OUP user menu

Evaluation of microbial biofilm communities from an Alberta oil sands tailings pond

Susanne Golby, Howard Ceri, Lisa M. Gieg, Indranil Chatterjee, Lyriam L.R. Marques, Raymond J. Turner
DOI: http://dx.doi.org/10.1111/j.1574-6941.2011.01212.x 240-250 First published online: 1 January 2012


Bitumen extraction from the oil sands of Alberta has resulted in millions of cubic meters of waste stored on-site in tailings ponds. Unique microbial ecology is expected in these ponds, which may be key to their bioremediation potential. We considered that direct culturing of microbes from a tailings sample as biofilms could lead to the recovery of microbial communities that provide good representation of the ecology of the tailings. Culturing of mixed species biofilms in vitro using the Calgary Biofilm Device (CBD) under aerobic, microaerobic, and anaerobic growth conditions was successful both with and without the addition of various growth nutrients. Denaturant gradient gel electrophoresis and 16S rRNA gene pyrotag sequencing revealed that unique mixed biofilm communities were recovered under each incubation condition, with the dominant species belonging to Pseudomonas, Thauera, Hydrogenophaga, Rhodoferax, and Acidovorax. This work used an approach that allowed organisms to grow as a biofilm directly from a sample collected of their environment, and the biofilms cultivated in vitro were representative of the endogenous environmental community. For the first time, representative environmental mixed species biofilms have been isolated and grown under laboratory conditions from an oil sands tailings pond environment and a description of their composition is provided.

  • Calgary Biofilm Device
  • environmental isolates
  • mixed species biofilm
  • oil sands tailings pond
  • pyrosequencing


Covering an area of over 100 000 km2, the oil sands region of northern Alberta yield 1.3 million barrels of bitumen per day, a volume that continues to grow (CAPP 2010). The waste byproducts of the oil sands mining are collected in large manufactured settling basins called tailing ponds (Nix & Martin, 1992). Tailings contain water, sand, clay, residual bitumen, heavy metals, naphtha diluent, and naphthenic acids (NAs). The latter contribute to the toxicity of the tailings ponds, which are thus managed under a zero discharge policy (Quagraine et al., 2005). In the tailings ponds, the solids are left to settle and consolidate, leaving a layer of water overtop loosely packed sediments. Tailings ponds contain active microbial communities; their population numbers as well their aerobic and anaerobic biodegradation activities have been the focus of many studies (Wyndham & Costerton, 1981; Foght et al., 1985; Nix & Martin, 1992; Herman et al., 1993, 1994; Holowenko et al., 2000; Fedorak et al., 2002; Del Rio et al., 2006; Siddique et al., 2006; Penner & Foght, 2010; Ramos-Padron et al., 2011).

Biofilms are consortia of microorganisms of one or more species that adhere to a surface and are enveloped in an extra polymeric substance. Biofilms have been applied in biotreatment efforts, particularly in wastewater treatment facilities, for decades (Nicolella et al., 2000). Biofilms exhibit an altered physiological state from their planktonic counterparts, which likely contribute to their robustness in natural environments (Hall-Stoodley et al., 2004). It is now widely believed that compared to planktonic cultures, biofilms have a decreased susceptibility to environmental stressors such as heavy metals or hydrocarbons (Hall-Stoodley et al., 2004; Harrison et al., 2007), which exist in many contaminated sites including oil sands tailings ponds (Nix & Martin, 1992; Allen, 2008a). Investigations of the microbial communities present in the natural environment once relied on culture-based techniques. However, because the majority of microorganisms are not easily grown in the laboratory, culture-based techniques are limited, which has lead to the dependence on molecular methods such as metagenomics instead. Biofilm studies have commonly been conducted by isolating individual planktonic species from an environmental sample either in a liquid or in an agar medium, then placing the isolates into a biofilm mode of growth in vitro. Using this technique, only a few select organisms are typically recovered and the microbes that prefer the biofilm mode of growth may be selected against in the initial isolation procedures. Previous studies that have attempted to isolate organisms using biofilm support models have been successful in recovering multiple different organisms (Puhakka et al., 1995; Griebler et al., 2002; Al-Awadhi et al., 2003; Ringelberg et al., 2011). Stach & Burns (2002) compared the difference between enrichment cultures and biofilm cultures, and proposed that biofilm culture methods are more appropriate for microbial community studies as it allows for better understanding of community level interactions.

The aim of this study was to cultivate microorganisms from oil sands tailings ponds by selecting them to grow as a mixed species biofilm directly from their environment in an in vitro model. The hypothesis was that this approach would provide for a different view of the microbes in this tailings pond environment in comparison with previously published studies. The Calgary Biofilm Device (CBD) has been used to establish mixed species biofilms from colonoscopy sections that closely reflect the diversity of microbial populations seen in individual patients (Sproule-Willoughby et al., 2010). We therefore investigated whether the CBD can be similarly used to obtain representative mixed species biofilms from tailings ponds environments. Biofilm formation was tested under different laboratory culture conditions and the resulting communities were compared using molecular tools. Our study demonstrates that a good representation of the tailings pond microorganisms can be cultured if grown directly as a biofilm. As expected, media augmentation and different growth conditions altered the community recovered.

Materials and methods

Tailings sample

A sediment sample (sludge) from a tailings pond in the Athabasca oil sands region of northern Alberta was collected in July 2009 from a depth of 0.45 m below the surface. The average temperature of the tailings pond is c. 18 °C, regardless of season. The sample was approximately 80–85% solid, and the sample jar was filled completely to limit oxygen exposure. The sample was transported to the University of Calgary and stored under anaerobic conditions without hydrogen (90% N2, 10% CO2) at room temperature. A single tailings sample was used for all inoculants to help maintain reproducibility as the tailings are very heterogeneous and different physical samplings can be quite unique to each other and can vary with depth, aeration, and company/mine type.

Biofilm culturing

The inoculum that seeded all biofilms was tailings sludge. The tailings sample was stirred before a subsample was removed and diluted twofold with sterile double distilled water (anoxic water was used with the anaerobic biofilms). Biofilms were grown on the CBD following published procedures (Ceri et al., 1999, 2001; Harrison et al., 2005, 2010). Briefly, wells of a 96-well microtiter plate were filled with 75 μL of the tailings/distilled water mixture and 75 μL of a specific growth medium or water. Plates were incubated on a gyrorotary shaker at 150 r.p.m. to provide shear force, which encourages cell attachment to the CBD pegs and development of biofilms. Wells were replenished with sterile fresh medium every 24 and 48 h for aerobic and anaerobic growth, respectively. Anaerobic biofilms were grown under atmospheric conditions of 90% N2 and 10% CO2/H2 mix. Microaerobic biofilms were grown in a candle jar (with an estimated O2 concentration between 17% and 18%) and replenished with sterile fresh growth medium every 72 h (Jensen & Trager, 1977; Bolton & Coates, 1983). The candle was relit every 24 h and left to extinguish. A minimum of 24 biofilms were grown from the same subsample of starting material at each time.

Because of the nature of the tailings sludge, the CBD pegs were coated with oil organics that adhered to the peg surface during inoculation. These oil organics were able to serve as carbon sources and/or electron donors in addition to the supplemented growth media. The media used were the following: R2B (Difco), trypticase soy broth (TSB) (Difco), and a general anaerobic medium for freshwater anaerobes (GA medium) adapted from McInerney et al. 1979. Per 100 mL: 5 mL Pfennig I (10 g L−1 K2HPO4), 5 mL Pfennig II (6.6 g L−1 MgCl2; 8.0 g L−1 NaCl; 8.0 g L−1 NH4Cl; 1.0 g L−1 CaCl2·2H2O), 1 mL Wolin metals (0.5 g L−1 EDTA; 3.0 g L−1 MgSO4·6H2O; 0.5 g L−1 MnSO4·H2O; 1.0 g L−1 NaCl; 0.1 g L−1 CaCl2·2H2O; 0.1 g L−1 ZnSO4·7H2O; 0.1 g L−1 FeSO4·7H2O; 0.01 g L−1 CuSO4·7H2O; 0.01g L−1 Na2MoO4·2H2O; 0.01 g L−1 H3BO3, 0.005 g L−1 Na2SeO4; 0.003 g L−1 NiCl2·6H2O), 1 mL Balch vitamins [2.0 mg L−1 biotin; 2.0 mg L−1 folic acid; 10.0 mg L−1 pyridoxine–HCl; 5.0 mg L−1 thiamine–HCl; 5.0 mg L−1 riboflavin; 5.0 mg L−1 nicotinic acid; 5.0 mg L−1 dl-calcium pantothenate; 0.1 mg L−1 vitamin B12; 5.0 mg L−1 p-aminobenzoic acid (PABA); 5.0 mgL−1 lipoic acid mercaptoethane; 5.0 mg L−1 methyl ester sulfonic acid (MESA)], 0.1 mL 0.1% resazurin, 0.35 g NaHCO3. Under anaerobic incubations, cysteine sulfide (0.1 mL of a 2.5% solution) and 2.5 mM nitrate were added to promote the growth of nitrate reducers. To encourage the growth of sulfate-reducing bacteria, the growth medium contained 16 mM of sulfate and 2 mL of 2.5% cysteine sulfide was added per 100 mL.

Culture plates and viable cell counts

A subsample of the tailings sludge was diluted twofold in sterile double distilled water, and 100 μL was spread plated on either R2A or TSA. Plates were evaluated for growth after 4–7 days of incubation at 25 °C. Pegs from the CBD were removed and rinsed twice in sterile 0.9% saline to remove planktonic and loosely attached cells from the cultured biofilms. Pegs were then placed in 200 μL 0.9% saline with 1% Tween 20 and sonicated for 15 min to disrupt and remove the biofilms off the pegs (model 250T sonicator; VWR International). The detached biofilm suspensions were serially diluted, and 20 μL of each dilution was spot plated on the same type of medium used as the liquid medium in the CBD plate. For the biofilms grown with no growth medium, R2A plates were used. Growth was evaluated based on viable cell counts, recognizing that this is a selection of only those organisms able to grow under this culture condition, yet it is a typical way to evaluate the growth of biofilms (Ceri et al., 1999, 2001; Harrison et al., 2005, 2010).

DNA extraction

Total genomic DNA was extracted from 250 mg of the starting sludge inoculum using environmental sample extraction kits. DNA from biofilms was extracted using two biofilm-covered CBD pegs per extraction to yield higher DNA concentrations with an isolation kit optimized for the type of sample. For aerobic and microaerobic biofilms, DNA was extracted using the MO Bio Power Soil™ DNA isolation kit (Carlsbad CA), whereas the MP Bio Soil DNA Extraction™ kit (Solon, OH) that includes a bead beating step was used for the anaerobic biofilms and the sludge inoculum. DNA extractions were carried out according to the manufacturers’ instructions.

Denaturing gradient gel electrophoresis (DGGE) analysis

A region of the 16S rRNA gene was targeted with bacterial primers 341f-GC (5′-CGCCCGCCGCGCCCCGCGCCCG TCCCGCCGCCCCCGCCCGCCTACGGGAGGCAGCAG-3′) and 907r (5′-CCGTCAATTCMTTTGAGTT-3′) (Muyzer et al., 1995). The PCR was run with 30 s at 94 °C, 30 s at 47 °C (0.5 °C increase after each cycle), 1 min at 72 °C × 9 and 30 s at 94 °C, 30 s at 52 °C and 1 min at 72 °C × 19. For analysis, 6% polyacrylamide gels were used with a denaturant gradient of 30–60% composed of deionized formamide and urea. Gels were run 1× TAE buffer (40 mM Tris, 20 mM acetic acid, 1 mM EDTA pH 8) at 60 °C and 60 V for 17 h. Bands were visualized by staining with 1× SYBR® Gold and illuminated on a UV-transilluminator under a SYBR® gold filter.

Quantitative PCR (Q-PCR) analysis

DNA was amplified with the same primers used in the DGGE analysis without the GC clamp with cycle conditions 30 s at 94 °C, 30 s at 52 °C, and 1 min at 72 °C × 30. The SYBR® Green detection method (Bio-Rad) was used to quantify 16S rRNA gene copy numbers. Gene copy numbers are reported as gene copy number per biofilm-covered peg as calculated by formula . Gene copy numbers per gram of tailings sample was calculated by formula (Ritalahti et al., 2006). The correlation coefficient for all standard curves was ≥ 0.99, and PCR efficiency was between 83% and 84%. Gene copy numbers were extrapolated from a standard curve prepared from a known concentration of genomic Legionella pneuomophila DNA. Embedded Image Embedded Image

454 pyrosequencing

Extracted DNA was amplified with 16S rRNA gene primers targeted for a range of both bacteria and archaea [926Fw (AAACTYAAAKGAATTGRCGG) and 1392R (ACGGGCGGTGTGTRC)], designed by the Joint Genome Institute for use in the Roche 454 system (Ramos-Padron et al., 2011). The first PCR was run for 25 cycles with 30 s at 95 °C, 45 s at 55 °C, and 1.5 min at 72 °C, and products were verified on an agarose gel. This PCR product was used as the template for a second PCR with FLX titanium amplicon primers 454T_RA_X and 454T-FBw that have the sequences for 16S primers 926Fw and 1392R at their 3′-ends. Primer 454T_RA_X contains a 25 nucleotide A-adaptor (CGTATCGCCTCCCTCGCGCCATCAG) and a 10 nucleotide multiplex identifier barcode sequence X, and primer 454T-FB has a 25 nucleotide B-adaptor sequence (CTATGCGCCTTGCCAGCCCGCTCAG). The second PCR was run for 10 cycles with 45 s at 96 °C, 30 s at 65 °C, and 1 min at 72 °C. This PCR product was checked on an agarose gel and purified with QIAquick PCR Purification kit (Qiagen). PCR products were quantified by fluorescence spectrophotometer using a Quanti-iT™ PicoGreen dsDNA kit (Invitrogen). Samples (5 ng uL−1) were processed at Genome Quebec and McGill University Innovation Centre, in Montreal, Quebec, Canada, for pyrosequencing with a Genome Sequencer FLX Instrument, using a GS FLX Titanium Series kit XLR70 (Roche Diagnostics Corporation).

All the 16S rRNA data were processed using the Phoenix pipeline that was developed in the Sun Center of Excellence for Visual Genomics (COE) in Calgary (Ramos-Padron et al., 2011) (Available from authors upon request). In the Phoenix pipeline, pyrosequence data were subjected to stringent systematic checks to remove low-quality reads and minimize sequencing errors (Huse et al., 2007). Eliminated sequences included those that (i) did not perfectly match the adaptor and primer sequences, (ii) had ambiguous bases, (iii) had a sequence which is < 200 bp or > 520 bp, (iv) had an average quality score below 25, and (v) contained homopolymer lengths > 8 bp. The remaining high-quality sequences were compared against the nonredundant SSU Ref data set of silva102 (Pruesse et al., 2007) using the Tera-blast algorithm on a TimeLogic Decypher system (Active Motif, Inc.) consisting of 12 boards. Tera-blast results were used to screen for problematic, chimeric, and eukaryotic sequences. Sequences having a best alignment covering (alignment length divided by the trimmed query sequence length) < 80% or best blast search hit e-value greater than e−50 were excluded as problematic sequences. Putative chimeras were identified using a two-stage approach. The sequences having a best alignment covering < 90% of the trimmed read length, with > 90% sequence identity to the best blast match, were identified as potential chimeras. The potential chimeras were excluded from further analysis if they were also identified as chimeras at minimum 80% bootstrap support in chimera-slayer implemented in the Mothur software package (Schloss et al., 2009). The filtered sequences, passing the quality control, problematic, chimerical, eukaryotic sequence removal, were clustered into operational taxonomic units (OTUs) at 3% distance using the complete linkage algorithm in Mothur. A taxonomic consensus of each representative sequences from each OTU was derived from the recurring species within 5% of the best bitscore from a blast search against silva database. Of the good reads generated by pyrosequencing, XYZ (different in different sample batch) were assigned taxonomic identifiers, which were identified at the phylum, class, and/or genus level. Chao1 estimates and Shannon indexes were calculated using the Mothur software.

Confocal scanning laser microscopy (CSLM)

Biofilm-covered pegs were individually broken off the CBD, rinsed in 200 μL sterile 0.9% saline and fixed in 200 μL 2.5% glutaraldehyde solution for 30 min (Harrison et al., 2006). Biofilms were stained in the dark with 10 μM Syto® Red 62 (Molecular Probes) nucleic acid stain for 15 min. Images were taken with a 63× water immersion objective and Leica Confocal Software.


Using the CBD, mixed species biofilms were successfully cultivated under aerobic, microaerobic, and anaerobic growth conditions. Evaluation of biofilm growth and comparison of cultured biofilms incubated under different conditions indicated selection of unique biofilm communities from within the original microbial community in the tailings sludge inoculum. To visualize biofilm growth on the pegs of the CBD, pegs were stained with a cell-permeable nucleic acid stain and visualized by CSLM. Figure shows an example of the progression and expansion of an aerobic biofilm after a period of 3, 4, and 7 days. Biofilm development was seen mostly at the air–liquid interface of the peg, in addition to the areas around the peg that had droplets of the oily sludge adhered to it (arrows, Fig. a and b). CSLM images mostly showed flat biofilms that were 1–3 cell layers thick (Fig. ).

CSLM images (63×) of aerobically grown biofilms on the CBD pegs with R2B medium. (a) Three-day-old biofilms stained with acridine orange nucleic acid stain (Ex 506 nm). Autoflouresence of hydrocarbons from the tailings can be seen in the background and small clusters of cells are indicated with a white arrow. (b) Four-day-old biofilm stained with SytoRed nucleic acid stain (Ex 652). No autoflouresence from the hydrocarbons is seen at this wavelength. Void spots within the biofilm are oil droplets adhered to the peg that the bacteria congregate around (indicated with yellow arrows). (c) One-week-old biofilm stained with SytoRed. Scale bars, 20 μm.

To test whether the tailings material that served as the microbial inoculum contained sufficient carbon and nutrient sources for biofilm development, biofilm growth was tested both with and without the addition of any growth medium. 16S rRNA gene copy numbers were used to estimate bacterial numbers in each biofilm using Q-PCR (Fig. ). After incubation, the biofilms grown with no growth medium [Aerobic− (A−), Microaerobic− (M−) Anaerobic− (An−)] had lower gene copy numbers per peg (in the 105–106 range) than when a growth medium was provided (Fig. ). Aerobic and anaerobic biofilms amended with growth medium reached gene copy numbers of over 107 whereas that of the microaerobic biofilms amended with growth medium reached between 105 and 106 gene copy numbers per peg (Fig. ). The tailings sludge inoculum had a 16S rRNA gene copy number of 5.4 × 108 g−1.

16S rRNA gene copy numbers in the tailings biofilms as determined by Q-PCR. Data are reported as gene copy number/biofilm peg A + R2B, aerobic with R2B; A + TSB, aerobic with TSB; A−, aerobic no medium (tailings only); M + R2B, microaerobic with R2B; M + TSB, microaerobic with TSB; Mn, microaerobic no medium (tailings only); An+, anaerobic with GA medium; An−, anaerobic no medium (tailings only); AnS, anaerobic sulfate-reducing bacteria supporting medium; AnN, anaerobic nitrate reducers supporting medium. For comparison, the tailings sludge inoculum had a 16S rRNA gene copy number of 5.4 × 108 per gram of sludge. n = 3, error bars represent standard error of the mean.

DGGE was used as an initial measure of the diversity within each biofilm and to assess how this diversity changed with different incubation conditions. Different banding patterns and intensities were seen in the DGGE gels under all growth conditions (Fig. a–c). Addition of the growth media generated different banding patterns from the biofilms with no growth medium. Replicate analyses of all growth conditions tested suggest equivalent biofilms formed on each peg.

DGGE of biofilms and sludge inoculum. (a) Aerobic biofilms grown for 96 h at 25 °C. A + R2B, aerobic with R2B; A + TSB, aerobic with TSB; A−−, aerobic no medium (tailings only). (b) M + R2B, microaerobic with R2B; M + TSB, microaerobic with TSB; M−−, microaerobic no medium (tailings only). (c) An+, anaerobic with GA medium; An−−, anaerobic no medium (tailings only); AnS, anaerobic SRB supporting medium; AnN, anaerobic NRB supporting medium. (d) Sludge inoculum.

Biofilms were sonicated off the pegs, serially diluted and spot plated on agar plates for CFU per peg enumerations and to observe how many different colony morphologies could be recovered. Aerobic and microaerobic agar plates showed three to five different colony morphologies could be recovered and grown per biofilm, and CFU per peg were between 102 and 105 (data not shown). Plate culturing of organisms from the anaerobic biofilms were unsuccessful, and growth was very inconsistent between replicates (data not shown). Spread plate analysis of the sludge inoculum revealed four or five different colony morphologies with a CFU count of between 105 and 106 (data not shown). Some of these colonies were similar to the morphologies recovered from the biofilms.

Pyrosequencing was conducted on the tailings sludge inoculum, aerobic biofilms with (A+) and without (A−) the addition of R2B medium, and anaerobic biofilms with (An+) and without (An−) the addition of GA medium (Table , Figs and ). The sequencing results further confirmed that the biofilms growing on the CBD were composed of multiple genera (Fig. ).

(a) Predominant phyla in each biofilm and sludge inoculum as determined by 454 pyrosequencing. (b) Proteobacterial classes. Sludge, inoculum; A+, aerobic with R2B; A−, aerobic no growth medium; An+, anaerobic with GA medium; An−, anaerobic no growth medium.

Predominant genera identified in the sludge inoculum and biofilms as determined by 454 pyrosequencing. Only those with a representation of 1% or higher are shown. (a) Sludge inoculum; (b) aerobic without growth medium; (c) aerobic with R2B medium addition (tailings only); (d) anaerobic without growth medium (tailings only); (e) anaerobic with a GA growth medium.

View this table:

Total number of reads, OTUs, and biodiversity indexes for the sludge inoculum and biofilms at 3% dissimilarity

SampleTotal no. of readsTotal OTUsTotal OTU estimate (Chao1)Shannon index (H′)
Sludge (inoculum)652482314104.9
  • Aerobic−, no addition of growth medium; Aerobic+, addition of R2B; Anaerobic−, no addition of growth medium; Anaerobic+, addition of GA medium.

  • Based on an average of two samples.

At the phylum level, good similarities were observed between the sludge inoculum and the biofilms (Fig. ). At the genus level, more differences were observed (Fig. ). The majority of the sequences within the sludge inoculum aligned within the Proteobacteria, representing 59% of the total community. In all biofilms, the Proteobacteria were also the most dominant, with community percentages between 65 and 73 (Fig. a). In the biofilms, the Deltaproteobacteria diminished to < 1%, from the starting 7% in the sludge, whereas the Alphaproteobacteria population doubled in all biofilms from the original 2.25% in the sludge (Fig. b). The Betaproteobacterial and Gammaproteobacterial populations stayed comparable between the biofilms and sludge inoculum with population percentages around 45 and 10, respectively. The Euryarchaeota population almost disappeared in all biofilms compared to the sludge inoculum, and the addition of a growth medium under aerobic and anaerobic conditions doubled the Firmicute populations (Fig. a). Little population change was seen among the other phyla.

There were a total of 18 genera with a community percentage of 1 or greater in the sludge inoculum (Fig. a). Aerobic biofilms had between 12 and 18 genera with a community percentage of 1 or greater, and the anaerobic biofilms had between 12 and 15 genera (Fig. b–e).

The most abundant genus in the sludge was Brachymonas at 17.2%, followed by Acidovorax (6.2%), Variovorax (5.7%), Rhodoferax (3.7%) and Thioalkalispira (3.7%) (Fig. a, Supporting Information, Table S1). Under aerobic conditions, the addition of the growth medium strongly selected for one dominating genus, Pseudomonas (35%), followed by Thauera (13.1%), Hydrogenophaga (10.4%), Tolumonas (7.8%), and Alishewanella (4.8%) (Fig. c, Table S1). The aerobic biofilms without the addition of a growth medium had a more even population distribution, rather than one genera dominating (Fig. b, Table S1). The most abundant genus was Rhodoferax (7.6%), followed closely by Acidovorax (7.2%), Acinetobacter (5.7%), Pseudomonas (4.8%), and Thioalkalispira (4.3%) (Fig. b). Under anaerobic conditions, both biofilms showed a similar spread of organisms, with one genera representing close to 20% of the total community and the others beginning at 10% representation and decreasing at close increments (Fig. d–e, Table S1). In anaerobic biofilms with no growth medium, Hydrogenophaga was the most abundant with 19.5% followed by Rhodoferax (9.9%), Methyloversatilis (9.9%), Magnetospirillum (6.5%), and Acidovorax (4.0%) (Fig. d, Table S1). Addition of the GA medium resulted in Methyloversatilis being the most prevalent genus at 17.6%, followed by Pseudomonas (8.4%), Thauera (8.0%), Azoarcus (6.0%), and Acholeplama (5.1%) (Fig. e, Table S1). A complete list of all genera identified can be found in supplementary information.

The microbial community within the tailings inoculum had a Shannon Diversity Index value of 4.9 (Table ). All the biofilms grown had a Shannon index between 4.1 and 4.3, which suggest the biofilms retained similar evenness and diversity to the starting community (Table ). Total OTUs were calculated with the nonparametric richness estimator Chao1 and can be compared to the number of OTUs generated from the pyrosequencing results (Table ).


In the present study, we sought to establish mixed species biofilms from an Alberta oil sands tailings pond that reflected the in situ microbial community using an in vitro model. This approach in its own right would provide selection for those organisms that prefer to grow as a surface attached consortium. Culturing biofilms directly from a tailings sludge sample on the CBD produced replica biofilms and thus is a tool to evaluate microbial communities from such an environment. Gene copy numbers of six independent biofilms (combined into three replicates) were calculated with Q-PCR, and standard error of the mean values was between 0.02 and 0.23, suggesting that replicate biofilms contained similar numbers of cells. Four to six biofilms (combined into two or three replicates) from each growth condition were analyzed by DGGE and compared with gel compare software to observe how similar the replicates were. The percent similarities between replicate lanes were between 73% and 95%. Although some differences are expected between replicates when working with a heterogeneous sample, the low standard errors of the Q-PCR between replicate biofilms indicate equivalent growth. The DGGE similarities indicate comparable community compositions among replicates (Hume et al., 2008).

This approach potentially selected for microorganisms that preferentially grow as biofilms, and also those that may naturally interact synergistically to form a community. As a result, our study recovered over 10 different genera of microbes from an oil sands tailings sample per biofilm and changes to the culture conditions such as growth media and oxygen tension selected for or against certain species resulting in a slightly different community. Biodiversity assessment revealed good evenness and diversity in our mixed species biofilms as seen from the Shannon indexes. The Chao1 richness estimator was selected to estimate total OTUs from each sample at 3% dissimilarity (Lemos et al., 2011). The cultured biofilms all contained fewer OTUs than the sludge inoculum, as expected from culturing techniques. The anaerobic biofilms with no growth medium were estimated by Chao1 to contain 844 of the 1410 OTUs estimated in the sludge inoculum (Table ). This growth condition had the highest recovered and highest Chao1 estimated OTUs. Being that this growth condition was the most similar to the indigenous conditions, it was expected that it would support most of the original organisms and be the most similar to the sludge inoculum. This, however, was not observed, as the communities do not appear to be significantly similar. Two possible explanations are that our model and growth conditions strongly selected against certain species that were abundant in the sludge, or a number of the organisms within the sludge inoculum were not viable, and thus could not cultivate the biofilm.

Foght et al. (1985) predicted that the aerobic community in the tailings ponds was dominated by either Alcaligenes spp. or Acinetobacter spp. depending on the depth. No Alcaligenes spp. were found in this study. Acinetobacter was present at a very low percentage in the sludge inoculum, and biofilms grown aerobically with no growth media had an increase in this genus to 5.7% (Fig. , Table S1). Under aerobic conditions with R2B medium, glucose and other carbohydrates were present as a carbon source in addition to the oil-associated organics. This resulted in selection for some fermentative microorganisms such as Tolumonas (7.7%) and Trichococcus (4.4%) that were at < 1% in the sludge inoculum (Fig. , Table S1). Hydrogenophaga was abundant in three of the biofilms, but was found at < 1% in the sludge. Hydrogenophaga has commonly been identified in wastewater treatment communities (Kaempfer et al., 1991; Magic-Knezev et al., 2009) and has been isolated from microbial consortia from benzene-contaminated sites in conjunction with Rhodoferax, Acidovorax, and Pseudomonas, which were also found in our study (Fahy et al., 2006; Aburto et al., 2009). The metabolically diverse iron-reducing Rhodoferax and denitrifying Acidovorax were found consistently in all biofilms and are commonly associated with anoxic hydrocarbon contaminated environments (Singleton et al., 2009; Penner & Foght, 2010). Acidovorax isolates have also been found in conjunction with Pseudomonas and Hydrogenophaga (You et al., 2002). Pseudomonas was detected in all biofilms in a higher abundance than in the sludge inoculum. Pseudomonas is a well-known biofilm former, as well as hydrocarbon degrader (Stringfellow & Aitken, 1994; Christofoletti Mazzeo et al., 2010), and thus it is not surprising to see such an abundance of this genus in our surface attached biofilms. Thauera, also present in all biofilms, can biodegrade toluene (Anders et al., 1995; Mao et al., 2010), and a single isolate of this genus has been shown to do so under both aerobic and anaerobic conditions (Shinoda et al., 2004). Bacteria in natural environments are prone to the sessile mode of growth, and the consistent finding of Hydrogenophaga, Acidovorax, Thauera, Rhodoferax, and Pseudomonas as dominant organisms in all biofilms suggests that these organisms have a preference to grow as a biofilm in an in vitro model (Beveridge et al., 1997).

Thioalkalispira was only found in the aerobic biofilms with no growth medium at 4% representation, which is similar to its representation in the sludge (3.7%) (Fig. , Table S1). Currently, to the authors’ knowledge, a single species belongs to the genus Thioalkalispira and is described as an obligate, microaerophilic sulfur oxidizing bacterium that was isolated from alkaline (soda) lakes (Sorokin & Kuenen, 2005). The oils sands tailings ponds are alkaline, typically with a pH of > 7.5 (Allen, 2008a, b), and may represent another environment for this genus.

Of the organisms found to exist in this tailings sample, the most prevalent one, Brachymonas (17.2%), was only detected in two of the four biofilms at less than half of this percentage. The archaea, which in this tailings sample were near 7% of the community, were not represented in any of the biofilms. Both Brachymonas and species of Archaea have previously been found in other tailings ponds at greater depths (> 10 m and > 4 m respectively) than the depth of sample this study used (0.45 m) (Penner & Foght, 2010; Ramos-Padron et al., 2011). This culture technique appeared to be selective against these organisms under the growth conditions tested here.

The heterogeneity of the tailings environment unavoidably creates variance among sampling efforts. PCR-based approaches, such as used in this study, also create biases, and thus this study is not meant to present the absolute ecology of the tailings environment, but rather present another approach through which this unique environment can be evaluated. Some earlier studies of the oil sands tailings used a combination of enrichment and colony isolations to recover NA degraders and found only a few species that belonged to Pseudomonas, Acinetobacter, and Alcaligenes (Herman et al., 1993, 1994; Del Rio et al., 2006). Two of these three were also recovered in this study. By examining the few different colony morphologies from our culture plates, it appeared that culturing many of the organisms from the sludge and cultured biofilms would take significant effort and wide range of growth conditions and agar types than were used here. Using only a few biofilm growth conditions, multiple organisms and diverse communities were recovered. As the majority of the bacteria in this environment are presumed to be oil associated (attached to oil and sand particles) (Bordenave et al., 2010), it gave us reason to try a growth support model to recover organisms from this environment. Molecular methods are now commonly used to investigate the oil sands, and a number of the organisms identified in this study corroborate well with those previously reported from direct 16S DNA sequencing of other areas of the Alberta oil sand tailings ponds, including Pseudomonas, Thauera, Rhodoferax, Acidovorax, Thiobacillus, and Brachymonas (Penner & Foght, 2010; Ramos-Padron et al., 2011). Unique to this study was the large selection of Hydrogenophaga in the biofilms.

A review by Singh et al. (2006) describes many applications of biofilm technology in the bioremediation of contaminated environments as well as many characteristics of biofilms that make them practical in this field. This study has demonstrated that endogenous species from an oil sands tailings pond will form mixed species biofilms on the CBD and that the cultured community can reflect the endogenous one. Isolating organisms by promoting their growth as a biofilm can select for a wide range of organisms, particularly those that may prefer to form biofilms as well as prefer to be in close assemblies or communities with each other. This could provide added value when assessing for microbial consortia that could be used for bioremediation and other purposes.

Supporting Information

Additional Supporting Information may be found in the online version of the article:

Table S1. Summary of organisms identified in the sludge inoculum and each biofilm as determined by 454 sequencing.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.


This work was supported by discovery grants from the Natural Science and Engineering Research Council to H.C. and R.J.T. Additionally, funding was provided by Genome Canada, Genome Alberta, the Government of Alberta and Genome BC to L.M.G. H.C. also acknowledges support from Alberta Advanced Education and Technology. S.G. was supported through an industrial collaboration with HydroQual Laboratories Ltd and Golder Associates, Calgary AB. We are thankful to Dr Gerrit Voordouw from the University of Calgary for his helpful support and discussions, as well as to Pernilla Stenroos from HydroQual Laboratories. We thank Dr Chris Sensen and Xiaoli Dong for the bioinformatics analyses. An oil sands operator from the Athabasca region is gratefully acknowledged for supplying the tailing pond sample.


  • Editor: Max Häggblom


View Abstract