant hosted in the past various wastewater treatment procedures that released along it the year’s contaminants to the sub-surface environment. For example, forced evaporation of effluents by sprinklers was practiced over hillslope located next to observation borehole KN4 until the late 1980s [26
], and unlined wastewater evaporation ponds were used adjacent to borehole KN201 until 1982 [27
]. The locations of the sampled boreholes are summarized in Table 1
]. See SI for a detailed description of groundwater sampling procedures.
2.3. DNA Extraction, PCR Amplification and Sequencing
Seventeen groundwater samples from September 2015 were extracted for total genomic DNA after filtration through 0.22 µm sterilized filters. DNA was extracted by using PowerSoil®
kits (MoBio) (Carlsbad, CA, USA), with slight modifications to the manufacturer’s instructions: The 0.22 µm sterilized filters were added to the PowerBead Tubes instead of 0.25 g soil. In the elution, 100 µL of solution C6 was added in two steps (twice 50 µL), and centrifuged at 10,000× g
for 30 s in-between. The 16S rRNA gene was amplified by the polymerase chain reaction (PCR) with the primers CS1-341F-CS2-806R for Bacteria and CS1_Arch344F-CS2_Arc806R for Archaea [28
]. The thermocycler conditions for DNA amplification were: 5 min at 95 °C, followed by 26 cycles of 95 °C (30 s) → 55 °C (45 s) → 68 °C (30 s) and finalizing the reaction with 68 °C (7 min); a set of PCR products was run and visualized on agarose gel (2%) to verify product specificity. The DNA concentration was determined using a NanoDrop 1000 spectrophotometer (Thermo Scientific, Wilmington, NC, USA).
PCR products were sent to the DNA Services Facility, Research Resources Center, University of Illinois at Chicago (UIC), for sequencing of 16S rRNA using the Illumina MiSeq platform. Operational taxonomic units (OTUs) were assigned according to the company’s pipeline: Through the NGS analysis pipeline of the SILVA rRNA gene database project (SILVAngs 1.3) [29
]. BLAST (version 2.2.30+) in combination with SILVA SSU or LSU Ref datasets were used for theclassification of reference sequences [30
2.4. Data Analysis of Sequencing Results
Statistical analysis of the sequencing data was implemented using various R packages (version 3.3.1) [31
]. The alpha biodiversity of the microbial community was estimated for each sample using the abundance-based diversity indices Shannon and Chao1, through the plot_richness function of the phyloseq package in R (version 1.14.0) [32
]. Significant differences between diversity indices to the sample location were tested by the Kruskal-Wallis Analysis of Variance (ANOVA) by ranks test (STATISTICA software (version 10). Weighted UniFrac distances of the community data was calculated by the UniFracfunction of the phyloseq package, and the Bray-Curtis distance matrix by the bray function of the phyloseq and vegan (version 2.4.0) packages [33
]. Nonmetric multidimensional scaling (NMDS) OTU-based community composition, based on Bray-Curtis and weighted UniFrac distances was conducted by the ordinate function in the phyloseq package, and by the metaMDS function of the vegan package. Significant differences in the microbial community composition between samples were tested by Permutational multivariate ANOVA (PERMANOVA) using the adonis function of the vegan package. Envfit function in the vegan package was applied to illustrate the correlations between environmental parameters and the microbial communities. Microbial phylotypes with their corresponding read counts were imported into METAGENassist [34
] to analyze the metabolic features of the microbial communities.
2.5. Microcosm Experiments
The biodegradation potential of groundwater bacteria to degrade TBNPA, DBNPG, BCE, EDB, and EDC was tested in a 100 mL anaerobic groundwater microcosm batch experiment. In all experiments, a RH49P groundwater sample (two liters in a sterile polypropylene bottle) was retrieved from the continuously pumping RH49 borehole (Figure 1
, RH49P) [35
]. The TBNPA; DBNPG (50 mg/L)/or both TBNPA and DBNPG (25 mg/L each) experiment included three different supplied electron donors: H2
(1.7–3.8 mL), lactate (0.4 g/L), or acetate (6.8 g/L) to encourage degradation (i.e., each electron donor treatment was given to three bottles that contained 1-TBNBA; 2-DBNPG; and 3-TBNPA+DBNPG), and was done under anaerobic conditions. Three negative controls contained sterile (autoclaved) groundwater with TBNPA (50 mg/L); DBNPG (50 mg/L); or both TBNPA and DBNPG (25 mg/L each). The microcosms were sampled anaerobically at the beginning at a period of weeks, then months, and finally after three years. For BCE, EDB, and EDC (1 mM) microcosms, water was sampled from RH49P in February 2019 and lactate was added to all bottles (0.39 gr/L) (in duplicates). One bottle was autoclaved and served as a control. Samples were taken anaerobically every week. All bottles were held at room temperature and were constantly shaken.
2.6. Analytical Methods
TBNPA and DBNPG concentrations were determined by high performance liquid chromatography (HPLC; Agilent 1100 series, Palo Alto, CA, USA). BCE, EDB and EDC concentrations were determined by Gas Chromatograph-Mass Spectrometer (6890–5975 Agilent Palo Alto, CA, USA. Samples were automatically diluted 100 times, and 5 mL of the solution was purged for 11 min (He flow 40 mL/min). Desorption was performed at 190 °C for 4 min. For GC separation a DB-5 capillary column (30 m × 0.25 mm id × 0.25 μm) was used. The following conditions were applied: Split injection mode 20:1, injector temp. 220 °C; oven heating from 50 °C to 200 °C at a rate of 10 °C/min; He flow was 1 mL/min. The concentration was compared to chloroform as an internal standard. DOC was quantified by a Multi N/C 21005 (Analytic Jena, Jena, Germany).
This study exhibits the importance of combining multiple techniques when addressing the natural bioremediation potential of a microbial community. When dealing with a site with multi-contaminants, it is crucial to test various pollutants’ fate, to gain a better understanding of this potential. In addition, this study exemplifies how the characteristics of the groundwater matrix govern the pollutant distribution, microbial community structure, and perhaps the natural bioremediation potential.
The complex microbial community that lies in the polluted groundwater within a chalk aquitard that underlies the Neot Hovav industrial site was presented. Hydrologic discontinuity in the aquitard results in spatial heterogeneity of hydro-chemical conditions in the site, leading to variations in microbial community structure. The connection between contaminant loads to bacterial community structure shows that the contamination differences in the distinct groundwater section under separate sections of the industrial site lead to the variation in the microbial community. High contaminant load leads to lower communal diversity and the presence of more tolerant bacteria. When contamination decreases, diversity increases, with pollutant mixture seemingly affecting community structure. Anaerobic de-halogenating activity and potential of the groundwater microbial community from the production plant area were exhibited. While three of the tested compounds were degraded, two were not, thus exemplifying the complication of remediating multi-contaminant locations.