Abstract
Sediments from the Athabasca River and its tributaries naturally contain bitumen at various concentrations, but the impacts of this variation on the ecology of the river are unknown. Here, we used controlled rotating biofilm reactors in which we recirculated diluted sediments containing various concentrations of bituminous compounds taken from the Athabasca River and three tributaries. Biofilms exposed to sediments having low and high concentrations of bituminous compounds were compared. The latter were 29% thinner, had a different extracellular polysaccharide composition, 67% less bacterial biomass per μm2, 68% less cyanobacterial biomass per μm2, 64% less algal biomass per μm2, 13% fewer protozoa per cm2, were 21% less productive, and had a 33% reduced content in chlorophyll a per mm2 and a 20% reduction in the expression of photosynthetic genes, but they had a 23% increase in the expression of aromatic hydrocarbon degradation genes. Within the Bacteria, differences in community composition were also observed, with relatively more Alphaproteobacteria and Betaproteobacteria and less Cyanobacteria, Bacteroidetes, and Firmicutes in biofilms exposed to high concentrations of bituminous compounds. Altogether, our results suggest that biofilms that develop in the presence of higher concentrations of bituminous compounds are less productive and have lower biomass, linked to a decrease in the activities and abundance of photosynthetic organisms likely due to inhibitory effects. However, within this general inhibition, some specific microbial taxa and functional genes are stimulated because they are less sensitive to the inhibitory effects of bituminous compounds or can degrade and utilize some bitumen-associated compounds.
INTRODUCTION
Oil sands are unconventional petroleum deposits where bitumen, a dense and viscous form of petroleum, is found in combination with sand, clay, and water. One of the largest bitumen reservoirs, the Athabasca oil sands, is located in northeastern Alberta, Canada, along the Athabasca River. Here, as the Athabasca River and its tributaries (e.g., Ell's, Steepbank, and Firebag) cut through the Athabasca oil sands formation, oil sands are eroded, making the water and sediments naturally enriched in various bituminous compounds. These compounds are mainly asphaltenes but also include various aromatic hydrocarbons (typically high-molecular-weight, partially saturated polycyclic aromatic hydrocarbons similar to hopanoids), naphthenic acids (NA), and alkanes (1, 2). The concentrations of these compounds vary naturally due to geological and hydrological factors, as well as because of aerial deposition from oil-upgrading activities (3, 4, 58). There are also some indications of a potential hydraulic connectivity between the oil sands tailings ponds (which are often above grade, setting up a hydraulic head) and the at-grade or subsurface natural water bodies (5). Many microorganisms, often belonging to the Proteobacteria, can degrade and utilize hydrocarbon compounds that are found in bitumen (6–8). The Betaproteobacteria were previously shown to be positively correlated with various hydrocarbon indicators in a field survey of sediments from the Athabasca River, suggesting that they are able to use some of the bituminous compounds as carbon sources (9). However, the Athabasca River is generally considered to be nitrogen and phosphorus limited (10), which probably strongly limits the utilization of bituminous compounds by microorganisms, as even with abundant carbon sources microbial growth cannot occur without N and P.
Although some bituminous compounds like asphaltenes and alkanes generally are not toxic to microbial activity, other compounds, like polycyclic aromatic hydrocarbons and naphthenic acids, can inhibit microbial growth. Although some bacteria can metabolize them (11–13), naphthenic acids can have negative effects on bacteria (14, 15), algae (16, 17), and cyanobacteria (18). Aromatic hydrocarbons are also known to be toxic to a variety of microbial processes (19, 20) and to bacteria (21, 22). Photosynthetic organisms were reported as being more strongly inhibited by contaminants than other organisms (9, 23) but also as being stimulated indirectly by oil addition through decreased grazing pressure (24). In river ecosystems, biofilms and flocs containing photosynthetic organisms are in large part responsible for ecosystem productivity and are used as a nutrient source by higher trophic levels. Any changes in microbial community productivity and nutrient cycling activities will have a direct influence on other trophic levels.
In freshwater ecosystems, the majority of microbial activity is associated with surfaces. Biofilms form at a specific location and respond to local environmental conditions, such as current velocity, nutrients (25), and pollutants (26, 27), which makes them ideal resident communities for environmental monitoring. They also regulate the physical and chemical microhabitat and contribute to ecosystem processes as a whole (28). In addition, there is recruitment of populations to form the biofilm that represents the selection pressure for those conditions and locations (29). Therefore, the biofilm community is more representative than the planktonic community of what would be happening within a region of the river and in/on the sediment in particular. Therefore, the response of river biofilms to contaminant presence could impact biodiversity, ecosystem health, and functioning. The presence of various contaminants was shown previously to influence biofilm aquatic microbial communities, even at concentrations that were generally considered to be subinhibitory (26, 27).
Our objective was to observe, under controlled in vitro conditions, the type of biofilm communities that would develop from sediments taken from various parts of the Athabasca watershed, which would be indicative of the impact of the variation in bituminous compounds on microbial communities. To do this, sediments from various locations were diluted in Athabasca River water taken from a reference location and recirculated in rotating annular reactors, and the biofilm architecture, community composition, and activities were measured after 8 weeks. A large array of techniques was used to describe the functional and biological differences in microbial communities. We hypothesized that the hydrocarbons contained in bitumen would be inhibitory for microbial growth, but that dilution in the reactor would decrease this inhibition and provide a supplementary carbon source for adapted microbial communities.
MATERIALS AND METHODS
Sediment and water samples.
Sediment and water for reactor set-up (see below) were sampled on 23 September 2010. A single 10-kg sediment sample (top 15 cm of the sediment) was taken from each of the sites (12 samples) using a stainless steel shovel in less than 1 m of water. See Yergeau et al. (9) for a map of the sampling sites. Three sampling sites were chosen for the Athabasca River: a reference site near Fort McMurray (AR; 56.72°N, 111.40°W; 19.41 km from the nearest tailings pond), one site upstream of the oil sands mining activities (US; 56.87°N, 111.44°W; 4.69 km), and one site directly downstream of Suncor mining activities (DS; 57.06°N, 111.52°W; 6.90 km). Nine other sampling sites were chosen in three of the Athabasca tributaries: upper (EU; 57.23°N, 111.89°W; 22.69 km), middle (EM; 57.24°N, 111.77°W; 19.41 km), and lower (EL; 57.26°N, 111.72°W; 21.12 km) Ell's River (just outside the mining area); upper (FU; 57.34°N, 110.48°W; 73.15 km), middle (FM; 57.44°N, 110.89°W; 59.67 km), and lower (FL; 57.52°N, 111.11°W; 57.75 km) Firebag River (outside the mining area); upper (SU; 56.86°N, 111.13°W; 16.16 km), middle (SM; 56.99°N, 111.34°W; 10.69 km), and lower (SL; 57.02°N, 111.47°W; 10.80 km) Steepbank River (directly in the mining area). Sediment samples were collected, transferred to plastic bags, transported in coolers to the laboratory within 48 h, and kept at 4°C in the dark until they were used in reactor set-up (14 days). Water for the experiment was taken at the reference site near Fort McMurray (AR). The chemical characteristics of the water were the following: dissolved organic carbon (DOC), 4.5 mg/liter; dissolved inorganic carbon, 28.2 mg/liter; SO4, 34.6 mg/liter; Ca, 35.7 mg/liter; Mg, 9.87 mg/liter; K, 1.15 mg/liter; Cl, 3.22 mg/liter; conductivity, 0.298 μS; pH, 8.61; O2, 9.05 mg/liter; NH3, 0 mg/liter. Examination of an extensive database on reference sites in the Athabasca indicates that median concentrations were 5, 13, and 26 μg liter−1 total phosphorus and 2, 2, and 8 μg liter−1 total dissolved phosphorus for the upper, middle, and lower Athabasca River, respectively. Total nitrogen concentrations at reference sites were 200 and 361 μg liter−1 for the middle and lower Athabasca River, respectively (10). The reference site is considered to be both nitrogen and phosphorus limited based on periphyton assessments (30). The water used for dilution (AR site) contained negligible or undetectable levels of the various hydrocarbon indicators.
Experimental design and sampling.
Rotating annular reactors containing 450 ml were set up as previously described (31, 32). Sediment slurry (500 ml; >60% solid) was added to a 10-liter container and then filled with Athabasca River reference water (from the AR site; this represented a 1/20 dilution). The overlying water plus sediment was then continuously recirculated through 3 replicate reactors for each treatment (each replicate reactor was supplied independently via a peristaltic pump but from the same reservoir and was randomly placed on the reactor bench) for about 50 days at a rate of 500 ml per day at 22 ± 2°C, giving 36 bioreactors. The reactors were run with continuous illumination (2.1 μE s−1 m−2) as described in Neu et al. (33). Each replicate reactor contained 12 identical coupons, made in-house out of polycarbonate as previously described (32). Each analysis was done on subsamples of three randomly selected replicate biofilm coupons. Confocal laser-scanning microscopic (CLSM) imaging was performed at 5 random locations on transects across a 1-cm2 subsample piece of the biofilm coupons. Subsampling for other analyses (protozoan counts, thymidine incorporation, carbon utilization, and molecular/genomic) was also carried out using randomly selected subsamples from among the 12 identical coupons in each replicate reactor. For DNA-RNA work, biofilm was harvested by scraping off one coupon from each replicate reactor and flash freezing the biomass in liquid nitrogen. Since all replicate reactors were inoculated with the same dilution for each sediment source and concentrations were not measured directly in the reactors, statistical significance cannot be tested. For correlation analyses (see below), we used the same concentrations for the three replicate reactors.
CLSM.
Examination of all stained and control materials was carried out with an MRC 1024 confocal laser-scanning microscope (Zeiss [formerly Bio-Rad], Jena, Germany) attached to a Microphot SA microscope (Nikon, Tokyo, Japan) equipped with a Nikon water-immersible lens (magnification, 10×; numeric aperture, 0.3). Coupons from each of the replicate reactors were cut into 1-cm2 pieces and mounted in small petri dishes using Dow Corning 3140 acid-free silicone (WPI, Inc., Sarasota, FL) and then stained and analyzed according to Lawrence et al. (31). In brief, bacteria were stained with a fluorescent nucleic acid stain (SYTO 9; green), a lectin probe (Triticum vulgaris-TRITC [tetramethyl rhodamine isothiocyanate]; red) was used to visualize exopolymer, and autofluorescence was used to detect algal and cyanobacterial cells (blue) as described previously in detail (34). The lectins Arachis hypogaea, Canavalia ensiformis, Glycine max, Triticum vulgaris, and Ulex europaeus conjugated to fluorescein isothiocyanate (FITC) were applied individually for in situ analyses of polymer composition, as described by Neu et al. (35). Digital image analysis of the CLSM optical thin sections was performed by using NIH Image version 1.61 (http://rsb.info.nih.gov/nih-image/) with macros written for semiautomated quantification as described in Manz et al. (36). This allowed determination of biofilm thickness, bacterial cell area (biomass), exopolymer biomass, cyanobacterial biomass, and total photosynthetic biomass at various depths. Three-color red-green-blue projections (bacterial cells, green; polymer, red; algal autofluorescence, blue) of the biofilms were computed. Based on controls, there was no background detectable that could be interpreted as hydrocarbon, and the threshold setting for algal and cyanobacterial autofluorescence limits detection of weaker signals.
Protozoan and micrometazoan enumeration, carbon utilization pattern, and thymidine incorporation.
Protozoa and micrometazoa were enumerated as follows. Samples were removed from the reactors on a weekly basis, and the numbers of protozoa and micrometazoa were manually counted on replicate 2-cm2 subsamples using phase-contrast microscopy. The abundance values presented are the cumulative abundance over the course of the 8 weeks of the experiment. Carbon utilization patterns were determined for biofilm samples using commercial Eco-plates (Biolog, Hayward, CA) as described previously (37). Thymidine incorporation was carried out using tritiated thymidine by following the standard protocol of Robarts and Wicks (38). All negative controls were killed with formaldehyde at a 0.4% final concentration.
Nucleic acid extraction and rRNA subtraction.
Frozen biofilm strips were thawed in 2 volumes (approximately 5 ml) of RNAlater (Ambion, Life Technologies, Burlington, Ontario, Canada) and centrifuged at 8,000 × g for 12 min. The DNA and RNA from resulting cell pellets were simultaneously extracted using a homemade bead-beating protocol with subsequent phenol-chloroform purification, as previously described (39, 40). Extracts were separated in two parts, one was treated with DNase to produce RNA and the second was treated with RNase to produce DNA. Total RNA was amplified using the Message Amp kit (Ambion), and rRNA was subtracted by following the protocol described by Stewart et al. (41), with the exception that the T7 promoter was coupled to the forward primer instead of the reverse primer because the MessageAmp procedure produces antisense RNA.
Ion Torrent 16S rRNA gene sequencing.
Partial 16S rRNA gene amplicons were produced using the universal primers E786 (5′-GATTAGATACCCTGGTAG-3′) and U926 (5′-CCGTCAATTCCTTTRAGTTT-3′) (42) containing the 10-bp multiplex identifiers (MID) and adaptor sequences for Ion Torrent sequencing described previously (9, 43). Reactions were performed in 25-μl volumes containing 1 μl of template DNA, 0.3 μM each primer, 0.4 mg/ml of bovine serum albumin (BSA), 0.2 mM deoxynucleoside triphosphates (dNTPs), and 0.05 U/μl of rTaq DNA polymerase (GE Healthcare, Baie d'Urfé, Canada). Cycling conditions involved an initial 5-min denaturing step at 95°C, followed by 25 cycles of 30 s at 95°C, 30 s at 55°C, and 45 s at 72°C and a final elongation step of 3 min at 72°C. All PCR products were purified on agarose gels using the QIAquick gel extraction kit (Qiagen, Valencia, CA) and quantified using the PicoGreen double-stranded DNA quantitation assay (Invitrogen, Carlsbad, CA). All of the 36 amplification products from the different samples were pooled in an equimolar ratio and sequenced together. A total of 3.50 × 107 molecules were used in an emulsion PCR using the Ion OneTouch 200 template kit (Life Technologies) and the OneTouch instruments (Life Technologies) according to the manufacturer's protocol. The sequencing of the pooled library was done using the Personal Genome Machine (PGM) system and a 314 chip with the Ion Sequencing 200 kit according to the manufacturer's protocol. Sequences were binned by MID, after which MIDs were trimmed from each sequence. Sequences having an average Phred quality score below 20 or a length of less than 100 bp were then filtered out of the data set. Taxonomic identities were assigned to sequences using the “multiclassifier,” which is the local multisample version of the RDP Pipeline Classifier (http://pyro.cme.msu.edu/). Weight-normalized Unifrac distances between each sample pair were calculated using the FastUnifrac website (44) based on the GreenGene core data set. For operational taxonomic unit (OTU) calculation, sequence data were randomly normalized to 1,964 sequences and then denoised using the procedure of Quince et al. (45).
RNA sequencing.
Total rRNA-subtracted RNA was reverse transcribed using the SuperScript III kit (Invitrogen). Illumina libraries were prepared by following the protocol of Meyer and Kircher (46), with indices 7 to 42 pooled and sent for three lanes of Illumina HiSeq 2000 paired-end 2 × 101 bp sequencing at The Center for Applied Genomics of the Hospital for Sick Children, Toronto, Canada. Resulting data were split into 216 files (36 samples times 2 reads times 3 lanes). Data from the different lanes were pooled, and the resulting 72 files were filtered using a custom-made Perl script. Sequences were trimmed to the first occurrence of an undefined base (N) or of a low-quality base (below a Phred-like score of 20), and sequences shorter than 75 bp were filtered out. The resulting high-quality sequences were submitted to MG-RAST 3.0 (47) for automated annotation. No attempt was made to assemble the paired reads, as overlaps were not observed. Annotation results from the two paired files were summed after preliminary analyses that revealed no significant differences between the paired data sets. We also looked more specifically at key genes related to aerobic hydrocarbon utilization: alkane hydroxylases and ring-opening dioxygenases. For alkane hydroxylase, we used the MG-RAST “all annotations” functionality with the GenBank database and summed the abundance of reads matching either “alkane monooxygenase” or “alkane hydroxylase.” For the ring-opening dioxygenase, we used the MG-RAST “hierarchical classification” functionality with the M5NR database and summed the abundance of reads for three categories of ring-opening dioxygenases. The intradiol ring-opening dioxygenase represents a sum of the following MG-RAST functions: catechol 1,2-dioxygenase (EC 1.13.11.1), catechol 1,2-dioxygenase 1 (EC 1.13.11.1), intradiol ring-cleavage dioxygenase (EC 1.13.11.1), protocatechuate 3,4-dioxygenase alpha chain (EC 1.13.11.3), and protocatechuate 3,4-dioxygenase beta chain (EC 1.13.11.3). The extradiol ring-opening dioxygenase represents a sum of the following MG-RAST functions: catalytic subunit of meta cleavage enzyme, catechol 2,3-dioxygenase (EC 1.13.11.2), extradiol dioxygenase large subunit, protocatechuate 4,5-dioxygenase alpha chain (EC 1.13.11.8), protocatechuate 4,5-dioxygenase beta chain (EC 1.13.11.8), biphenyl-2,3-diol 1,2-dioxygenase (EC 1.13.11.39), and 1,2-dihydroxynaphthalene dioxygenase. The gentisate/homogentisate ring-opening dioxygenase represents a sum of the following MG-RAST functions: gentisate 1,2-dioxygenase (EC 1.13.11.4) and homogentisate 1,2-dioxygenase (EC 1.13.11.5).
Data analysis.
All statistical analyses were carried out in R (v 2.13.2; The R Foundation for Statistical Computing). Normal distribution and variance homogeneity of the data were tested using the “shapiro.test” and “bartlett.test” functions, respectively. If the data were not normally distributed or did not show homogeneous variance, they were log transformed before analyses of variance (ANOVA). ANOVA were carried out using the “aov” function, with the post hoc Tukey's honestly significant difference (HSD) tests being performed with the “TukeyHSD” function. Spearman rank-order correlations were carried out using the “cor” and “cor.test” functions. Similarity matrices were calculated using the “vegdist” function of the “vegan” package using “Bray” for Biolog substrate utilization patterns and for mRNA functional classification. These matrices and the Unifrac matrix were then used for principal coordinate analyses (PCoA) that were carried out using the “pcoa” function of the “ape” package. Vectors of summed relative abundance at the phylum/class level, substrate, or SEED functional hierarchy were overlaid on the ordinations.
Nucleotide sequence accession numbers.
The sequence data determined in the course of this work have been submitted to the NCBI SRA database under accession no. SRP026549 (NCBI BioProject PRJNA210445). The annotated data sets are available in MG-RAST under accession numbers 4482884 to 4482889, 4482995 to 4483000, 4483004 to 4483009, 4483079 to 4483084, 4483471 to 4483476, 4483480 to 4483485, 4483488 to 4483493, 4483780 to 4483785, 4483791 to 4483796, 4483821 to 4483824, 4483827, 4483829, 4483835, 4483841 to 4483845, 4483870 to 4483873, 4483881, and 4483882.
RESULTS
Sediment chemical analyses.
A subset of the sediments analyzed in Yergeau et al. (9) was diluted in Athabasca River water (1:20) and used to inoculate rotating annular reactors. Since the same sediment was recirculated in triplicate reactors and hydrocarbon concentrations were not measured directly in the recirculated water, chemical data cannot be tested for significance. The sediments had very different concentrations of bituminous compounds, with Ell's River lower, Firebag River lower, and Steepbank River middle and lower sediment samples generally having values one or more orders of magnitude higher than other samples. In these samples with high levels of bituminous compounds, up to 12.76% of the total organic matter was due to the presence of total petroleum hydrocarbons (TPH) (Table 1). In contrast, the samples showing the lowest values for bituminous compounds generally had less than 1% of their organic matter being related to hydrocarbons (Table 1).
Table 1.
Organic component concentrations of the sediments used in the biofilm reactorsa
Sediment source | Organic matter (%, wt/wt) | [TPH] (μg/g) | TPH/total C (%) | [TSH] (μg/g) | [TAH] (μg/g) | [NA] (mg/ml) | [EPA PAH] (ng/g) | Concn. of total alkylated PAHs (ng/g) | Concn. of total aromatic compounds (ng/g) | Concn. of total n-alkanes (ng/g) |
---|---|---|---|---|---|---|---|---|---|---|
Athabasca River | ||||||||||
Reference site | 2.10 | 251.3 | 1.20 | 104.0 | 147.3 | ND | 1,176.3 | 1,235.3 | 1,406.3 | 5,831.7 |
Suncor upstream | 2.20 | 104.9 | 0.48 | 52.2 | 52.7 | 1 | 188.6 | 188.2 | 248.6 | 4,241.7 |
Suncor downstream | 1.35 | 68.5 | 0.51 | 32.6 | 35.9 | ND | 127.7 | 130.9 | 172.8 | 1,014.3 |
Tributaries | ||||||||||
Ell's River upper | 3.18 | 66.3 | 0.21 | 35.1 | 31.2 | 1 | 218.7 | 197.7 | 299.5 | 2,252.2 |
Ell's River middle | 1.06 | 139.6 | 1.32 | 69.5 | 70.1 | 3 | 433.6 | 677.1 | 711.9 | 660.1 |
Ell's River lower | 2.19 | 2,792.4 | 12.76 | 1,638.4 | 1,154.0 | 4 | 9,497.4 | 14,661.6 | 15,138.0 | 2,111.3 |
Firebag River upper | 2.49 | 34.1 | 0.14 | 19.6 | 14.5 | 1 | 35.9 | 33.6 | 39.9 | 3,576.2 |
Firebag River middle | 12.96 | 129.1 | 0.10 | 87.1 | 42.0 | 1 | 89.4 | 73.0 | 100.6 | 7,815.9 |
Firebag River lower | 5.05 | 1,995.4 | 3.96 | 977.9 | 1,017.5 | 3 | 854.2 | 762.5 | 932.0 | 1,703.5 |
Steepbank River upper | 5.53 | 121.2 | 0.22 | 52.7 | 68.5 | 4 | 82.3 | 105.4 | 125.7 | 4,522.2 |
Steepbank River middle | 3.69 | 840.3 | 2.28 | 420.1 | 420.2 | 0 | 5,854.3 | 8,925.4 | 9,525.8 | 1,146.1 |
Steepbank River lower | 4.26 | 3,879.8 | 9.12 | 1,947.7 | 1,932.1 | 3 | 9,209.7 | 13,348.1 | 13,866.6 | 16,570.7 |
All values except NA are per dry weight of sediments. TPH, total petroleum hydrocarbon; TSH, total straight chain hydrocarbon; TAH, total aromatic hydrocarbon; NA, naphthenic acids; EPA-PAH, U.S. Environmental Protection Agency 16 priority polycyclic aromatic hydrocarbon; PAH, polycyclic aromatic hydrocarbon. ND, not determined. Data are modified from Yergeau et al. (9).
Biofilm architecture.
Biofilms were 29.0% thinner in the samples exposed to the sediments containing the highest concentration of hydrocarbon (black bars) than in the other treatments (white and gray bars), and many of the pairwise differences were significant (Fig. 1). Consequently, biofilm thickness was significantly and negatively correlated to most of the hydrocarbon indicators, with stronger correlations (r of up to −0.77; P < 0.001) to EPA-PAH, alkylated PAHs, and aromatic compounds (Table 2). The composition of the extracellular polysaccharide (EPS) matrix varied considerably between the different treatments based on lectin binding assays (Fig. 1). Glc(NAc)2 (N-acetylglucosamine) and NeuNAc (N-acetylneuraminic acid) were 35.9% less abundant in the samples exposed to the sediments containing the highest concentration of hydrocarbon (black bars) than in the other treatments (white and gray bars) (Fig. 1), resulting in significant (P < 0.05) negative correlations with concentrations of naphthenic acids, EPA-PAH, alkylated PAHs, and aromatic compounds (Table 2). In contrast, terminal α- or β-linked N-acetylgalactosamine (terminal α or β galNAc) was 207.6% more abundant in the samples exposed to the sediments containing the highest concentration of hydrocarbon (black bars) than in the other treatments (white and gray bars) (Fig. 1), resulting in significant (P < 0.01) positive correlation with TPH, TSH, and TAH (Table 2).
Fig 1.
Biofilm thickness and composition of the extracellular polysaccharide (EPS) matrix based on lectin binding assays for biofilm grown in rotating reactors inoculated with sediments from the Athabasca River and its tributaries. Values are means of triplicate measurements, and different letters indicate significant differences in Tukey's HSD post hoc tests (at P < 0.05). α man, α mannose; α glc, α glucose; glc(NAc)2, N-acetylglucosamine; neuNAc, N-acetylneuraminic acid; terminal α or β galNAc, terminal α- or β-linked N-acetylgalactosamine; terminal β gal, terminal β-galactose; galNAc, N-acetylgalactosamine. Sediment [TPH] bar colors: white fill, [TPH] < 125 μg/g; gray fill, 125 μg/g < [TPH] < 750 μg/g; black fill, [TPH] > 750 μg/g.
Table 2.
Correlation coefficients between sediment chemical concentrations and biofilm biological dataa
Biofilm component | TPH | TSH | TAH | NA | EPA-PAH | Alkylated PAHs | Aromatic compounds | n-alkanes |
---|---|---|---|---|---|---|---|---|
Protozoa | −0.71 | −0.69 | −0.74 | −0.74 | −0.51 | −0.56 | −0.56 | −0.01 |
Algae | −0.60 | −0.55 | −0.66 | −0.24 | −0.72 | −0.75 | −0.75 | 0.06 |
Cyanobacteria | −0.33 | −0.32 | −0.33 | −0.02 | −0.38 | −0.40 | −0.40 | −0.06 |
Bacteria | −0.50 | −0.47 | −0.54 | −0.22 | −0.63 | −0.66 | −0.66 | 0.18 |
glc(NAc)2, neuNAc | −0.22 | −0.17 | −0.28 | −0.42 | −0.39 | −0.43 | −0.43 | 0.25 |
Terminal β gal, galNAc | −0.08 | −0.08 | −0.09 | 0.05 | −0.07 | −0.09 | −0.09 | 0.05 |
α man, α glc | 0.18 | 0.17 | 0.21 | 0.26 | 0.38 | 0.37 | 0.37 | −0.25 |
Terminal α or β galNAc | 0.50 | 0.52 | 0.45 | −0.01 | 0.29 | 0.27 | 0.27 | 0.11 |
α−l−fucose | 0.35 | 0.36 | 0.28 | −0.17 | 0.24 | 0.21 | 0.21 | 0.05 |
Thickness | −0.54 | −0.50 | −0.59 | −0.36 | −0.74 | −0.77 | −0.77 | 0.19 |
Bacterial productivity | −0.17 | −0.15 | −0.25 | −0.15 | −0.36 | −0.41 | −0.41 | −0.23 |
Chlorophyll a | −0.28 | −0.25 | −0.33 | −0.17 | −0.55 | −0.58 | −0.58 | 0.15 |
mRNA | ||||||||
Photosynthesis | −0.31 | −0.30 | −0.31 | −0.03 | −0.47 | −0.48 | −0.48 | −0.14 |
Alkane monooxygenase | 0.06 | 0.05 | 0.07 | 0.06 | 0.05 | 0.07 | 0.07 | 0.19 |
Cytochrome P450 | −0.19 | −0.25 | −0.20 | 0.06 | −0.19 | −0.21 | −0.21 | −0.36 |
Extradiol | 0.31 | 0.31 | 0.37 | 0.31 | 0.37 | 0.40 | 0.40 | 0.25 |
Intradiol | 0.29 | 0.30 | 0.27 | 0.38 | 0.32 | 0.33 | 0.33 | 0.21 |
Gentisate/homogentisate | 0.27 | 0.24 | 0.28 | 0.36 | 0.16 | 0.20 | 0.20 | 0.21 |
Sum aromatic degradation genes | 0.41 | 0.39 | 0.42 | 0.49 | 0.38 | 0.41 | 0.41 | 0.23 |
Underline, P value below 0.05; italics, P value below 0.01; boldface, P value below 0.001. glc(NAc)2, N-acetylglucosamine; neuNAc, N-acetylneuraminic acid; terminal β gal, terminal β-galactose; GalNAc, N-acetylgalactosamine; α man, α mannose; α glc, α glucose; terminal α or β GalNAc, terminal α- or β-linked N-acetylgalactosamine. Intradiol represents a sum of the following MG-RAST functions: catechol 1,2-dioxygenase (EC 1.13.11.1), catechol 1,2-dioxygenase 1 (EC 1.13.11.1), intradiol ring-cleavage dioxygenase (EC 1.13.11.1), protocatechuate 3,4-dioxygenase alpha chain (EC 1.13.11.3), and protocatechuate 3,4-dioxygenase beta chain (EC 1.13.11.3). Extradiol represents a sum of the following MG-RAST functions: catalytic subunit of meta cleavage enzyme, catechol 2,3-dioxygenase (EC 1.13.11.2), extradiol dioxygenase large subunit, protocatechuate 4,5-dioxygenase alpha chain (EC 1.13.11.8), protocatechuate 4,5-dioxygenase beta chain (EC 1.13.11.8), biphenyl-2,3-diol 1,2-dioxygenase (EC 1.13.11.39), and 1,2-dihydroxynaphthalene dioxygenase. Gentisate/homogentisate represents a sum of the following MG-RAST functions: gentisate 1,2-dioxygenase (EC 1.13.11.4) and homogentisate 1,2-dioxygenase (EC 1.13.11.5).
Biofilm community composition.
The biofilm community composition was analyzed by Ion Torrent 16S rRNA gene sequencing, direct counts, and confocal laser-scanning microscopy. The microbial communities exposed to the sediments having the highest concentrations of hydrocarbons (Ell's lower, Firebag lower, and Steepbank middle and lower) were visually very different from the other biofilms when observed by confocal laser-scanning microscopy (Fig. 2). Similar differences were found upon analysis of each biofilm for relative abundance of algae, Cyanobacteria, and Bacteria using confocal laser-scanning microscopy (Fig. 3). Cyanobacteria, algae, and Bacteria were 67.7, 63.6, and 67.2% less abundant, respectively, in the samples exposed to the sediments containing the highest concentrations of hydrocarbon (black bars) compared to the other treatments (white and gray bars) (Fig. 3). Significant negative correlations were observed between algae, Cyanobacteria, and Bacteria and various hydrocarbon indicators (P < 0.001, P < 0.05, and P < 0.01, respectively), with the strongest correlations being with EPA-PAH, alkylated PAHs, and aromatic compounds (Table 2). The cumulative number of protozoa observed over the course of the 8-week experiment (with observation every week) was also 13.4% lower in the samples exposed to the sediments containing the highest concentration of hydrocarbon (black bars) than in the other treatments (white and gray bars) (Fig. 3). Protozoan counts were also negatively correlated to most of the hydrocarbon indicators measured (P < 0.01), with total aromatic hydrocarbon, total petroleum hydrocarbon, and naphthenic acid concentrations having the strongest correlations (Table 2).
Fig 2.
Montage of confocal laser-scanning microscopy images of biofilm grown in rotating reactors inoculated with sediments from the Athabasca River and its tributaries. Images were selected based on their similarity to the mean results obtained for algal, bacterial, and cyanobacterial biomass for each treatment, so they are the images that best represent the results obtained for any treatment via image analyses of the stack. Bacteria, green; exopolymer, red; cyanobacteria and algae, blue.
Fig 3.
Amount of bacterial, cyanobacterial, and algal biofilm components based on confocal laser-scanning microscopy imaging and numbers of protozoa based on direct counts over a period of 8 weeks for biofilm grown in rotating reactors inoculated with sediments from the Athabasca River and its tributaries. Values are means from triplicate measurements, and different letters indicate significant differences in Tukey's HSD post hoc tests (at P < 0.05). Sediment [TPH] bar colors: white fill, [TPH] < 125 μg/g; gray fill, 125 μg/g < [TPH] < 750 μg/g; black fill, [TPH] > 750 μg/g.
The bacterial community, as measured by Ion Torrent 16S rRNA gene sequencing, was dominated by Proteobacteria, with the Alphaproteobacteria and Betaproteobacteria classes being highly abundant in most treatments (19.6 to 36.0% and 14.7 to 43.8% of the total community, respectively) (Fig. 4a; also see Table S1 in the supplemental material). Most of the phyla/classes were significantly (P < 0.05) affected by the experimental treatments, with the exception of the Cyanobacteria (see Table S1). The Alphaproteobacteria and the Betaproteobacteria were often significantly more abundant in the biofilms exposed to sediments containing the highest hydrocarbon concentrations (Alphaproteobacteria, 29.0% [high] versus 23.3% [medium-low]; Betaproteobacteria, 32.4% versus 28.9%) (Fig. 4a; also see Table S1 in the supplemental material) and were positively correlated to several hydrocarbon indicators (P < 0.01 and P < 0.001, respectively), with the highest coefficient being for EPA-PAH, alkylated PAHs, and aromatic compounds (Table 3). The Gammaproteobacteria class was significantly more abundant in two treatments, downstream of Suncor and Steepbank River upper (29.0 and 27.2% of the total community, respectively) (Fig. 4a; also see Table S1), but only showed a significant positive correlation (P < 0.01) with naphthenic acid concentration (Table 3). The Bacteroidetes were significantly more abundant in the biofilms exposed to sediments from Firebag River (15.7 to 23.2% of the total community) (Fig. 4a; also see Table S1), which resulted in significant negative correlations (P < 0.001) with EPA-PAH, alkylated PAHs, and aromatic compounds (Table 3). The Actinobacteria were significantly more abundant in the biofilms that developed from Firebag River lower and upper, Ell's River middle, and Steepbank River upper sediments (6.8 to 9.3%) (see Table S1), which resulted in a significant positive correlation (P < 0.01) with naphthenic acids and negative correlations (P < 0.001) with EPA-PAH, alkylated PAHs, and aromatic compounds (Table 3).
Fig 4.
Bacterial community composition based on Ion Torrent 16S rRNA gene amplicon (∼120 bp) sequencing (a) or on the taxonomic affiliation of the sequenced mRNA (b) for biofilm grown in rotating reactors inoculated with sediments from the Athabasca River and its tributaries. For each treatment, means from triplicates are presented. See Tables S1 and S2 in the supplemental material for ANOVA and Tukey test results.
Table 3.
Spearman correlation coefficients between the most abundant phylum/class and sediment chemical concentrationsa
Genetic material | TPH | TSH | TAH | NA | EPA-PAH | Alkylated PAHs | Aromatic compounds | n-alkanes |
---|---|---|---|---|---|---|---|---|
16S rRNA gene | ||||||||
Actinobacteria | −0.30 | −0.30 | −0.29 | 0.46 | −0.51 | −0.47 | −0.47 | −0.09 |
Bacteroidetes | −0.22 | −0.22 | −0.29 | −0.05 | −0.56 | −0.59 | −0.59 | 0.08 |
Cyanobacteria | −0.18 | −0.17 | −0.19 | −0.05 | −0.26 | −0.27 | −0.27 | 0.17 |
Firmicutes | 0.00 | 0.07 | −0.08 | −0.41 | 0.10 | 0.04 | 0.04 | 0.40 |
Alphaproteobacteria | 0.56 | 0.51 | 0.60 | 0.29 | 0.70 | 0.73 | 0.73 | −0.48 |
Betaproteobacteria | 0.21 | 0.24 | 0.25 | −0.26 | 0.55 | 0.56 | 0.56 | 0.04 |
Gammaproteobacteria | 0.07 | 0.04 | 0.10 | 0.52 | −0.15 | −0.11 | −0.11 | 0.14 |
Deltaproteobacteria | 0.21 | 0.24 | 0.13 | −0.08 | 0.05 | 0.01 | 0.01 | −0.03 |
mRNA | ||||||||
Actinobacteria | −0.04 | −0.05 | 0.02 | 0.37 | −0.28 | −0.23 | −0.23 | 0.17 |
Bacteroidetes | 0.08 | 0.07 | 0.00 | 0.09 | −0.12 | −0.14 | −0.14 | 0.36 |
Cyanobacteria | −0.27 | −0.27 | −0.28 | 0.03 | −0.56 | −0.57 | −0.57 | 0.05 |
Firmicutes | −0.34 | −0.32 | −0.39 | −0.05 | −0.52 | −0.54 | −0.54 | −0.16 |
Alphaproteobacteria | 0.30 | 0.29 | 0.30 | −0.16 | 0.60 | 0.60 | 0.60 | −0.12 |
Betaproteobacteria | 0.26 | 0.27 | 0.31 | −0.21 | 0.58 | 0.59 | 0.59 | 0.06 |
Gammaproteobacteria | 0.25 | 0.21 | 0.26 | 0.24 | 0.26 | 0.29 | 0.29 | 0.01 |
Deltaproteobacteria | −0.10 | −0.18 | −0.09 | −0.03 | 0.18 | 0.17 | 0.17 | −0.39 |
Underline, P value below 0.05; italics, P value below 0.01; boldface, P value below 0.001.
The relationships between phyla/classes and samples mentioned above were also visible in the principal coordinate analyses of the Unifrac distance onto which the phylum/class data were projected. The ordination also revealed that the biofilms exposed to sediments from Firebag River were harboring similar bacterial communities, with relatively more Actinobacteria (average for Firebag River of 5.04% versus an average for other samples of 4.29%), Deltaproteobacteria (2.42% versus 1.79%), and Bacteroidetes (12.59% versus 8.70%) than other treatments (Fig. 5a). These samples were also relatively richer in Acidobacteria (1.58% versus 1.39%) and Cyanobacteria (0.06% versus 0.02%) and poorer in Betaproteobacteria (29.02% versus 33.22%) and Alphaproteobacteria (25.73% versus 25.96%) (Fig. 5a). In contrast, most of the biofilms exposed to the lower Ell's River and the Steepbank River sediments were richer in Betaproteobacteria (average lower Ell's River and the Steepbank River of 34.27% versus an average for the other samples of 27.97%), Alphaproteobacteria (27.70% versus 24.00%), and Gammaproteobacteria (17.76% versus 17.28%) (Fig. 5a). The remaining biofilms were relatively enriched in Planctomycetes (average for other samples of 1.54% versus an average for lower Ell's River and the Steepbank River of 0.90%), Firmicutes (2.32% versus 1.39%), and Cyanobacteria (0.07% versus 0.005%).
Fig 5.
Principal coordinate analysis (PCoA) of the community composition based on 16S rRNA gene amplicon sequencing (Unifrac values) (A), Biolog assay color intensity (B), and the functional composition based on mRNA sequencing (C) for biofilm grown in rotating reactors inoculated with sediments from the Athabasca River and its tributaries. Arrows indicate samples having the highest values for that variable and were added as supplementary variables; they are not involved in the calculation of the ordination. Abbreviations for panel B are the following: pol, polymer, sum of the color intensity for α-cyclodextrin, Tween 40, and Tween 80; cb, carbohydrate, sum of the color intensity for d-xylose, I-erythritol, glycogen, ß-methyl-d-glucoside, N-acetyl-d-glucosamine, d-cellobiose, α-d-lactose, and d-mannitol; ca, carboxylic acid, sum of the color intensity for 2-hydroxy benzoic acid, α-keto butyric acid, itaconic acid, d-malic acid, d-galactonic acid γ-lactone, d-glucosaminic acid, 4-hydroxy benzoic acid, γ-hydroxybutyric acid, and d-galacturonic acid; aa, amino acid, sum of the intensity for l-threonine, glycyl-l-glutamic acid, l-phenylalanine, l-serine, l-arginine, and l-asparagine; am, amine, sum of color intensity for putrescine and phenylethylamine; ph, phosphorylated, sum of the color intensity for d,l-α-glycerol phosphate and glucose-1-phosphate; es, ester, color intensity for pyruvic acid methyl ester.
We compared the microbial communities of the biofilms to the ones detected in a previous microbial field survey of the sediments used in this study (9). When looking at each phylum/class individually, no significant Spearman correlations were found between the biofilm and sediments, suggesting that, across the different samples, the relative abundance of particular taxa in the sediment was not predictive of its relative abundance in the biofilm. When comparing the relative abundance of all phyla/classes across pairs of samples, significant Spearman correlations were found between sediments and their associated biofilm and between almost all biofilm-sediment pairs, indicating that the relative community composition at the phylum/class level was not more similar between a sediment sample and its associated biofilm than a random biofilm-sediment pair. Similarly, in terms of Unifrac distances, the biofilm bacterial communities were not more similar to their associated sediment bacterial communities than to other sediment bacterial communities. Altogether, these results indicate that biofilm microbial communities could not be predicted from the sediment microbial communities, and that some selection of microorganisms had occurred because not all microorganisms present in the sediment can live in a biofilm.
The number of OTUs detected in the microbial communities was highly variable among replicate samples (see Fig. S1 in the supplemental material). This variability precluded the identification of any significant differences between the treatments. The number of OTUs was generally negatively correlated with the different hydrocarbon indicators, but again, none of these trends were significant at P < 0.05.
Biofilm activities.
Biofilm activities were measured by bacterial thymidine incorporation, chlorophyll a measurement, Biolog assays, and Illumina mRNA sequencing. Bacterial production, as measured by thymidine incorporation rates, was 206.6% higher in the biofilms exposed to sediments from Firebag River (middle and lower) and Ell's River (middle) (Fig. 6), which were not among the sediments having the lowest hydrocarbon concentrations. However, significant (P < 0.05) negative correlations were observed between bacterial production and different hydrocarbon compounds, with the strongest negative correlations being with the concentration of total aromatic compounds, EPA-PAHs, and alkylated PAHs (Table 2).
Fig 6.
Bacterial productivity, chlorophyll a content, and relative abundance of mRNA related to photosynthesis for biofilm grown in rotating reactors inoculated with sediments from the Athabasca River and its tributaries. Values are means from triplicate measurements, and different letters indicate significant differences in Tukey's HSD post hoc tests (at P < 0.05). Sediment [TPH] bar colors: white fill, [TPH] < 125 μg/g; gray fill, 125 μg/g < [TPH] < 750 μg/g; black fill, [TPH] > 750 μg/g.
The active community composition, based on the taxonomic affiliation of the sequenced mRNA in MG-RAST, was dominated by Proteobacteria, mainly the Alphaproteobacteria, Betaproteobacteria, and Gammaproteobacteria classes (12.9 to 20.6%, 11.7 to 19.2%, and 17.1 to 22.8% of total activity, respectively) (Fig. 4b; also see Table S2 in the supplemental material). The Firmicutes and Cyanobacteria were much more active than expected based on their relative abundance in 16S rRNA gene libraries (10.4 to 17.2% versus 0.9 to 4.8% and 3.6 to 13.4% versus 0.0 to 0.4%, respectively) (Fig. 4a versus b). The Cyanobacteria were significantly less active (by 56.4%) when exposed to the sediments from the reference site, the lower Ell's River, and the middle and lower Steepbank River (Fig. 4b; also see Table S2) and showed significant negative correlations (P < 0.001) with EPA-PAH, alkylated PAHs, and aromatic compound concentrations (Table 3). The activities of Firmicutes were also significantly (P < 0.05) and negatively correlated with TPH, TAH, EPA-PAH, alkylated PAHs, and aromatic compound concentrations, even though most of the differences between treatments were not significant (Table 3). Alphaproteobacteria and Betaproteobacteria were 7.6 and 21.3% more active, respectively, in biofilms exposed to sediments containing higher concentrations of hydrocarbons (black lines) than the other treatments (white and gray lines) (Fig. 4b; also see Table S2), which resulted in significant positive correlations (P < 0.001) between the activities of these two groups and EPA-PAH, alkylated PAHs, and aromatic compound concentrations (Table 3). The relative activity of Actinobacteria was significantly (P < 0.05) and positively correlated with the concentration of naphthenic acids (Table 3).
Chlorophyll a and photosynthesis-related transcripts were 33.3 and 19.5% less abundant, respectively, in the samples exposed to the sediments containing the highest concentration of hydrocarbons (black bars) compared to the other treatments (white and gray bars) (Fig. 6). Chlorophyll a was significantly (P < 0.001) and negatively correlated with the concentration of total aromatic compounds, EPA-PAHs, and alkylated PAHs (Table 2). Similarly, the expression of photosynthesis-related genes was significantly (P < 0.01) and negatively correlated with the concentration of total aromatic compounds, EPA-PAHs, and alkylated PAHs (Table 2). A large majority of the Cyanobacteria families were negatively correlated to the concentrations of TPH and total aromatic compounds, while almost as many families were positively and negatively correlated to naphthenic acids and n-alkanes (see Fig. S2 in the supplemental material). Very few phyla/classes showed a clear majority of families positively correlated to TPH, n-alkanes, and aromatic compounds, with only Verrucomicrobia and Actinobacteria having a clear majority of families showing positive correlations (see Fig. S2). In contrast, for naphthenic acids, most of the phyla/classes had a majority of their families positively correlated with NA concentration (see Fig. S2).
The biofilms that developed in the presence of the Firebag River lower sediments degraded a higher percentage of the substrates in Biolog assays and clustered away from most other treatments (Fig. 5b). Within the other treatments, there was clustering of the Steepbank and Ell's River lower samples, which were observed in the lower left quadrant of the PCoA ordination plot (Fig. 5b). The relative abundance of the different functions detected following mRNA sequencing was also used to create an ordination of the samples (Fig. 5c). The biofilms exposed to the Firebag River sediment were clearly distinct from the other samples, and this was partly due to a higher relative abundance of transcripts related to photosynthesis and lower relative abundance of transcripts related to phosphorus, fatty acids, aromatics, secondary metabolism, and stress (Fig. 5c). The biofilms that developed in the presence of sediments from the Steepbank River, lower and upper Ell's River, Athabasca River reference site, and downstream of Suncor clustered together on the left side of the first axis, having relatively higher expression of genes related to fatty acids, aromatics, potassium, secondary metabolism, virulence, iron, and amino acids and relatively lower expression of genes related to photosynthesis, respiration, and sulfur (Fig. 5c). The biofilms exposed to sediments from the upper Steepbank River, the middle Ell's River, and upstream of Suncor clustered together, with relatively more transcripts associated with sulfur metabolism, photosynthesis, respiration, and carbohydrates and fewer related to potassium, virulence, iron, amino acids, and mobile elements (Fig. 5c). For the most part, photosynthesis-related functions were negatively correlated to different hydrocarbon indicators (see Fig. S3 in the supplemental material), while the relative abundance of functions related to potassium were mostly negatively correlated to TPH and naphthenic acids and mostly positively correlated to n-alkanes (see Fig. S3). For the other functional groups (carbohydrates and N, P, Fe, and S metabolism), a majority of functions were positively correlated to TPH, naphthenic acids, n-alkanes, and aromatic compounds (see Fig. S3).
No significant pairwise differences among the treatments were observed for the relative expression of alkane monooxygenase genes (Fig. 7). Similarly, very few significant pairwise differences were observed between treatments for the various classes of aromatic ring-cleaving dioxygenases (Fig. 7). For the extradiol-type dioxygenases, the only significant difference was between the lower Ell's River and Firebag River treatments (Fig. 7). For the intradiol-type dioxygenases, the Suncor upstream treatment induced significantly less expression than the reference site, the lower Ell's River, and the lower Steepbank River treatments (Fig. 7). For the gentisate/homogentisate-type dioxygenases, the lower Steepbank River treatment induced significantly more expression than the Suncor upstream treatment (Fig. 7). The relative expression of alkane monooxygenase genes was not correlated to any of the chemical parameters measured, while the expression of cytochrome P450 alkane hydroxylase (CYP153) was significantly (P < 0.05) and negatively correlated to the concentration of n-alkanes (Table 2). The expression of aromatic ring-cleaving dioxygenase genes was significantly (P < 0.05) and positively correlated to the concentration of a variety of chemical compounds, with the strongest correlations often being observed with EPA-PAH, alkylated PAHs, and aromatic compounds (Table 2).
Fig 7.
Relative abundance of mRNA related to alkane monooxygenase and extradiol-, intradiol-, and gentisate/homogentisate-type aromatic ring-cleaving enzymes for biofilm grown in rotating reactors inoculated with sediments from the Athabasca River and its tributaries. Intradiol represents a sum of the following MG-RAST functions: catechol 1,2-dioxygenase (EC 1.13.11.1), catechol 1,2-dioxygenase 1 (EC 1.13.11.1), intradiol ring-cleavage dioxygenase (EC 1.13.11.1), protocatechuate 3,4-dioxygenase alpha chain (EC 1.13.11.3), and protocatechuate 3,4-dioxygenase beta chain (EC 1.13.11.3). Extradiol represents a sum of the following MG-RAST functions: catalytic subunit of meta cleavage enzyme, catechol 2,3-dioxygenase (EC 1.13.11.2), extradiol dioxygenase large subunit, protocatechuate 4,5-dioxygenase alpha chain (EC 1.13.11.8), protocatechuate 4,5-dioxygenase beta chain (EC 1.13.11.8), biphenyl-2,3-diol 1,2-dioxygenase (EC 1.13.11.39), and 1,2-dihydroxynaphthalene dioxygenase. Gentisate/homogentisate represents a sum of the following MG-RAST functions: gentisate 1,2-dioxygenase (EC 1.13.11.4) and homogentisate 1,2-dioxygenase (EC 1.13.11.5). Values are means from triplicate measurements, and different letters indicate significant differences in Tukey's HSD post hoc tests (at P < 0.05). Sediment [TPH] bar colors: white fill, [TPH] < 125 μg/g; gray fill, 125 μg/g < [TPH] < 750 μg/g; black fill, [TPH] > 750 μg/g.
DISCUSSION
In the present study, biofilms were grown in rotating annular reactors to provide a representation of the periphyton of the Athabasca River and its tributaries. Biofilms form at a specific location in response to local environmental conditions; as such, they make an ideal resident community for environmental monitoring. They also regulate the physical and chemical microhabitat and contribute to ecosystem processes as a whole (28). In addition, there is recruitment of populations to form the biofilm that represents the selection pressure for those conditions and location (29). Therefore, the biofilm community is more representative of what would occur within a region of the river and in/on the sediment in particular than the planktonic community. Thus, we focused our attention on these communities. Our results suggest that variation in bituminous compound content of the river sediments leads to major differences in biofilm microbial communities. In agreement with our initial hypothesis, two mechanisms, namely, inhibitory and increased energy and carbon effects, appear to be at play during the exposure of biofilm microbial communities to bitumen-containing river sediments. First, inhibitory effects were observed, as the river sediments having the highest concentrations of bituminous compounds generally led to the development of less productive biofilms, where photosynthetic organisms were particularly repressed compared to biofilms that developed from river sediments having lower bituminous compound concentrations. Second, within the microbial communities, some taxonomic groups (mainly Proteobacteria) and functional genes (ring-opening dioxygenases) increased their relative abundance and activity with increasing concentrations of hydrocarbon compounds. Thus, for general inhibition of the microbial communities, some microorganisms appeared to be stimulated by higher hydrocarbon concentrations, either because they could take advantage of this carbon source or because they were more tolerant to the inhibitory effect of some bituminous compounds.
Inhibitory effects.
Constant exposure of sediments to higher concentrations of bituminous compounds resulted in a decrease in all of the productivity indicators (thymidine incorporation, chlorophyll a, and photosynthesis gene expression) and reduced biofilm thickness. These results suggest that the hydrocarbons found in the Athabasca River watershed negatively affect the indigenous biofilm-forming microbial communities. Several compounds present in bitumen are inhibitory to organisms. Naphthenic acids (NA) are generally thought to be the most toxic components (48, 49) and were shown to negatively influence not only plants, birds, fish, toads, frogs, and rats (50–54) but also bacteria (14, 15), algae (16, 17), cyanobacteria (18), and protozoa (55). However, in the present study, the strongest negative correlations were often observed with aromatic compounds, while very few significant correlations were found with NA and alkanes, suggesting that the aromatic fraction of bitumen is the most inhibitory or the most available for microorganisms. Aromatic hydrocarbons are recognized to be toxic to a variety of microbial processes. For instance, methanogenesis in sludge was 50% inhibited by a variety of aromatic compounds at concentrations ranging from 3.4 to 57.3 mM (5 mM ≈ 500 ppm) (19). For nitrification in soils, the no-observed-effect concentrations (NOECs) ranged from 22 to 1,100 mg kg−1 (20). However, when directly confronted with aromatic compounds, bacteria appeared to be more sensitive, with half-maximal effective concentrations (EC50) of luminescent bacterial assays for different PAHs ranging from 0.53 to 24.39 ppm (μg ml−1) (21, 22). In contrast, the EC50 of naphthenic acids in luminescent bacterial assays was reported to range from 41.9 to 64.9 ppm (mg liter−1) (14), and even the more toxic aromatic alkanoic naphthenic acids and their degradation products had EC50s ranging from 9.4 to 69.2 ppm (mg liter−1) (13). In the current study, NAs were measured as a class using liquid chromatography-tandem mass spectrometry and no species identification data are available, but the concentrations estimated in the reactors (≤0.2 ppm, assuming a complete dissolution in the recirculated water) are considerably lower than those reported in the toxicological tests described above. In contrast, for aromatic compounds, the concentrations estimated in the reactors (e.g., EPA-PAH at ≈1 ppm, assuming a complete dissolution in the recirculated water) are within the range reported as inhibitory in laboratory assays. This might explain the greater number and the higher correlation coefficients for negative correlations between our microbiological indicators and aromatic compounds compared to those of other hydrocarbon compounds. However, although estimated concentrations of PAH and NA were mostly lower than the levels reported as inhibitory in toxicology tests, a long-term cumulative exposure to a persistent suite of contaminants in the environment might affect microbial communities regardless of relatively low concentration.
The activities of photosynthetic organisms were strongly repressed when exposed to sediments having the highest content of bituminous compounds. A previous field survey of the bacterial and archaeal communities in sediments from the Athabasca River, its tributaries, and oil sand tailings ponds revealed a significant decrease in Cyanobacteria relative abundance with increasing sediment hydrocarbon content (9). Cyanobacteria appear to be particularly sensitive to environmental disturbance, as exposure of reactor-grown biofilms to some pharmaceutical products also resulted in a reduction of the relative abundance of Cyanobacteria and a reduction of the expression of photosynthetic genes (27). Thus, Cyanobacteria could be used as a bio-indicator to monitor the impact of oil sands mining operations. However, this might be confounded by the apparent insensitivity of some algae to PAHs (56). In aquatic ecosystems, Cyanobacteria are involved in major processes (C and N fixation) that are at the base of ecosystem productivity. The lowered photosynthetic activity observed in most biofilms exposed to sediments having higher bituminous compound content could cascade effects on the entire microbial community as well as on higher trophic levels. This is consistent with the generally lower productivity and biomass in these biofilm samples. One indication of this cascade effect on higher trophic levels is the lower abundance of protozoa in reactors inoculated with sediments having high bituminous compound concentrations. However, a reduction in protozoa also results in diminished grazing pressure on the microbial communities (57), which would lead to thicker biofilms if nothing else was influencing the microbial community.
Increased carbon and energy.
The microbial communities present in the Athabasca watershed are constantly exposed to various components of bitumen through erosion of the oil sands by the river flow and atmospheric deposition through the activity of bitumen upgrading (3, 4, 58). One of our hypotheses was that these microbial communities were adapted to this constant flow of organic compounds and were taking advantage of this increased carbon and energy input. The concentration of TPH observed could represent an important carbon and energy source for the microbial community, as in the sediments containing the highest concentrations of TPH, a large portion of the organic matter was due to the presence of bituminous compounds (up to 12.76%) (Table 1). However, in the case of an increased carbon and energy effect, we would have expected the biofilms developed from sediments containing higher concentrations of bituminous compounds to produce more biomass, which was not the case. In fact, increased bituminous concentrations had mainly inhibitory effects on biofilms, with lowered thickness; lower bacterial, cyanobacterial, and algal biomass; and lower bacterial productivity. The decrease in biomass/productivity indicators could also be related to an increased selection pressure for the specialist subset of the microbial community that can use bituminous compounds as carbon sources, which might be less productive than other members of the community. In fact, the relative expression of functional genes related to extradiol ring-opening dioxygenases and the relative abundance and activities of Alphaproteobacteria and Betaproteobacteria was positively correlated with hydrocarbon concentration. Microorganisms of the Athabasca watershed are known to be able to degrade bituminous compounds, many of which belong to the Proteobacteria (6–8). The Betaproteobacteria class and the Proteobacteria phylum were previously shown to be positively correlated to various hydrocarbon indicators in a field survey of sediments from the Athabasca River, its tributaries, and oil sands tailings ponds (9). Alternatively, an increased proportion of bituminous compounds among carbon sources could reduce biofilm productivity, as many bituminous compounds are difficult/recalcitrant substrates.
The stretch of the Athabasca River near the oil sands mining operations is considered to be both nitrogen and phosphorus limited based on periphyton assessments (30), and an increased carbon and energy source might not have an impact if microbial communities are otherwise limited. The carbon and energy effect and the inhibitory effect due to bituminous compounds simultaneously influence Athabasca watershed microorganisms. Inhibitory effects are probably more visible when growth is limited by available nutrients and the input of carbon compounds from bitumen cannot be fully exploited. It appears that in the first 8 weeks of biofilm formation, the inhibitory effects of bituminous compounds override any potential increases in carbon and energy, as all productivity indicators were negatively correlated to bituminous compound concentration. However, in the longer term, the relative increases observed in some bacterial taxa (e.g., Alphaproteobacteria and Betaproteobacteria) and in the expression of hydrocarbon-degrading genes with increasing bituminous compound concentrations might lead to an adapted community and to biofilms of thickness and productivity similar to those exposed to lower bituminous compound concentrations.
Inhibitory effects and increased carbon and energy effects of bituminous compounds were not the only mechanisms influencing biofilm microbial communities, as the correlation coefficients observed with hydrocarbon concentrations were not perfect and some of the biofilms exposed to sediments did not behave exactly as other biofilms exposed to similarly contaminated sediments. For instance, cyanobacteria, algae, and bacteria levels were lower in the upper Steepbank River, Suncor downstream, and Athabasca reference site, where TPH and NA concentrations were also low, while bacterial production was higher in the lower Firebag River, where hydrocarbon levels were higher. Other mechanisms that were not explored here are differences in N and P concentration of the sediments, highly tolerant microorganisms in the original sediment microbial population, sediment granulometry, and interactions with other biota. This is illustrated by the clustering of samples coming from the Firebag River in some of the ordinations shown in Fig. 5. The incomplete solubility and organic matter binding of the bituminous compounds, especially PAHs, could also explain some of the discrepancies observed, as the concentrations in the sediments do not perfectly reflect the concentrations available to the biofilms in the reactors. Even with all these confounding factors, several highly significant correlations, with correlation coefficients below −0.7 and P values well below 0.001, were found between biological indicators and the concentration of various bituminous compounds. Since most of these correlations were negative, this clearly indicates that biofilm-forming organisms in the Athabasca watershed are strongly inhibited by increasing concentrations of bituminous compounds.
In conclusion, the results presented here showed that after 8 weeks of incubation, increased exposure to bituminous compounds reduced biofilm productivity and biomass while stimulating particular microbial taxa and some aerobic hydrocarbon-degrading genes. Natural and anthropogenic increases in bituminous substance release in the Athabasca watershed will, in the short term, strongly reduce microbial productivity, but in the longer term it could lead to the stimulation of already occurring microbes that can utilize the carbon present in bituminous compounds for growth and partly restore productivity if other nutrients are not limiting.
Supplementary Material
ACKNOWLEDGMENT
This project was undertaken with the financial support of Environment Canada's Strategic Applications of Genomics in the Environment (STAGE) program.
Footnotes
Published ahead of print 20 September 2013
Supplemental material for this article may be found at http://dx.doi.org/10.1128/AEM.02216-13.
REFERENCES
- 1.Strausz OP, Lown EM, Morales-Izquierdo A, Kazmi N, Montgomery DS, Payzant JD, Murgich J. 2011. Chemical composition of Athabasca bitumen: the distillable aromatic fraction. Energy Fuels 25:4552–4579 [Google Scholar]
- 2.Strausz OP, Morales-Izquierdo A, Kazmi N, Montgomery DS, Payzant JD, Safarik I, Murgich J. 2010. Chemical composition of Athabasca bitumen: the saturate fraction. Energy Fuels 24:5053–5072 [Google Scholar]
- 3.Kelly EN, Schindler DW, Hodson PV, Short JW, Radmanovich R, Nielsen CC. 2010. Oil sands development contributes elements toxic at low concentrations to the Athabasca River and its tributaries. Proc. Natl. Acad. Sci. U. S. A. 107:16178–16183 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kurek J, Kirk JL, Muir DCG, Wang X, Evans MS, Smol JP. 2013. Legacy of a half century of Athabasca oil sands development recorded by lake ecosystems. Proc. Natl. Acad. Sci. U. S. A. 110:1761–1766 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Holden AA, Donahue RB, Ulrich AC. 2011. Geochemical interactions between process-affected water from oil sands tailings ponds and North Alberta surficial sediments. J. Contam. Hydrol. 119:55–68 [DOI] [PubMed] [Google Scholar]
- 6.Del Rio LF, Hadwin AKM, Pinto LJ, MacKinnon MD, Moore MM. 2006. Degradation of naphthenic acids by sediment micro-organisms. J. Appl. Microbiol. 101:1049–1061 [DOI] [PubMed] [Google Scholar]
- 7.Wyndham RC, Costerton JW. 1981. Heterotrophic potentials and hydrocarbon biodegradation potentials of sediment microorganisms within the Athabasca oil sands deposit. Appl. Environ. Microbiol. 41:783–790 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wyndham RC, Costerton JW. 1981. In vitro microbial degradation of bituminous hydrocarbons and in situ colonization of bitumen surfaces within the Athabasca oil sands deposit. Appl. Environ. Microbiol. 41:791–800 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yergeau E, Lawrence JR, Sanschagrin S, Waiser MJ, Korber DR, Greer CW. 2012. Next-generation sequencing of microbial communities in the Athabasca River and its tributaries in relation to oil sands mining activities. Appl. Environ. Microbiol. 78:7626–7637 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Chambers PA, Culp JM, Glozier NE, Cash KJ, Wrona FJ, Noton L. 2006. Northern rivers ecosystem initiative: nutrients and dissolved oxygen–issues and impacts. Environ. Monit. Assess. 113:117–141 [DOI] [PubMed] [Google Scholar]
- 11.Blakley ER. 1978. The microbial degradation of cyclohexanecarboxylic acid by a β-oxidation pathway with simultaneous induction to the utilization of benzoate. Can. J. Microbiol. 24:847–855 [DOI] [PubMed] [Google Scholar]
- 12.Blakley ER. 1974. The microbial degradation of cyclohexanecarboxylic acid: a pathway involving aromatization to form p-hydroxybenzoic acid. Can. J. Microbiol. 20:1297–1306 [Google Scholar]
- 13.Johnson RJ, Smith BE, Sutton PA, McGenity TJ, Rowland SJ, Whitby C. 2011. Microbial biodegradation of aromatic alkanoic naphthenic acids is affected by the degree of alkyl side chain branching. ISME J. 5:486–496 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Frank RA, Kavanagh R, Kent Burnison B, Arsenault G, Headley JV, Peru KM, Van Der Kraak G, Solomon KR. 2008. Toxicity assessment of collected fractions from an extracted naphthenic acid mixture. Chemosphere 72:1309–1314 [DOI] [PubMed] [Google Scholar]
- 15.Herman DC, Fedorak PM, MacKinnon MD, Costerton JW. 1994. Biodegradation of naphthenic acids by microbial populations indigenous to oil sands tailings. Can. J. Microbiol. 40:467–477 [DOI] [PubMed] [Google Scholar]
- 16.Headley JV, McMartin DW. 2004. A review of the occurrence and fate of naphthenic acids in aquatic environments. J. Environ. Sci. Health Part A 39:1989–2010 [DOI] [PubMed] [Google Scholar]
- 17.Leung SS, MacKinnon MD, Smith REH. 2003. The ecological effects of naphthenic acids and salts on phytoplankton from the Athabasca oil sands region. Aquat. Toxicol. 62:11–26 [DOI] [PubMed] [Google Scholar]
- 18.Quagraine EK, Peterson HG, Headley JV. 2005. In situ bioremediation of naphthenic acids contaminated tailing pond waters in the Athabasca oil sands region–demonstrated field studies and plausible options: a review. J. Environ. Sci. Health A Tox. Hazard Subst. Environ. Eng. 40:685–722 [DOI] [PubMed] [Google Scholar]
- 19.Sierra-Alvarez R, Lettinga G. 1991. The effect of aromatic structure on the inhibition of acetoclastic methanogenesis in granular sludge. Appl. Microbiol. Biotechnol. 34:544–550 [Google Scholar]
- 20.Sverdrup LE, Ekelund F, Krogh PH, Nielsen T, Johnsen K. 2002. Soil microbial toxicity of eight polycyclic aromatic compounds: effects on nitrification, the genetic diversity of bacteria, and the total number of protozoans. Environ. Toxicol. Chem. 21:1644–1650 [PubMed] [Google Scholar]
- 21.El-Alawi YS, McConkey BJ, Dixon DG, Greenberg BM. 2002. Measurement of short- and long-term toxicity of polycyclic aromatic hydrocarbons using luminescent bacteria. Ecotoxicol. Environ. Saf. 51:12–21 [DOI] [PubMed] [Google Scholar]
- 22.McConkey BJ, Duxbury CL, Dixon DG, Greenberg BM. 1997. Toxicity of a pah photooxidation product to the bacteria Photobacterium phosphoreum and the duckweed Lemna gibba: effects of phenanthrene and its primary photoproduct, phenanthrenequinone. Environ. Toxicol. Chem. 16:892–899 [Google Scholar]
- 23.Gilde K, Pinckney JL. 2012. Sublethal effects of crude oil on the community structure of estuarine phytoplankton. Estuaries Coasts 35:853–861 [Google Scholar]
- 24.Chronopoulou P-M, Fahy A, Coulon F, Païssé S, Goñi-Urriza M, Peperzak L, Acuña Alvarez L, McKew BA, Lawson T, Timmis KN, Duran R, Underwood GJC, McGenity TJ. 2013. Impact of a simulated oil spill on benthic phototrophs and nitrogen-fixing bacteria in mudflat mesocosms. Environ. Microbiol. 15:242–252 [DOI] [PubMed] [Google Scholar]
- 25.Larson CA, Passy SI. 2013. Rates of species accumulation and taxonomic diversification during phototrophic biofilm development are controlled by both nutrient supply and current velocity. Appl. Environ. Microbiol. 79:2054–2060 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yergeau E, Lawrence JR, Korber DR, Waiser MJ, Greer CW. 2010. Meta-transcriptomic analysis of the response of river biofilms to pharmaceutical products using anonymous DNA microarrays. Appl. Environ. Microbiol. 76:5432–5439 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Yergeau E, Sanschagrin S, Waiser MJ, Lawrence JR, Greer CW. 2012. Sub-inhibitory concentrations of different pharmaceutical products affect the meta-transcriptome of river biofilm communities cultivated in rotating annular reactors. Environ. Microbiol. Rep. 4:350–359 [DOI] [PubMed] [Google Scholar]
- 28.Battin TJ, Kaplan LA, Denis Newbold J, Hansen CME. 2003. Contributions of microbial biofilms to ecosystem processes in stream mesocosms. Nature 426:439–442 [DOI] [PubMed] [Google Scholar]
- 29.Besemer K, Peter H, Logue JB, Langenheder S, Lindstrom ES, Tranvik LJ, Battin TJ. 2012. Unraveling assembly of stream biofilm communities. ISME J. 6:1459–1468 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Chambers P, Dale A, Scrimgeour G, Bothwell M. 2000. Nutrient enrichment of northern rivers in response to pulp mill and municipal discharges. J. Aquat. Ecosyst. Stress Recov. 8:53–66 [Google Scholar]
- 31.Lawrence JR, Chenier MR, Roy R, Beaumier D, Fortin N, Swerhone GDW, Neu TR, Greer CW. 2004. Microscale and molecular assessment of impacts of nickel, nutrients, and oxygen level on structure and function of river biofilm communities. Appl. Environ. Microbiol. 70:4326–4339 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Lawrence JR, Swerhone GDW, Neu TR. 2000. A simple rotating annular reactor for replicated biofilm studies. J. Microbiol. Methods 42:215–224 [DOI] [PubMed] [Google Scholar]
- 33.Neu TR, Swerhone GDW, Bockelmann U, Lawrence JR. 2005. Effect of CNP on composition and structure of lotic biofilms as detected with lectin-specific glycoconjugates. Aquat. Microb. Ecol. 38:283–294 [Google Scholar]
- 34.Neu TR, Woelfl S, Lawrence JR. 2004. Three-dimensional differentiation of photo-autotrophic biofilm constituents by multi-channel laser scanning microscopy (single-photon and two-photon excitation). J. Microbiol. Methods 56:161–172 [DOI] [PubMed] [Google Scholar]
- 35.Neu TR, Swerhone GDW, Lawrence JR. 2001. Assessment of lectin-binding analysis for in situ detection of glycoconjugates in biofilm systems. Microbiology 147:299–313 [DOI] [PubMed] [Google Scholar]
- 36.Manz W, Wendt-Potthoff K, Neu TR, Szewzyk U, Lawrence JR. 1999. Phylogenetic composition, spatial structure, and dynamics of lotic bacterial biofilms investigated by fluorescent in situ hybridization and confocal laser scanning microscopy. Microb. Ecol. 37:225–237 [DOI] [PubMed] [Google Scholar]
- 37.Lawrence JR, Zhu B, Swerhone GDW, Roy J, Wassenaar LI, Topp E, Korber DR. 2009. Comparative microscale analysis of the effects of triclosan and triclocarban on the structure and function of river biofilm communities. Sci. Total Environ. 407:3307–3316 [DOI] [PubMed] [Google Scholar]
- 38.Robarts RD, Wicks RJ. 1989. [Methyl-3H] thymidine macromolecular incorporation and lipid labeling: their significance to DNA labelling during measurements of aquatic bacterial growth rate. Limnol. Oceanogr. 34:213–222 [Google Scholar]
- 39.Yergeau E, Bokhorst S, Huiskes AHL, Boschker HTS, Aerts R, Kowalchuk GA. 2007. Size and structure of bacterial, fungal and nematode communities along an Antarctic environmental gradient. FEMS Microbiol. Ecol. 59:436–451 [DOI] [PubMed] [Google Scholar]
- 40.Yergeau E, Kowalchuk GA. 2008. Responses of Antarctic soil microbial communities and associated functions to temperature and freeze-thaw cycle frequency. Environ. Microbiol. 10:2223–2235 [DOI] [PubMed] [Google Scholar]
- 41.Stewart FJ, Ottesen EA, DeLong EF. 2010. Development and quantitative analyses of a universal rRNA-subtraction protocol for microbial metatranscriptomics. ISME J. 4:896–907 [DOI] [PubMed] [Google Scholar]
- 42.Baker GC, Smith JJ, Cowan DA. 2003. Review and re-analysis of domain-specific 16S primers. J. Microbiol. Methods 55:541–555 [DOI] [PubMed] [Google Scholar]
- 43.Bell TH, Yergeau E, Maynard C, Juck D, Whyte LG, Greer CW. 2013. Predictable bacterial composition and hydrocarbon degradation in Arctic soils following diesel and nutrient disturbance. ISME J. 7:1200–1210 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Hamady M, Lozupone C, Knight R. 2010. Fast UniFrac: facilitating high-throughput phylogenetic analyses of microbial communities including analysis of pyrosequencing and PhyloChip data. ISME J. 4:17–27 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Quince C, Lanzen A, Davenport RJ, Turnbaugh PJ. 2011. Removing noise from pyrosequenced amplicons. BMC Bioinformatics 12:38. 10.1186/1471-2105-12-38 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Meyer M, Kircher M. 2010. Illumina sequencing library preparation for highly multiplexed target capture and sequencing. Cold Spring Harb. Protoc. 2010:pdb.prot5448. 10.1101/pdb.prot5448 [DOI] [PubMed] [Google Scholar]
- 47.Meyer F, Paarmann D, D'Souza M, Olson R, Glass EM, Kubal M, Paczian T, Rodriguez A, Stevens R, Wilke A, Wilkening J, Edwards RA. 2008. The metagenomics RAST server–a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinformatics 9:386. 10.1186/1471-2105-9-386 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Clemente JS, Fedorak PM. 2005. A review of the occurrence, analyses, toxicity, and biodegradation of naphthenic acids. Chemosphere 60:585–600 [DOI] [PubMed] [Google Scholar]
- 49.Han X, MacKinnon MD, Martin JW. 2009. Estimating the in situ biodegradation of naphthenic acids in oil sands process waters by HPLC/HRMS. Chemosphere 76:63–70 [DOI] [PubMed] [Google Scholar]
- 50.Crowe AU, Plant AL, Kermode AR. 2002. Effects of an industrial effluent on plant colonization and on the germination and post-germinative growth of seeds of terrestrial and aquatic plant species. Environ. Pollut. 117:179–189 [DOI] [PubMed] [Google Scholar]
- 51.Gentes M-L, Waldner C, Papp Z, Smits JEG. 2006. Effects of oil sands tailings compounds and harsh weather on mortality rates, growth and detoxification efforts in nestling tree swallows (Tachycineta bicolor). Environ. Pollut. 142:24–33 [DOI] [PubMed] [Google Scholar]
- 52.Lister A, Nero V, Farwell A, Dixon DG, Van Der Kraak G. 2008. Reproductive and stress hormone levels in goldfish (Carassius auratus) exposed to oil sands process-affected water. Aquat. Toxicol. 87:170–177 [DOI] [PubMed] [Google Scholar]
- 53.Pollet I, Bendell-Young LI. 2000. Amphibians as indicators of wetland quality in wetlands formed from oil sands effluent. Environ. Toxicol. Chem. 19:2589–2597 [Google Scholar]
- 54.Rogers VV, Wickstrom M, Liber K, MacKinnon MD. 2002. Acute and subchronic mammalian toxicity of naphthenic acids from oil sands tailings. Toxicol. Sci. 66:347–355 [DOI] [PubMed] [Google Scholar]
- 55.Scarlett AG, West CE, Jones D, Galloway TS, Rowland SJ. 2012. Predicted toxicity of naphthenic acids present in oil sands process-affected waters to a range of environmental and human endpoints. Sci. Total Environ. 425:119–127 [DOI] [PubMed] [Google Scholar]
- 56.Djomo JE, Dauta A, Ferrier V, Narbonne JF, Monkiedje A, Njine T, Garrigues P. 2004. Toxic effects of some major polyaromatic hydrocarbons found in crude oil and aquatic sediments on Scenedesmus subspicatus. Water Res. 38:1817–1821 [DOI] [PubMed] [Google Scholar]
- 57.Parry JD. 2004. Protozoan grazing of freshwater biofilms. Adv. Appl. Microbiol. 54:167–196 [DOI] [PubMed] [Google Scholar]
- 58.Kelly EN, Short JW, Schindler DW, Hodson PV, Ma M, Kwan AK, Fortin BL. 2009. Oil sands development contributes polycyclic aromatic compounds to the Athabasca River and its tributaries. Proc. Natl. Acad. Sci. U. S. A. 106:22346–22351 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.