Article pubs.acs.org/est
Direct Evaluation of in Situ Biodegradation in Athabasca Oil Sands Tailings Ponds Using Natural Abundance Radiocarbon Jason M. E. Ahad*,† and Hooshang Pakdel‡ †
Geological Survey of Canada, Natural Resources Canada, Québec City, Québec, G1K 9A9, Canada INRS Eau Terre Environnement, Québec City, Québec, G1K 9A9, Canada
‡
S Supporting Information *
ABSTRACT: Compound-specific stable (δ13C) and radiocarbon (Δ14C) isotopes of phospholipid fatty acids (PLFAs) were used to evaluate carbon sources utilized by the active microbial populations in surface sediments from Athabasca oil sands tailings ponds. Algal-specific PLFAs were absent at three of the four sites investigated, and δ13CPLFA values were generally within ∼3‰ of that reported for oil sands bitumen (∼−30‰), suggesting that the microbial communities growing on petroleum constituents were dominated by aerobic heterotrophs. Δ14CPLFA values ranged from −906 to −586‰ and pointed to significant uptake of fossil carbon, particularly in PLFAs (e.g., cy17:0 and cy19:0) often associated with petroleum hydrocarbon degrading bacteria. The comparatively heavier Δ14C values found in other, less specific PLFAs (e.g., 16:0) indicated the preferential uptake of younger organic matter by the general microbial population. Since the main carbon pools in tailings sediment were essentially “radiocarbon dead” (i.e., Δ14C ∼ −1000‰), the principal source for this relatively modern carbon is considered to be the Athabasca River, which provides the bulk of the water used in the bitumen extraction process. The preferential utilization of the minor amount of younger and presumably more labile material present in systems otherwise dominated by petroleum carbon has important implications for remediation strategies, since it implies that organic contaminants may persist long after reclamation has begun. Alternatively, this young organic matter could play a vital and necessary role in supporting the microbial utilization of fossil carbon via cometabolism or priming processes.
■
INTRODUCTION With approximately 170 billion barrels of oil remaining,1 the bitumen contained within the Peace River, Cold Lake, and vast Athabasca oil sands deposits of Northern Alberta, Canada, represents one of the world’s largest known energy reserves. The continued expansion and development of this “unconventional” resource, however, present a variety of air and water quality challenges.2−6 One of the key concerns is the fate of contaminants contained within the large volumes of oil sands process-affected water (OSPW) stored in surface impoundments (i.e., tailings ponds). As a result of the bitumen extraction process, typical OSPW contains high levels of total dissolved solids (TDS), petroleum hydrocarbons, and acid extractable organics containing “naphthenic acids” (NAs); the compounds generally considered to be the principal source of toxicity in OSPW.7 NAs may seep into the shallow groundwater near tailings ponds8,9 and thus pose a potential risk to the Athabasca River. Responsible management and stewardship of the Athabasca oil sands require that operators eventually return the land disturbed by mining operations to their original state. Reclamation plans include both dry and wet landscape options.7 For many tailings ponds, one of the leading options following mine closure is to convert them into end-pit lakes, which are designed to mimic natural dimictic lakes by capping tailings with freshwater.10 One of the primary functions of end© 2013 American Chemical Society
pit lakes is to aerobically biodegrade organic contaminants released by tailings (i.e., the mixture of sand, silt, clay, water, and residual hydrocarbons remaining after bitumen extraction) and any runoff contaminated by passing over the reclaimed landscape.10 Given the important role of natural attenuation in tailings pond reclamation plans, an understanding of active microbial degradation processes in these environments is essential. Using controlled experimental microcosms, previous research has demonstrated the anaerobic degradation of petroleum hydrocarbons into methane by the indigenous microorganisms present in mature fine tailings (MFT).11−13 The compounds degraded include BTEX (benzene, toluene, ethylbenzene, and xylenes), naphtha (low molecular weight aliphatics and aromatics in the C3 to C14 range used as a diluent in bitumen processing),12 and both short-chain (C6, C7, C8, and C10)11 and longer-chain (C14, C16, and C18) n-alkanes.13 The aerobic microbial degradation of other tailings pond organic contaminants such as NAs is also well documented under controlled laboratory conditions (e.g., refs 14−17), although commercially available NAs have been shown to be Received: Revised: Accepted: Published: 10214
May 22, 2013 August 9, 2013 August 19, 2013 August 19, 2013 dx.doi.org/10.1021/es402302z | Environ. Sci. Technol. 2013, 47, 10214−10222
Environmental Science & Technology
Article
considerably more biodegradable than those found in OSPW.18 Despite the potential for biodegradation of the suite of organic contaminants found in tailings ponds, however, the direct verification of in situ natural attenuation processes remains a challenge. The use of natural abundance radiocarbon (14C; half-life = 5730 years) analysis can provide valuable insight into microbial biodegradation of organic contaminants. This is due to the fact that organic compounds derived from petroleum are millions of years old and hence contain no detectable 14C (i.e., “fossil” carbon; Δ14C = −1000‰) whereas, in contrast, the 14C content of recently fixed organic matter (OM) will reflect that of modern atmospheric CO2 (Δ14C = +55‰19). Determination of 14C contents of active cellular membrane components such as phospholipid fatty acids (PLFAs) and comparison to potential carbon sources thus allows for direct in situ assessment of microbial uptake of petroleum-derived carbon.20−25 The measurement of the 14C contents of PLFAs can also shed light into the dominant microbial communities responsible for fossil carbon assimilation. For instance, relatively lower 14 C contents in PLFAs indicative of heterotrophic bacteria (16:1, 17:0, 18:n, and 20:0) compared to those associated with phototrophic organisms (20:n) was attributed to petroleum hydrocarbon uptake by heterotrophic but not phototrophic organisms.24 This study utilizes natural abundance stable (δ13C) and radiocarbon (Δ14C) isotopes to determine microbial carbon sources and assess intrinsic biodegradation of organic contaminants in Athabasca oil sands tailings ponds. Surface sediment samples were collected from four different sites representing tailings material from varying stages of development and use. The relative distributions of PLFAs were used to assess differences in microbial populations between the different tailings ponds. This is the first study to use this approach to provide a direct evaluation of in situ biodegradation of organic contaminants in these systems.
and recycled water. TP3 has been in operation for the past several decades as a storage area for coarse tailings, although it is also used for temporary fluid storage. The site is expected to eventually provide storage for ∼1000 Mm3 of sand. TP4 has been in operation for over a decade and like TP1 is used primarily as a fluid storage facility. TP4 currently provides storage for coarse tailings, MFT, and recycled water. Petroleum Hydrocarbons. Between 40 and 60 g of ovendried tailings sediment was Soxhlet extracted with dichloromethane (DCM), treated with activated copper to remove elemental sulfur, and separated into fractions by gravity column chromatography using precombusted (450 °C for 8 h) fully activated silica gel (70−230 mesh, Silicycle) following a method adapted from Ahad et al.20 Total petroleum hydrocarbon (TPH) concentrations were determined by weighing the material eluting in 1:1 hexane/DCM following the removal of solvent by gentle evaporation under a stream of ultrahigh purity (UHP) N2. An Agilent gas chromatograph mass spectrometer (GC-MS) system (MSD 5975C and GC 7890A) equipped with a 30 m × 0.25 mm i.d. DB-5 column (0.25 μm film thickness) was used to qualitatively characterize the TPHs. The following GC oven temperature program was used: 35 °C (0.2 min), 30 °C/min to 120 °C, and 5 °C/min to 300 °C (20 min). Tailings Sediment Carbon Pools. The stable and radiocarbon isotope signatures of several different fractions corresponding to the main potential carbon pools in tailings sediment were examined: TLE (total lipid extract), solventextracted residue (EXT-RES), and total organic carbon (TOC). In addition to TPHs, the TLE pool includes all organic compounds which are solvent-extractable. TLE was extracted from 10 to 20 g of oven-dried tailings sediment with 1:1 acetone/hexane using a microwave accelerated reaction system (MARS, CEM Corporation). A second extraction using 9:1 DCM/methanol was also carried out and combined with the first one to ensure complete removal of solvent-extractable components. EXT-RES is defined as the residual organic carbon remaining after solvent extraction (i.e., TLE removal) and is a proxy for nonpetroleum-derived sedimentary OM.20,22,23,27 Percentages of organic carbon in TOC and EXT-RES were determined using a Costech elemental analyzer following decarbonation using H2SO3. Microbial PLFAs. For PLFA concentrations and δ13C analyses, 30−40 g of tailings sediment was extracted by the modified Bligh and Dyer method28 using 2:1 methanol/DCM. Large extractions (2.6−2.8 kg) were used for Δ14CPLFA analyses (sites TP1, TP2, and TP4) in order to obtain a sufficient mass of carbon (i.e., >∼50 μg) required for accurate measurements.29 Samples were filtered and phase separated, and the organic fraction was subsequently separated into three fractions (DCM, acetone, methanol) by gravity column chromatography using precombusted (450 °C for 8 h) fully activated silica gel (70−230 mesh, Silicycle). The phospholipid fraction (dissolved in methanol) was evaporated to dryness using N2 and converted to fatty acid methyl esters (FAMEs) via mild alkaline methanolysis.28,30 A secondary silica gel chromatography step was used to isolate FAMEs prior to identification and quantification using the same GC/MS and column described above. The following GC oven temperature program was used: 40 °C (1 min), 20 °C/min to 130 °C, 4 °C/min to 160 °C, and 8 °C/min to 300 °C (5 min). FAMEs were identified using a bacterial reference standard (Bacterial Acid Methyl Ester Mix, Sigma-Aldrich), mass-fragmentation patterns, and retention times and quantified using external FAME standards (12:0,
■
MATERIALS AND METHODS Sample Collection and Site Description. Surface (∼0−2 cm) sediment samples were collected in September 2010 from four different tailing ponds (TP1, TP2, TP3, and TP4) situated in the Athabasca oil sands region ∼30−60 km north of the city of Fort McMurray, Alberta, Canada. Samples were collected near the edge of the ponds and close to the water surface (water depth ∼ 0.1−0.3 m) to target aerobic biodegradation processes in these systems. No vegetation was observed in the vicinity where sampling occurred. Water samples were collected to determine concentrations of total acid extractable organics (AEOs) containing NAs following the protocol described by Headley et al.26 Additional samples were collected in June 2011 for δ13C measurements of dissolved inorganic carbon (DIC). The hydrogeochemical parameters of tailings pond water samples are provided in the Supporting Information (Table S1). TP1, which has been in operation for over three decades, is used primarily as a fluid storage facility. It currently provides storage for MFT, coke, flotation tailings, froth treatment tailings, and recycled water for plant operations. Originally an open pit mine in operation until 1999, TP2 has since been filled with coarse tailings and composite tailings, which combines MFT with gypsum and sand as the tailings are deposited. Parts of TP2 are currently undergoing dry landscape reclamation and are being revegetated, although the site still receives some MFT 10215
dx.doi.org/10.1021/es402302z | Environ. Sci. Technol. 2013, 47, 10214−10222
Environmental Science & Technology
Article
Figure 1. The relative distributions (mole %) of 22 of the most abundant PLFAs found in tailings pond sediments (sites TP1, TP2, TP3, and TP4). The full suites of PLFAs identified in each sample are provided in Table S3, Supporting Information.
14:0, 16:0, 18:0, and 20:0). PLFAs are identified as Z:nΔx, where Z is the total number of carbon atoms in the fatty acid chain, n is the number of double bonds, and Δx is the position of the double bond, if known. Cyclopropyl PLFAs are represented by the prefix “cy”, and methyl group branching is indicated by the prefixes “i” (iso), “a” (anteiso), and “br” (unknown position). The letters A, B, and C next to unsaturated PLFAs denote different isomers (double bond positions not determined). Large-scale process blanks extracted using identical solvent volumes, conditions, and procedures as samples targeted for Δ14CPLFA analysis yielded no detectable amounts of background FAME contamination. Preparative Capillary Gas Chromatography (PCGC). Individual FAMEs for radiocarbon analysis were separated and collected using an Agilent 7890A GC equipped with two 30 m × 0.53 mm i.d. ZB-5 columns (0.5 μm film thickness; Phenomenex) and a flame ionization detector (FID) coupled to a Gerstel preparative fraction collection (PFC) system. The injection volume was 5 μL, and each sample was injected ∼100 times. The following GC oven temperature program was used: 40 °C (1 min), 20 °C/min to 130 °C, 4 °C/min to 250 °C, and 25 °C/min to 300 °C (5 min). The Gerstel cold injection system (CIS) was kept at 10 °C for 0.18 min at the start of the run. Approximately 5% of the effluent was passed to the FID, and the remaining 95% went to the PFC system. The transfer line between the GC and PFC was set at 300 °C, and the PFC interface was set at 310 °C. Samples were collected cryogenically in glass U-traps at −20 °C using chilled methanol. U-traps were eluted with DCM and transferred to amber 2 mL GC vials. The samples were then cleaned up by silica gel and analyzed by GC-MS to check for potential impurities, which were insignificant in all cases. δ13C and Δ14C Analysis. δ13C ratios of individual PLFAs were determined at the Delta-Lab of the Geological Survey of Canada using a PRISM-III dual inlet isotope ratio mass spectrometry (IRMS) system equipped with a Hewlett-Packard GC (HP 5890 Series II) and a HP-5 column (30 m × 0.32 mm × 0.25 μm). The combustion interface was packed with CuO only and kept at 900 °C. The following GC oven temperature program was used: 40 °C (1 min), 20 °C/min to 130 °C, 3 °C/ min to 226 °C (0.1 min), and 40 °C/min to 290 °C (2 min). δ13C values were analyzed using CO2 calibrated against international carbonate standards (NBS 18 and NBS 19) that were converted to CO2 by reaction with phosphoric acid using a traditional vacuum line. The gas mixture was kept at 25 °C for
24 h, and the CO2 was cryogenically purified and isolated in a break-seal tube prior to analysis. A mixture containing 5-αandrostane obtained from the Biogeochemical Laboratories at Indiana University and three in-house FAME standards (16:0, 18:0, and 20:0) were injected into the GC-IRMS after every three sample injections to assess accuracy. Samples were analyzed at several different injection volumes to obtain peak sizes which were within the calibration range of the IRMS system. The standard deviation of triplicate analyses was TP4 > TP3 > TP2). The substantial difference in TPH levels between sites is likely the result of the larger sand and lower background TOC components in TP2 and TP3 compared to TP1 and TP4 (Table S2, Supporting Information). The full scan GC/MS chromatograms for the fractions containing TPHs were characterized by a large 10216
dx.doi.org/10.1021/es402302z | Environ. Sci. Technol. 2013, 47, 10214−10222
Environmental Science & Technology
Article
unresolved complex mixture “hump” in all four samples (Figure S1, Supporting Information). The TPH fractions at TP2, TP3, and TP4 contained low levels (
