Sequential biodegradation of 1,2,4-trichlorobenzene at oxic-anoxic groundwater interfaces in model laboratory columns

Abstract

Halogenated organic solvents such as chlorobenzenes (CBs) are frequent groundwater contaminants due to legacy spills. When contaminated anaerobic groundwater discharges into surface water through wetlands and other transition zones, aeration can occur from various physical and biological processes at shallow depths, resulting in oxic-anoxic interfaces (OAIs). This study investigated the potential for 1,2,4-trichlorobenzene (1,2,4-TCB) biodegradation at OAIs. A novel upflow column system was developed to create stable anaerobic and aerobic zones, simulating a natural groundwater OAI. Two columns containing (1) sand and (2) a mixture of wetland sediment and sand were operated continuously for 295 days with varied doses of 0.14–1.4 mM sodium lactate (NaLac) as a model electron donor. Both column matrices supported anaerobic reductive dechlorination and aerobic degradation of 1,2,4-TCB spatially separated between anaerobic and aerobic zones. Reductive dechlorination produced a mixture of di- and monochlorobenzene daughter products, with estimated zero-order dechlorination rates up to 31.3 μM/hr. Aerobic CB degradation, limited by available dissolved oxygen, occurred for 1,2,4-TCB and all dechlorinated daughter products. Initial reductive dechlorination did not enhance the overall observed extent or rate of subsequent aerobic CB degradation. Increasing NaLac dose increased the extent of reductive dechlorination, but suppressed aerobic CB degradation at 1.4 mM NaLac due to increased oxygen demand. 16S-rRNA sequencing of biofilm microbial communities revealed strong stratification of functional anaerobic and aerobic organisms between redox zones including the sole putative reductive dechlorinator detected in the columns, Dehalobacter. The sediment mixture column supported enhanced reductive dechlorination compared to the sand column at all tested NaLac doses and growth of Dehalobacter populations up to 4.1×10⁸ copies/g (51% relative abundance), highlighting the potential benefit of sediments in reductive dechlorination processes. Results from these model systems suggest both substantial anaerobic and aerobic CB degradation can co-occur along the OAI at contaminated sites where bioavailable electron donors and oxygen are both present.

Graphical Abstract

1. Introduction

Chlorobenzenes (CBs) are aromatic chlorinated organic solvents historically produced and used in industry. Exposure at low concentrations has been associated with negative human health effects including oxidative stress, respiratory inflammation, and possible carcinogenesis (Feltens et al. 2010). Due to spills and improper disposal at manufacturing and end-use sites, CBs in the subsurface often become persistent sources of legacy groundwater contamination, occurring as dense nonaqueous phase liquids (DNAPL) and resulting in high (mg/L) aqueous CB concentrations. Persistent DNAPL source zones can result in large dissolved aqueous plumes where contaminant concentrations are often in excess of 1% aqueous solubility (Feenstra et al. 1991). At least 491 EPA National Priorities List sites have documented CB contamination, most frequently mono- (MCB), di- (DCB), tri- (TCB), and hexachlorobenzene congeners (Agency for Toxic Substances Disease Registry 2017).

Biodegradation, either through natural attenuation or engineered in situ or ex situ treatment (bioremediation), is a significant environmental fate process for many organic contaminants, including CBs (Perelo 2010). Bioremediation is an appealing treatment approach for CBs due to its relatively low cost, minimal site impact, and potential to be self-sustaining (Boopathy 2000, National Research Council 2013). Under anaerobic conditions, CBs can be reductively dechlorinated to less chlorinated congeners by strict organohalide respiring bacteria (OHRB) in genera Dehalococcoides (Hölscher et al. 2003), Dehalobacter (Nelson et al. 2011), Dehalobium (Wu et al. 2002), and Dehalogenimonas (Qiao et al. 2018). Although complete anaerobic mineralization of CBs through benzene metabolism has been demonstrated in laboratory microcosms and in trace amounts in situ (Liang et al. 2013, Nijenhuis et al. 2007, Qiao et al. 2017), dechlorination of MCB to benzene is a known rate-limiting bottleneck in the degradation process (Fung et al. 2009). Anaerobic bioremediation of CBs, in general, has been avoided in field-scale applications due, in part, to concerns regarding possible inadvertent accumulation of MCB and benzene in groundwater. CBs with four or less chlorines and benzene, however, are particularly amenable to aerobic microbial mineralization, where no toxic daughter products are produced (Field and Sierra-Alvarez 2007). Aerobic CB degradation has been observed primarily in phylum Proteobacteria, including order Burkholderiales and genera Pseudomonas, Sphingomonas, and Xanthobacter (Field and Sierra-Alvarez 2007). Aerobic bioremediation is often employed for treatment of less-chlorinated CB congeners; however, the costs and difficulty of oxygenating groundwater can limit treatment effectiveness (Boopathy 2000, Qiao et al. 2017).

Groundwater discharged to surface water through wetland sediments and hyporheic zones often encounters redox gradients, which support diverse microbial communities and activity (Brune et al. 2000, Brunke and Gonser 1997). In these environments, organic carbon turnover can create reducing conditions in subsurface porewater while surface processes such as bioturbation, plant root oxygenation, and photosynthesis contribute to aerobic conditions (Braeckevelt et al. 2007, Brune et al. 2000). The oxic-anoxic interfaces (OAIs) formed in such environments offer a niche to potentially support both anaerobic and aerobic CB degradation pathways. The extent of such interfaces, and hence degradation, can vary considerably depending on the heterogenous characteristics of field sites such as organic carbon loading, competing electron acceptors, and mass transfer limitations (Brunke and Gonser 1997). Contaminant degradation has been described across a variety of interfacial scales in groundwater, from steep millimeter-length anoxic boundaries (Atashgahi et al. 2013, Kurt et al. 2012) to gradual microaerophilic permeation into deeper anaerobic zones (Burns et al. 2013, Gossett 2010).

At the Standard Chlorine of Delaware (SCD) Superfund site (New Castle, DE), over 1 million liters of mixed MCBs, DCBs, and TCBs were released into the environment from spill events (US EPA 2016). A previous site assessment conducted by the US Geological Survey found that subsurface CB DNAPL had contaminated the wetland adjacent to the industrial site, with mixed CB concentrations in excess of 50 mg/L and 1 g/kg in porewater and sediments respectively (Lorah et al. 2014). In situ Bio-Trap® microcosms as well as laboratory bioreactor experiments suggested that both reductive dechlorination and oxidative CB degradation pathways co-occur naturally in shallow SCD wetland sediments (Lorah et al. 2014). This and other studies (Braeckevelt et al. 2007, Burns et al. 2013) have provided empirical evidence of coupled degradation at OAIs through indicators such as ¹³C stable isotope fractionation, genetic markers, and shifts in congener distribution. However, the rates and extents to which these CB degradation pathways may occur at these interfaces remain poorly understood.

In this study, we aimed to quantify the biodegradation potential of CBs discharged through a model OAI within an upflow laboratory column system. Here, we investigated the dynamics of labile carbon amendment on the anaerobic and aerobic degradation of 1,2,4-TCB. Over a 295-day period, we varied the concentration of electron donor and measured CB degradation in site-relevant wetland sediment and idealized sand matrices. Additionally, we surveyed the spatial stratification of degradation activity and biofilm microbial communities within the columns to better understand the distribution of microbial taxonomy and functional potential across the interface.

2. Materials and Methods

2.1. Column Design and Setup

Modified 15 cm × 2.5 cm (L × ID) glass chromatography columns (Kimble Chromaflex®) were operated as upflow packed bed reactors to simulate groundwater discharge into surface water with a discrete OAI (Figure 1). A side-stream of oxygen-saturated Milli-Q® water was injected into each column at 67% of the nominal length to create aerobic conditions in the upper zone. Sample ports at 3.75 cm, 7.5 cm, and 11.25 cm from the inlet (25%, 50%, and 75% of column length, respectively) were added to vertically profile the interface. Caps, fittings, and tubing were constructed of fluoropolymers to minimize sorptive losses.

Figure 1.

Design schematic of experimental column systems. Samples were taken at positions designated 0, 25, 50, or 100% of nominal flowpath. At the end of the experiment, the columns were vertically segmented into eight equal segments (1–8) to characterize the column biofilm. Segments 1–5 represent the anaerobic zone and segments 6–8 represent the aerobic zone after oxygenated-side-stream addition.

Two model matrices were tested in parallel. The “Sand and Sediment Column” (SSC) was designed to reflect conditions similar to the SCD wetland, using sterilized sediment sampled from the site in 2016 (Lorah et al. 2014). The sediment, sieved to less than 1 mm grain size, had a total organic carbon content of 33.9 mg/g and a silt loam texture consisting of 3.1% sand, 76.4% silt, and 20.5% clay. Due to the high silt and clay content, SSC was packed with a mixture of sediment and autoclaved silica filter sand (0.55–0.65 μm) in a 75:25 wt% (84:16 dry wt%) ratio to facilitate more uniform flow characteristics. The “Sand Column” (SC) was designed to test microbial colonization and biodegradation potential on an inert matrix without the influence of site sediment. SC was packed solely with the filter sand described above. More detailed column design and packing details are provided in the Supporting Information (SI, Section S1.2).

A defined simulated anaerobic groundwater media was used to support microbial growth and avoid nutrient limitations during the experiment, as described in Section S1.3. Briefly, influent media were fed from two separate stocks containing (1) dissolved 1,2,4-TCB (from neat stock) and sodium lactate (NaLac) and (2) mineral salts, replenished every 20–30 days. Each stock was prepared from autoclaved water purged with N₂ to remove dissolved oxygen (DO). Media stocks were transferred into zero-headspace feed bags to minimize re-oxygenation, volatilization, and sorption. Media entering the columns contained 9.1 μM Cl⁻, 4.5 μM Fe, 0.26 mM SO₄²⁻, 1.5 mM total N (as NH₄⁺), and 1.27 mM total P (as phosphate buffer and counter-anions); a trace metal solution (Zeyer and Kearney 1982); and a vitamin mixture (Fathepure and Vogel 1991). Chloride salts were minimized to reduce background signal to allow detection of chloride released from biodegradation.

Media were delivered to the columns continuously by a multi-channel peristaltic pump (Ismatec IP 12) to achieve linear porewater velocities similar to the maximum of 3.0–4.5 cm/hr observed in the SCD wetland (Lorah et al. 2014). The anaerobic influent was diluted by a factor of 1.30 by the aerobic stream, and reported results accounted for this dilution. Column hydraulics were determined mid-experiment (Day 155) by a bromide tracer test. A step-input of 25 mg/L bromide was added to the column influent with effluent samples collected in discrete 20-minute intervals and analyzed by ion chromatography. Bromide breakthrough was fitted to a 1-D advection dispersion model with a semi-infinite boundary condition using tracer interpretation application TRAC to determine column hydrodynamics for the entirety of each column (Gutierrez et al. 2013). Hydraulic residence times and linear velocities were then back-calculated separately for anaerobic and aerobic zones assuming a constant effective porosity throughout the entirety of each column and proportionality of linear velocities to the flow rates in each zone. Calculated hydraulic parameters for each column are described in Table 1.

Table 1.

Hydraulic characteristics of experimental columns

Parameter	Sand + Sediment Column (SSC)	Sand Column (SC)
Effective porositya	0.34	0.48
Effluent flow (mL/hr)b	7.1 ± .1	6.8 ± .1
Linear velocity (cm/hr)a, c	3.3 / 4.3	2.3 / 2.9
Hydraulic residence time (hr)a, c	3.0 / 1.2 [4.2]	4.4 / 1.7 [6.1]

^aCalculated from hydraulic tracer test on Day 155

^bDirectly measured over course of experiment ±1 SD (n=5)

^cAnaerobic zone / Aerobic zone [Total column]

2.2. Column Inoculation

SSC was directly inoculated with both anaerobic and aerobic CB-degrading cultures under no-flow conditions to optimally stratify degraders across the OAI. The anaerobic inoculum was a 1,2,4-TCB-fed subculture derived from WBC-2 (SiREM Labs, Guelph, ON); WBC-2 was previously used to degrade CBs in field and laboratory tests at SCD (Lorah et al. 2014). The aerobic inoculum was a CB-fed culture enriched from filtered SCD wetland porewater fed with a mixture of MCB, 1,2-DCB, and 1,2,4-TCB as sole substrates. First, 40 mL of the anaerobic culture was injected by sterile syringe through the column influent port, displacing porewater media through the effluent. Ports were closed, and inoculation occurred statically for 4 days. Next, 20 mL of the aerobic culture was injected through the 50% sample port, with excess media displaced through the effluent. This inoculated the upper half of the column only, which occurred statically for 3 days. Media were then pumped continuously (Day 0) for 46 days in a startup phase to facilitate flushing of sediment-bound constituents. Detailed inoculum culture characteristics are presented in the SI (Section S1.4).

SC was inoculated passively by SSC effluent to simulate natural colonization of a clean matrix from upgradient microbial communities and determine whether coupled anaerobic- aerobic degradation could be established across the OAI in the absence of the optimized inoculation used in SSC. After flushing with media for the first 43 days, the SSC effluent was connected to the SC influent to provide a continuous flow-through inoculation of the full column for 3 days. On Day 46, SC was connected to a shared influent media bag with SSC and began independent operation. This marked the end of startup and beginning of operational Phase 1.

2.3. Experiment Operation

After start-up, columns were operated continuously at room temperature (18.0 °C ±3.1 °C SD) for 249 days. All data from column operation are available at Chow et al., 2020 and https://jscholarship.library.jhu.edu/handle/1774.2/62273. Influent 1,2,4-TCB concentrations varied slightly between batches, with an average value of 5.8±1.6 mg/L (32±8.8 μM). Aerobic zone DO concentrations after oxygenated side-stream injection were 7.0 ±0.9 mg/L (220 ±30 μM). NaLac, a readily fermentable source of labile carbon, was used as a model electron donor for the system. Column operation was divided into five discrete experimental phases where influent NaLac was cycled between (1) 0.28, (2) 1.4, (3) 0.14, (4) 14.4, and (5) 0.28 mM. This represented a dissolved organic carbon (DOC) range of 5–50 mg/L, similar to the range of non-volatile DOC observed at the SCD site (Lorah et al. 2014). A short phase (6) removed oxygen from the side-stream temporarily (by purging with N₂ gas) to verify oxygen dependence of CB biodegradation in the aerobic zone, while the final phase (7) restored the oxygenated side-stream and before sacrificial sampling. A graphical summary of operational phases is presented in Figure 2a.

Figure 2.

Time-series plots of experimental input phases and CB biodegradation for each column (A) Experimental input phases summary. Connected data points show influent 1,2,4-TCB concentrations measured at influent port (0% length). Target NaLac amendments annotated above. (B) Total anaerobic reductive dechlorination activity. (C) Aerobic degradation activity, represented as total CB removal. Bolded data points were samples used to calculate stable performance for each experimental phase.

Column samples were collected at three to four-day intervals at the influent (0%) and effluent (100%) ports. Less frequently, porewater samples were also collected to observe spatial trends along the column length. Porewater samples from ports at 0%, 25%, and 50% length represented the anaerobic zone of the column before oxygen addition, and samples at 75% and 100% length represented the aerobic zone of the column after the side-stream addition (Figure 1). However, because tracer measurements showed that complete mixing after the side-stream addition had not yet occurred at 75% column length, only the analytical data from the effluent (100%) port are discussed here as representative of the aerobic zone. Samples were collected in 5 mL glass Luer Lock syringes attached to sample ports. Immediately after collection, 100 μL subsamples were taken from the syringe barrel using a gastight microsyringe and analyzed for CBs (Section 2.4 below). The remaining sample volumes were passed through sterile 0.22 μm polyether sulfone syringe filters and stored in sterile polypropylene tubes for subsequent analysis.

At the end of the experiment (Day 295), columns were sacrificially sampled to analyze their biofilm microbial communities. The columns were divided into eight equal-length segments along the vertical profile, labeled 1–8 from entrance to exit (Figure 1). Column matrices from each segment were manually homogenized with a sterile cell scraper and stored in sterile 50 mL polypropylene tubes at −80 °C for microbial analyses.

2.4. Analytical Methods

Porewater CB concentrations were analyzed by gas chromatography-mass spectrometry (GC-MS; Agilent 7090B–5977). A single-step, miniaturized “in-vial” liquid-liquid extraction process was used to extract and analyze CBs directly in a single 1.5 mL autosample vial. CBs were quantitated with a calibration range of 0.05–10 mg/L. Benzene was analyzed separately by GC-MS using a purge and trap method described previously (Lorah et al. 2014). Ion chromatography (Thermo Fisher Dionex ICS-2100) was used to measure inorganic anions (chloride, sulfate, and bromide) with a calibration range of 0.1–20 mg/L. Detailed descriptions of analyte extraction and chromatographic protocols are presented in the SI (Sections S1.5–1.6).

A fiber-optic luminescence-based meter (PreSens OXY1-SMA; Regensburg, Germany) and an in-line sensor (PreSens FTC-PSt3) were used to measure DO concentrations at the oxygenated side-stream injection port. Column flowrates were determined gravimetrically by collection of effluent over hour-long time intervals.

2.5. Biodegradation Calculations

The overall biodegradation activity for the columns at each timepoint was estimated on the basis of reductive dechlorination and aerobic degradation using measured concentration profiles of influent and effluent CBs (Figure S1).

$$\text{Reductive Dechlorination}=\frac{\left(2\times {\text{MCB}}_{100}+\Sigma {\text{DCB}}_{100}\right)}{\Sigma {\text{CB}}_{100}}\times \Sigma {\text{CB}}_0$$ Equation 1

Transformation of 1,2,4-TCB by reductive dechlorination in the anaerobic zone was estimated as the total number of C-Cl bonds cleaved (dechlorinations) from the influent to produce the observed profile of daughter congeners in the effluent. This assumed that MCB was formed by 2 dechlorinations and DCB by 1, that dechlorination stalled at MCB, and that the difference in total influent and effluent CB concentrations were due to aerobic degradation (Section S3b). Dechlorinations were calculated according to Equation 1, where the degree to which total CBs in the effluent (ΣCB₁₀₀) were dechlorinated to MCB (MCB₁₀₀) and DCB (ΣDCB₁₀₀) was multiplied by the total influent CBs (ΣCB₀) to account for CBs removed aerobically.

$$\text{Aerobic Degradation}=\left(\Sigma {\text{CB}}_0÷DF\right)-\Sigma {\text{CB}}_{100}$$ Equation 2

Aerobic CB degradation was quantified as the removal of all CB congeners from the column assuming no anaerobic mineralization occurred (Equation 2). Since aerobic degradation occurred after the addition of the oxygen stream, the initial CB concentration was divided by the dilution factor (DF) described in Section 2.1 to reflect the actual removal occurring in the aerobic zone.

Sample data from the last 15 days of each experimental phase were used to estimate steady-state biodegradation activity, with the exception of shorter Phases 6 (5 days) and 7 (10 days; Table 2). Rates of anaerobic dechlorination and aerobic degradation were estimated based on degradation calculated in Equations 1 and 2 divided by the hydraulic retention times (HRT) of each respective redox zone (Table 1). Additional measurements and derived quantities (Table S1), a comprehensive explanation of all calculated parameters (Section S3), and example calculations (Section S4) are presented in the SI.

Table 2.

Summarized column biodegradation activity

Experimental Phase e− Donor Dose	n Samples	Column	Anaerobic Degradationa			Aerobic Degradationb			Cl− Balanceh
			Dechlorinations (μMc	Rate (μM/hr)c,d	e− Donor Utilization (meq %)e	Degradation (μM)c / [% Removal]	Rate (μM/hr)c,d	O2 Utilization (meq %)f / [mg CB/ mg DO]g
1	5	SSC	44.2 ± 13.6	14.7 ± 4.5	2.7%	13.1 ± 2.7 [47%]	11.3 ± 2.3	47.4% [0.41]	81%
0.28 mM NaLac		SC	9.4 ± 1.9	2.1 ± .4	0.6%	18.6 ± .5 [68%]	10.9 ± .3	62.3% [0.58]	50%
2	4	SSC	94.4 ± 6.2	31.3 ± 2.0	1.1%	1.9 ± 2.0 [5%]	1.7 ± 1.7	6.5% [0.05]	86%
1.4 mM NaLac		SC	86.9 ± 5.5	19.6 ± 1.2	1.0%	1.3 ± .4 [4%]	0.8 ± .3	4.5% [0.04]	95%
3	4	SSC	46.3 ± 3.7*	15.4 ± 1.2*	5.8%	5.7 ± 1.7 [31%]	4.9 ± 1.5	18.8% [0.15]	113%
0.14 mM NaLac		SC	0.7 ± 1.3	0.2 ± .3	0.1%	9.7 ± 1.4 [60%]	5.7 ± .8	27.6% [0.26]	112%
4	5	SSC	66.6 ± 3.2	22.1 ± 1.1	0.8%	2.3 ± 1.3 [8%]	1.9 ± 1.1	6.7% [0.05]	101%
1.4 mM NaLac		SC	52.5 ± 4.6	11.8 ± 1.0	0.6%	2.4 ± 1.0 [9%]	1.4 ± .6	7.0% [0.06]	105%
5	6	SSC	66.9 ± 2.1	22.2 ± .7	4.0%	10.0 ± 1.0 [31%]	8.7 ± .9	29.0% [0.24]	93%
0.28 mM NaLac		SC	34.4 ± 2.4	7.8 ± .5	2.1%	14.6 ± 1.1 [48%]	8.6 ± .7	40.0% [0.35]	99%
6		SSC	66.1 ± 3.3	21.9 ± 1.1	4.0%	0.6 ± .9 [0%]	0.6 ± .7	20.6% [0.17]	88%
0.28 mM NaLac, (O2 Off)i	4	SC	27.8 ± 2.6	6.3 ± .6	1.7%	0.5 ± .5 [0%]	0.3 ± .3	14.0% [0.12]	100%
7	5	SSC	69.3 ± 1.7	23.0 ± .6	4.2%	11.6 ± 2.5 [40%]	10.0 ± 2.2	35.0% [0.29]	96%
0.28 mM NaLac		SC	23.6 ± 1.3	5.3 ± .3	1.4%	13.6 ± 1.6 [52%]	8.0 ± 1.0	37.7% [0.34]	86%

^a.Based on undiluted concentrations in column anaerobic zone

^b.Based on diluted concentrations in column aerobic zone

^c.Results presented as average ±1 SD

^d.Calculated as Degradation ÷ HRT of respective column and zone

^e.CB reductive dechlorination e⁻ equivalent as a fraction of influent NaLac e⁻ equivalent, assuming 2 e⁻ per dechlorination and 12 e⁻ per NaLac oxidized

^f.CB oxidation e⁻ equivalent as a % of available O₂ assuming 24 (TCB), 26 (DCB), and 28 (MCB) e⁻ per CB oxidized and 4 e⁻ per O₂ reduced

^g.mg of equivalent influent 1,2,4-TCB degraded as a fraction of mg DO available in aerobic zone

^h.Measured Cl⁻ change ÷ Apparent Cl loss (from CB measurements)

^i.Though O₂ was purged from feed, some oxygen detected at column entrance resulting in low 0.7 mg/L DO at OAI

^*Effluent completely dechlorinated to MCB, indicating CB limitation

2.6. Microbial Community Analysis

DNA was extracted from 0.5 g wet subsamples of each sacrificed column segment using a MoBio PowerSoil® DNA extraction kit (Qiagen, Germantown, MD) following manufacturer instructions and quantified using a Qubit® 3.0 fluorometer dsDNA high sensitivity assay kit (Thermo Fisher Scientific, Waltham, MA). Extracts were prepared for Illumina amplicon sequencing of the V4 region of 16S rRNA gene (primers U515 and E786) for bacteria and archaea (Preheim et al. 2016). Samples were sequenced by 300-base pair paired-end sequencing using an Illumina MiSeq sequencing platform (JHU Genetic Resources Core Facility, Baltimore, MD). Each column segment was extracted, quantified, and sequenced in duplicate. Sequencing data were processed using the qiime2 pipeline (v.2018.4.0) (Bolyen et al. 2018), and taxonomy was assigned using Greengenes 13_8 99% OTUs reference sequences (DeSantis et al. 2006). Quantitative Polymerase Chain Reaction (qPCR) was used to quantify total microbial 16S rRNA gene copies using primers described by Puentes Jácome and Edwards (2017). Concentrations were normalized to the mass of column segment sampled. Details of the downstream sequence analysis and qPCR method can be found in SI (Section S1.7–1.8). Raw 16S reads for each analyzed sample have been uploaded to the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) under BioProject Accession ID PRJNA562559.

3. Results and Discussion

3.1. Sand and Sediment Column (SSC) Activity

After inoculation at Day 0, both anaerobic and aerobic biodegradation pathways were immediately apparent in SSC (Figure 2). At the initial 0.28 mM NaLac dose (Phase 1), 1,2,4-TCB degradation activity stabilized to approximately 44.2±13.6 μM reductive dechlorination and 13.1±2.7 μM aerobic degradation (47% removal), resulting in an effluent mix of DCB and MCB congeners. Vertical sampling showed that degradation pathways were spatially separated along the OAI, with reductive dechlorination exclusive to the anaerobic zone and aerobic degradation to the aerobic zone of the column (Figure 3a). Sulfate transformation also reflected the two distinct redox zones. In the anaerobic zone sulfate reduction co-occurred with CB dechlorination, and in the aerobic zone re-oxidation of reduced sulfur co-occurred with CB oxidation (Figure 3a).

Figure 3.

Representative SSC vertical CB biodegradation profiles during each input condition. (A) 0.28 mM NaLac [Day 256]; (B) 1.4 mM NaLac [Day 214]; (C) 0.14 mM NaLac [Day 179]; (D) 0.28 mM NaLac with imposed anoxia throughout the entire column [Day 277]. Stacked areas represent total molar CB concentrations plotted on left axes. Net degradative chloride release and apparent sulfate reduction relative to column influent plotted on right axes (sulfate reduction treated as a + value). Concentrations at the 100% port were multiplied by the side-stream dilution factor to correctly visualize mass balance of chemical processes, resulting in an exaggeration of true effluent concentrations.

Increasing the NaLac amendment to 1.4 mM (Phase 2) caused a rapid enhancement of reductive dechlorination, more than doubling the extent of reductive dechlorination to 94.4±6.2 μM. However, the extent of aerobic degradation decreased nearly tenfold to 1.9±2.0 μM (5% removal); nearly all 1,2,4-TCB mass entering the system was conserved as reduced MCB (Figure 3b). This five-fold increase in NaLac increased the share of fermentation side-reactions occurring in the column. The amount of electron donor theoretically used for reductive dechlorination decreased from 2.7% in Phase 1 to 1.1% in Phase 2, leaving the majority of the 16.6 millielectron equivalent (meq) NaLac to be used for its fermentation to acetate and propionate, which themselves are reduced electron donors. In the aerobic zone, only 0.81 meq O₂ was available as an electron acceptor to degrade approximately 1.1 meq MCB and the large fraction of excess fermented organic acids (approximately 9 meq). Under this high organic carbon load, residual organic acids likely outcompeted aerobic CB degradation for oxygen at the interface and led to inhibition.

After 26 days under high dosage, influent NaLac was decreased to 0.14 mM in Phase 3. Anaerobic reductive dechlorination activity decreased but remained stable at 46.3±3.7 μM, with all influent 1,2,4-TCB reduced to MCB (Figure 3c). Because nearly complete dechlorination occurred here, 1,2,4-TCB influent was limiting in this phase, and the estimated dechlorination activity was likely an underestimate of the total dechlorination capacity (Table 2). Nevertheless, high reductive dechlorination persisted through the entire 75-day operating period of Phase 3 at a rate equal to that of Phase 1, suggesting that the dechlorinating population acclimated over time to better utilize electron donor. Aerobic degradation slowly recovered over a 30-day period before stabilizing at 5.7±1.7 μM (31% removal) (Figure 2c). This was less than half the extent observed in Phase 1, which was unexpected since less NaLac would be expected to lower the amount of residual O₂ demand in the aerobic zone. The decreased aerobic degradation with decreased electron donor in Phase 3 may have been related to a change in the sulfur oxidation dynamics in the column. After a 20-day lag, sulfate reduction was suppressed in the anaerobic zone (Figure 3c) and effluent sulfate concentrations exceeded influent concentrations (Figure S2b), suggesting transient oxidation of previously immobilized reduced sulfide species. Although sulfides weren’t measured directly here, the column matrix darkened over time, a visual indicator that a sink of precipitated sulfide compounds such as iron sulfide were retained throughout the column. Conservatively, the difference in average sulfate change in Phase 3 (+28 μM) compared to Phase 1 (−12 μM) (Table S1) would indicate an increased theoretical oxygen demand of 0.32 meq (assuming S⁻ oxidized to SO₄²⁻), or 37.6% of total available oxygen (Section S4). These results highlight that even in low sulfate systems, accumulation of reduced sulfur over time can exert substantial oxygen demand on aerobic pathways downgradient as it is re-oxidized to sulfate.

Cycling influent NaLac back to 1.4 mM (Phase 4) again increased reductive dechlorination and disrupted aerobic degradation, as observed in Phase 2 (Figure 2). Sequential dechlorination and aerobic degradation re-occurred in Phase 5 when NaLac was restored to the initial 0.28 mM NaLac dose, indicating changes in organic carbon dosage did not permanently disrupt either degradation pathway (Figure 2). In Phase 6, oxygen was removed from the side-stream by bubbling N₂ gas so the entire column was made anaerobic. No change in reductive dechlorination was observed. However, aerobic CB degradation and sulfide oxidation ceased (Figure 2c), with a vertical profile nearly identical to the high 1.4 mM NaLac inputs (Figure 3d). This phase provided definitive evidence that a complete CB degradation was an oxygen-dependent process, and not an anaerobic process or artifact of dilution. The oxygen stream was restored to the column in Phase 7 after seven days of anoxia. Aerobic CB degradation was re-established within the first day of sampling and stabilized at levels observed prior to anoxia (Figure 2c).

3.2. Sand Column (SC) Activity

SC degradation activity was split across the OAI, similar to SSC (Figure 4a). However, this split was definitively a result of differences in redox zone as opposed to possibly being a result of a stratified inoculation. Despite these differences, the overall degradation patterns in SC in response to variable NaLac dosing were similar to SSC (Figure 2). In general, the extent of aerobic degradation was higher in SC than SSC. Anaerobic sulfate reduction and aerobic sulfur oxidation were both observed (Figure 4) across the OAI; however, the degree of sulfate reduction was consistently lower in SC compared to SSC (Figure S2b). Anaerobic reductive dechlorination activity was lower in SC compared to SSC and showed a comparatively lower electron donor utilization, especially at 0.14 and 0.28 mM NaLac doses (Table 2). These differences may be attributed to the absence of sediment in SC compared to SSC; sediment has been shown to provide a superior surface for microbial attachment to sand (Cápiro et al. 2014) and may have contained additional unquantified bioavailable reductants and nutrients favoring reductive processes (Maillard and Holliger 2016). However, the columns also had differing inoculation schemes, which could have played another important but confounding role in establishing microbial functionality (discussed in Section 3.4).

Figure 4.

Representative SC vertical CB biodegradation profiles during each input condition. (A) 0.28 mM NaLac [Day 256]; (B) 1.4 mM NaLac [Day 214]; (C) 0.14 mM NaLac [Day 179]; (D) 0.28 mM NaLac with imposed anoxia [Day 277]. Data presented in a similar manner to Figure 3.

Degradation activity during the 0.14 mM NaLac input in Phase 3 was notable. After inhibition of CB removal observed at 1.4 mM in Phase 2, aerobic degradation was immediately re-established, more quickly in SC than in SSC (Figure 2c). For the first 20 days of Phase 3 (78 pore volumes), reductive dechlorination activity persisted to a similar extent as SSC (Figure 2b); however, dechlorination activity and sulfate reduction ceased afterwards (Figure 4c), indicating that the 0.14 mM NaLac dose was inadequate to support long-term 1,2,4-TCB dechlorination in the sand matrix. The initial dechlorination activity observed may have been temporarily sustained by a slow decay of excess attached biomass developed during higher NaLac dose in Phase 2, which has been demonstrated by others previously (Sleep et al. 2005, Yang and McCarty 2000). Despite the lack of measured anaerobic activity, TCB degradation and sulfur oxidation were still observed in the aerobic zone (Figure 4c). As with SSC, effluent sulfate exceeded influent sulfate concentrations (Figure S2b), suggesting transient oxidation of retained sulfides may have outcompeted aerobic CB degradation for DO.

Influent NaLac was again increased to 1.4 mM in Phase 4, and reductive dechlorination in SC resumed (Figure 2b). Degradation activity followed similar trends in Phases 5–8 as described above for SSC (Figure 2). There was, however, a gradual but significant decrease in reductive dechlorination activity between Phases 5–6 (p =.0031, Welch’s 1-tailed t-test) and 6–7 (p=.020) through the end of the experiment (Figure 2b). This would suggest that dechlorination in SC may not have been sustainable at low NaLac doses without periodic electron donor spikes to re-stimulate activity.

3.3. Dynamics of Anaerobic and Aerobic Degradation

Anaerobic reductive dechlorination of 1,2,4-TCB resulted in nearly stoichiometric increases of 1,3-DCB, 1,4-DCB, and MCB production while 1,2-DCB was only a minor daughter product (Figure 3, Figure S1). The concentrations of benzene in the samples were below CB detection limits, indicating dechlorination effectively stalled at MCB. Dechlorination activity in both columns was positively correlated with NaLac dose, indicating anaerobic processes were electron donor-limited; conversely, electron donor utilization for dechlorination decreased with increasing NaLac dose (Table 2). Donor utilization varied from 5.8% at 0.14 mM to 0.6% at 1.4 mM, consistent with other reported values from lactate-amended dechlorinating column studies: 0.2% at 20 mM NaLac (Li et al. 2013) and 6.5% at 0.68 mM NaLac (Azizian et al. 2010).

Aerobic CB degradation appeared to occur for all CBs appearing in the column aerobic zones (Figure 3, Figure 4). Degradation was consistently highest at 0.28 mM NaLac, where the effects of excess electron donor and competitive sulfur oxidation were minimal. DO was depleted below detection limits (<0.1 mg/L) in column effluent during all phases, indicating aerobic processes were limited by electron acceptor availability. Oxygen utilization for CB degradation varied from 0.04 mg TCB/mg O₂ consumed at 1.4 mM NaLac to 0.58 mg TCB/mg O₂ at 0.28 mM NaLac (Table 2).

Influent and effluent column media had a circumneutral pH, although slight acidification occurred through the column. Influent pH was 7.0–7.6 in SSC and 7.0–7.5 in SC, and effluent pH was 6.7–7.3 in SSC and 6.8–7.3 in SC. This could be attributed to proton release from dechlorination and sulfide oxidation reactions. Measured chloride release (Figure S2a) closely followed CB degradation (Figure S1); during steady-state measurements in each phase, the mass balance between chloride release and CB transformations was closed (81%−113% in SSC and 86%−113% in SC), except during Phase 1 for SC (50%; Table 2).

In order to show degradation performance in a remediation context, zero-order degradation rates were estimated for all phases relative to HRT in Table 2. The highest observed rates of reductive dechlorination were 31.3±2.0 μM/hr in SSC and 19.6±1.2 μM/hr in SC during Phase 2 (1.4 mM NaLac). These rates were over an order of magnitude higher than the 0.4–1 μM/hr CB dechlorination rates reported by Bosma et al. (1988) in the only comparable anaerobic column study in the literature. Our results demonstrate that fixed-film 1,2,4-TCB dechlorination can occur at a rapid rate, which was only limited experimentally by electron donor dosage rather than degradation kinetics. The highest aerobic degradation rates (on a total CB basis) were 11.3±2.3 μM/hr in SSC and 10.9±.3 μM/hr in SC during Phase 1 (0.28 mM NaLac). These rates were much lower than rates of 44–87 μM/hr estimated from Kurt et al. (2012) in a short-run MCB-degrading column system simulating a sediment-water interface. In contrast to Kurt et al. (2012), apparent aerobic degradation in our study was impacted by limited oxygen availability for CB degradation.

A key question this experiment addressed was what effects the coupling of anaerobic reductive dechlorination to aerobic degradation would have on overall CB degradation. Since complete anaerobic dechlorination and mineralization of TCB were not observed in this experiment, the only path to mineralization was through the aerobic pathway. Although MCB, DCBs, and TCBs are all aerobically degradable, some bioremediation studies have found that less chlorinated CBs degraded at significantly faster rates than 1,2,4-TCB under aerobic conditions (Dermietzel and Vieth 2002, Elango et al. 2010). These prior studies would suggest that an initial dechlorination step could enhance the overall rates and extents of aerobic mineralization.

In Phase 3 of this experiment, two very different reductive dechlorination outcomes were observed between columns. In SSC, nearly all CBs entering the anaerobic zone were dechlorinated to MCB and 1,2- and 1,4-DCB (Figure 3c). In SC, however, little to no dechlorination occurred, leaving only 1,2,4-TCB (Figure 4c). In both cases, however, significant aerobic degradation occurred (Table 2). Additionally, in the vertical column profiles, it can be seen qualitatively that all CB congeners show a concentration decrease within the aerobic zone (Figure 3a, c; Figure 4a, c). This would imply that all CB congeners were subject to aerobic biodegradation.

To test this observation more rigorously, the measured aerobic degradation of each congener n was compared to the change in relative abundance nCB/ΣCB of that congener (Equation 3) across the aerobic zone (from 50% to 100% ports) at each vertical sample timepoint to determine whether certain congeners degraded at a greater extent compared to others (Figure S3).

$$\%\text{ Change in Congener }n\text{ Abundance}=\frac{n{\text{CB}}_{100}}{\Sigma {\text{CB}}_{100}}-\frac{n{\text{CB}}_{50}}{\Sigma {\text{CB}}_{50}}$$ Equation 3

Between MCB, total DCBs, and 1,2,4-TCB there was no statistically significant change (≤2%; p>.05) in relative abundances, indicating similar aerobic degradation rates and no apparent preferential degradation based on the degree of congener chlorination during aerobic degradation (Table S2). There was a significant preference for 1,2- and 1,4- DCB over 1,3- DCB (p<.05), however the small effect size (<2.8%), small samples size (n=37), and low amount of 1,2-DCB present adds uncertainty to this observation (Table S2). This is in agreement with previous work by Sander et al. (1991) showing very similar oxygen uptake and mineralization rates between MCB, DCBs, and 1,2,4-TCB with pure cultures of Ralstonia sp. strain PS12 and Burkholderia sp. strain PS14. Previous trends in degradation rates reported in the literature between MCB, DCBs, and TCBs are conflicting; their respective degradation potentials have been shown to be highly dependent on factors such as geochemical conditions, substrate acclimation, and microbial strain (Elango et al. 2010, Naziruddin et al. 1995, Sander et al. 1991, Schraa et al. 1986). Although not tested individually, our experimental results demonstrated 1,2,4-TCB and its reduced daughter congeners apparently degraded at a similar rate aerobically. Furthermore, the degree of CB chlorination and the extent of aerobic degradation from effluent column samples did not correlate (Figure S4). Based on our results, initial reductive dechlorination of 1,2,4-TCB did not appear to benefit subsequent aerobic degradation.

3.4. Microbial Community Analysis

Sequenced SSC inocula revealed distinct mixed communities. The anaerobic WBC-2-derived inoculum consisted primarily of classes containing obligate anaerobes including Clostridia, Bacteroidia, Methanomicrobia, Methanobacteria, Thermotogaea, and Anaerolineae (Figure S5a). Among the community, three known OHRB genera were identified at low relative abundances: Dehalobacter (1.3%), Dehalogenimonas (0.2%), and Dehalococcoides (0.05%). This was a substantial shift in OHRB population from the parent WBC-2 culture, which consisted primarily of Dehalococcoides (8.8%) and Dehalogenimonas (8.7%), but only a small fraction of Dehalobacter (0.012%; data not shown). This suggests that 1,2,4-TCB dechlorination may have selected for Dehalobacter enrichment over other OHRB. A similar shift to Dehalobacter in mixed communities was observed by others in a chloroethene enrichment KB-1 (SiREM Labs) when substrate was changed to 1,2,4-TCB (Puentes Jácome and Edwards 2017). The aerobic inoculum was dominated by Betaproteobacteria and Saprospirae (Figure S5a). Two Betaproteobacteria genera associated with aerobic CB degradation (Jiang et al. 2009) were highly abundant: Pandoraea (19.5%) and Burkholderia (7.2%), both in the Burkholderiaceae family.

Vertical taxa profiles in columns provided a detailed description of changing community structures and key functional populations across model OAIs. Class-level profiles are presented in the SI (Figure S6). Over 139 and 107 unique genera were identified in the SSC and SC column segments, respectively. Dehalobacter, Clostridium, Pseudomonas, Thiobacillus, Rhodocyclus, Zoogloea, and Geobacter were all at least 10% abundant in an SSC segment; Sediminibacterium, Comamonas, Rhodocyclus, Pseudomonas, Thiobacillus, Zoogloea, an unspecified Sphingomonadaceae genus, and an unspecified Comamonadaceae genus were at least 10% abundant in SC (Figure 5). The roles of many abundant but functionally diverse metabolic generalists such as Pseudomonas and Geobacter were unclear since functionality can be highly strain- and environment-dependent, beyond the resolution of short-read 16S sequencing (Dos Santos et al. 2004, Mahadevan et al. 2010). However, several genera that have been associated with specific redox functionality by others were identified with at least 1% relative abundance in various column segments. Dehalobacter (Clostridia) was the only OHRB identified in either column and was the presumed CB dechlorinator. Sulfate reducing bacteria (SRB) Desulfosporosinus (Pester et al. 2012) and methanogen Methanosarcina (De Vrieze et al. 2012) were also identified. Chemolithotrophic sulfur-oxidizer Thiobacillus (Kelly and Wood 2000) was prominently abundant in both columns. Several Proteobacteria groups implicated as aerobic CB degraders by others were also abundant in both columns, including Pandoraea, Xanthobacter, and an unspecified Sphingomonadaceae genus (Field and Sierra-Alvarez 2007, Jiang et al. 2009).

Figure 5.

Genus-level microbial community profiles in column biofilm segments. Composition expressed as relative abundance based on 16S rRNA amplicon sequences. Bacteria and Archaea classes with <2% representation aggregated as “Other”. Organisms that were “Unclassified” at the genus-level were grouped by their closest common taxonomic level.

Absolute abundances of these functionally-relevant genera along the vertical column profiles were estimated by multiplying the total 16S copy numbers from qPCR by relative abundances from sequencing (Figure 6). Total 16S abundances were fairly evenly distributed along the length of the SSC anaerobic zone (6.5×10⁸-1.4×10⁹ copies/g) compared to a more drastic decline after segment 1 of SC (1.9×10⁸-3.1×10⁹ copies/g), suggesting a rapid depletion of electron donor close to the bottom of SC (Figure 6). After the addition of O₂ in the aerobic zone, 16S abundances spiked slightly in segment 6 of SSC (8.7×10⁸ copies/g) before gradually decreasing through segment 8 (1.3×10⁹ copies/g; Figure 6a). In contrast, abundances increased along the length of the SC aerobic zone through segment 8 (Figure 6b), suggesting a more rapid depletion of oxygen in SSC following the side-stream addition. These distributions were also reflected in total DNA and bulk protein concentrations measured (Figure S7).

Figure 6.

Top: Vertical biofilm profiles of functional genera and total microbial 16S copy number in (A) SSC and (B) SC. Bottom: Vertical porewater profiles of CB biodegradation, cumulative chloride release, and net sulfate reduction in (C) SSC and (D) SC. All samples taken at Day 295.

In SSC, Dehalobacter was the most abundant genus in segments 1–3 of the anaerobic zone (≤3.6×10⁸ copies/g), nearly an order of magnitude greater than anaerobes Desulfosporosinus and Methanosarcina (Figure 6a). These anaerobes maintained high abundances in segments 1–3 before decreasing along the remaining length of the column. In segment 4, Thiobacillus succeeded Dehalobacter as the most abundant genus. This preceded an increase in potential aerobic CB degraders Pandoraea and Xanthobacter within the aerobic zone. Measured CB and sulfur redox activity at the time of sampling (Figure 6c) appeared to co-occur with corresponding segments of high functional genera abundances (Figure 6a). Although methane concentrations were not measured, the presence of high Methanosarcina abundances suggests methanogenesis was another prevalent anaerobic process occurring in this column.

In SC, the aforementioned functional anaerobe abundances were over 2 orders of magnitude less abundant than total 16S counts and thus accounted for a much smaller fraction of the total community compared to SSC (Figure 6b). The abundance of these genera quickly decreased after segment 1, but were stable through the remaining column length. Correspondingly, the majority of observed reductive dechlorination and sulfate reduction activity occurred within the first 25% of the column length (Figure 6d). Within the aerobic zone, Thiobacillus abundances peaked at segment 6, while Xanthobacter and Sphingomonadaceae abundances continued to increase through segment 8. The peak abundances of potential aerobic CB degraders at segment 8 would suggest that degradation did not occur immediately after the addition of O₂ in the SC aerobic zone, however the resolution of porewater sampling here was too coarse to confirm this through chemical activity.

The enrichment of Dehalobacter (26.2–50.9% relative abundance) in the bottom segments of SSC was unexpectedly high compared to previously reported literature observations. Dehalobacter abundances in CB-degrading batch cultures have been reported from less than 0.1% in site sediment microcosms (Qiao et al. 2017) up to approximately 25% in sequentially enriched fed-batch microcosms (Puentes Jácome and Edwards 2017). Though limited Dehalobacter-specific column data is available, Li et al. (2013) reported up to 2.4% enrichment in a PCP-dechlorinating column packed with sediment and glass. This was more comparable with the observed enrichment in SC (<1.2%). We posit that the combination of shorter HRT and addition of sediment may have selectively enriched Dehalobacter in SSC. Decreasing the HRT in a chloroethene-dechlorinating CSTR was found by others to decrease H₂ concentrations, making OHRB metabolism more thermodynamically favorable than that of acetogens, methanogens, and SRB (Yang and McCarty 1998). Additionally, Cápiro et al. (2014) found 2-fold lower interaction energy barriers and greater attachment rates for PCE dechlorinators to soil compared to fine sand.

Although the causes of differences between SSC and SC cannot be known with certainty, the resulting order of magnitude differences in observed CB degrader abundances help to explain the differences in activity observed throughout the time-series. SSC consistently demonstrated greater degrees of CB dechlorination (Figure 2b) than SC, corresponding to more highly enriched Dehalobacter. Conversely, the extent of aerobic CB degradation was generally greater in SC (Figure 2c), which corresponded to the potential aerobic degraders being more highly enriched in SC than in SSC.

Few other studies, to our knowledge, have carefully profiled the attached microbial communities in stratified systems actively degrading halogenated organic compounds (Li et al. 2013, Lima and Sleep 2007). Liquid samples have more frequently been analyzed due to the ease of producing and collecting representative samples, however they do not always correlate well with the true microbial community composition in the subsurface (Cápiro et al. 2014). In this experiment, column porewater at the time of sampling was dominated by Pseudomonas in SSC (81.1%) and Comamonas in SC (81.5%) (Figure S5b). Based on relative abundance, dominant and functionally relevant genera from the column matrices were underrepresented in porewater by nearly 3 log units, and some organisms of interest were not even detected in the porewater (Table S3). Notably, Dehalobacter abundance was underrepresented by 2.4 log units in SSC and Sphingomonadaceae by 2.8 log units in SC (Table S3). These results illustrate the importance of sampling solid matrices as well as porewater to more fully capture diversity, functional potential, and spatial heterogeneity of subsurface microbial communities.

3.5. Implications for Contaminated Sites

This proof-of-concept study demonstrated sustained degradation of dissolved CBs at OAIs in both site-relevant sediment (SSC) and clean sand bed (SC) matrices. These results showed a range of potential degradation outcomes that could be reasonably expected at shallow sites contaminated with concentrations of CBs in the micromolar range. Under lower NaLac concentrations of 0.14–0.28 mM (equivalent to 5–10 mg/L DOC), substantial aerobic degradation was observed; however, under a higher NaLac concentration of 1.4 mM (50 mg/L DOC), aerobic degradation activity was inhibited, instead favoring substantial reductive dechlorination to DCBs and MCBs. This narrow window in which DOC dose depleted available DO shows that aerobic bioremediation may not be viable at sites with modest concentrations of dissolved or sediment-bound organic carbon that exceed the available flux of oxygen.

In such cases, alternative strategies should be considered to manage contamination. Wetland plants have been shown to significantly increase oxygen flux and enhance aerobic CB degradation in constructed wetland studies (Braeckevelt et al. 2008, Braeckevelt et al. 2011). Sediment amendment with activated carbon has proven an effective means to immobilize hydrophobic organic contaminants and facilitate slow aerobic degradation over time, effectively stabilizing differences in oxygen supply and demand (Payne et al. 2017, Song et al. 2015, Tiehm et al. 2002). Additionally, amendment of solid-phase oxygen-releasing compounds has also shown effectiveness in enhancing aerobic degradation (Kao et al. 2003, Lin et al. 2004). Finally, future consideration should be given to the possibility of complete anaerobic mineralization at sites. Though dechlorination stalled at MCB in this work, recent studies by others have also implicated Dehalobacter and other OHRB strains with complete, albeit slow, reduction to benzene and subsequent anaerobic mineralization (Fung et al. 2009, Liang et al. 2013, Puentes Jácome and Edwards 2017, Qiao et al. 2017). In the future, enrichment of an efficient MCB-dechlorinating and benzene-oxidizing bioaugmentation culture may potentially facilitate oxygen-independent site treatment of CBs.

These columns demonstrated conditions where reductive dechlorination did not lead to enhanced CB mineralization outcomes. Although aerobic degradation appeared to be viable for all observed CB congeners, reductive dechlorination to CBs may have other potential benefits for site remediation provided dechlorination stalls at MCB. With lower sorption potentials to sediments and greater aqueous solubilities, less chlorinated CBs are more environmentally mobile and may be flushed from the subsurface more quickly than more chlorinated CBs. Yang and McCarty, for example, demonstrated a 3-fold increase in PCE DNAPL dissolution as a result of reductive dechlorination (2002). Some ecotoxicity studies have found that MCB is significantly less toxic than DCBs and TCB to Pseudomonas fluorescens (Boyd et al. 1998), Daphnia, and salmon (Calamari et al. 1983). However, this detoxification effect must be weighed against potentially higher dissolved concentrations of the dechlorinated congeners mobilized in the aqueous phase.

In this experiment, we assumed steady fluxes of contaminants and available oxygen, when these are often temporally and spatially heterogeneous in the field (Lorah et al. 2014). Degradation activity may therefore be quite transient. By cycling through long periods of alternative redox conditions, our study did demonstrate that temporary inhibition of degradation did not permanently alter the microbial community’s capacity to degrade CBs in the long-term, even on inert sand matrices (SC). This reflects positively on the potential for bioaugmented degraders to persist in the OAI. However, since conditions were still controlled in the lab, validation in the field, especially in the presence of native microbial communities, is warranted. In conjunction with this study, a pilot-scale bioaugmented reactive barrier at SCD is currently being evaluated by the authors to assess anaerobic and aerobic CB biodegradation in situ.

4. Conclusions

• The dual-flow column system developed here provided an effective method to carefully measure anaerobic-aerobic processes under stable conditions. Both anaerobic and aerobic CB degradation pathways were resilient and remained active through the end of the 295-day experiment. CB degradation mass balance was approximately 86%−113% through the majority of the experiment.

• Anaerobic reductive dechlorination and aerobic degradation of 1,2,4-TCB occurred sequentially across the simulated OAI. Reductive dechlorination of 1,2,4-TCB stalled at MCB. CB mineralization was only observed by aerobic degradation, which occurred for 1,2,4-TCB and all daughter DCB and MCB daughter products. Initial reductive dechlorination did not appear to enhance subsequent aerobic degradation rates in this system.

• Degradation outcomes varied with NaLac loading. Observations included reductive dechlorination inhibition at 0.14 mM NaLac, mixed degradation pathways at 0.28 mM, and aerobic degradation inhibition at 1.4 mM. This demonstrated a narrow and sensitive window in which degradation along the OAI was affected.

• Maximum observed aerobic CB removal was low – approximately 19 μM (3.5 mg/L) degraded with oxygen utilization of 0.58 mg TCB/mg O₂. This suggests that natural conditions may not facilitate adequate CB mineralization in highly contaminated water when DO is limiting.

• The column packed with sediment and sand (SSC) supported high reductive dechlorination across all tested NaLac doses (0.14–1.4 mM) and enriched OHRB Dehalobacter up to 3.6×10⁸ copies/g (50.9% relative abundance), which likely thrived on the sediment matrix. The sand column (SC) inoculated with SSC effluent retained similar degradation functionality, but with diminished dechlorination activity and Dehalobacter abundances. However, SC supported greater aerobic CB degradation and potential aerobic degrader abundances.

• Oxic-anoxic interfaces were formed in columns to study chlorobenzene biodegradation

• Anaerobic and aerobic degradation occurred sequentially along interfaces

• Dehalobacter in biofilms was responsible for anaerobic reductive dechlorination

• Less chlorinated daughter compounds were not more amenable to aerobic degradation

• Degradation outcomes depended on competing electron donor and oxygen availability

Abbreviations

OAI

Oxic-anoxic Interface

SC

Sand Column

SSC

Sand and Sediment Column

NaLac

Sodium Lactate

HRT

Hydraulic Retention Time

SCD

Standard Chlorine of Delaware Superfund site