Editor: Max Häggblom
Microbial consortia in mesocosm bioremediation trial using oil sorbents, slow-release fertilizer and bioaugmentation
Article first published online: 27 APR 2009
© 2009 Federation of European Microbiological Societies. Published by Blackwell Publishing Ltd. All rights reserved
FEMS Microbiology Ecology
Volume 69, Issue 2, pages 288–300, August 2009
How to Cite
Gertler, C., Gerdts, G., Timmis, K. N. and Golyshin, P. N. (2009), Microbial consortia in mesocosm bioremediation trial using oil sorbents, slow-release fertilizer and bioaugmentation. FEMS Microbiology Ecology, 69: 288–300. doi: 10.1111/j.1574-6941.2009.00693.x
- Issue published online: 6 JUL 2009
- Article first published online: 27 APR 2009
- Received 5 November 2008; revised 2 April 2009; accepted 11 April 2009.Final version published online 3 June 2009.
- marine oil degradation;
- oil sorbent materials;
- Top of page
- Materials and methods
- Supporting Information
An experimental prototype oil boom including oil sorbents, slow-release fertilizers and biomass of the marine oil-degrading bacterium, Alcanivorax borkumensis, was applied for sorption and degradation of heavy fuel oil in a 500-L mesocosm experiment. Fingerprinting of DNA and small subunit rRNA samples for microbial activity conducted to study the changes in microbial communities of both the water body and on the oil sorbent surface showed the prevalence of A. borkumensis on the surface of the oil sorbent. Growth of this obligate oil-degrading bacterium on immobilized oil coincided with a 30-fold increase in total respiration. A number of DNA and RNA signatures of aromatic hydrocarbon-degrading bacteria were detected both in samples of water body and on oil sorbent. Ultimately, the heavy fuel oil in this mesocosm study was effectively removed from the water body. This is the first study to successfully investigate the fate of oil-degrading microbial consortia in an experimental prototype for a bioremediation strategy in offshore, coastal or ship-bound oil spill mitigation using a combination of mechanical and biotechnological techniques.
- Top of page
- Materials and methods
- Supporting Information
Marine oil degradation is a process involving a multitude of specialized microorganisms with remarkable abilities to degrade complex mixtures of hydrocarbons at high rates. Among many hydrocarbon-utilizing microorganisms in marine ecosystems, Alcanivorax, Cyclocasticus, Oleispira, Oleiphilus, and Thalassolituus possess the unique skills to live solely on hydrocarbons in marine ecosystems while dominating the marine microbial communities during the process of biodegradation (Yakimov et al., 2005, 2007). Although many of these organisms are readily culturable, most of them have been detected in studies using culture-independent methods such as DNA fingerprinting or DNA clone libraries (Yakimov et al., 1998, 2005, 2007). Indeed, the biogeography of these organisms reveals a specialization of bacteria such as Oleispira antarctica to polar regions, whereas Alcanivorax borkumensis predominantly occurs in temperate zones (Yakimov et al., 2005, 2007). Other obligate hydrocarbon-degrading bacteria (OHCB) avoid concurrence of their kin by specialization to certain classes of hydrocarbons, such as Cycloclasticus sp., a very efficient decomposer of aromatic hydrocarbons (Kasai et al., 2002b). The best known and widest spread member of this group of Gammaproteobacteria is A. borkumensis (Golyshin et al., 2003). Its remarkable abilities for biosurfactant and biofilm production, a streamlined genome and metabolic pathways optimized for hydrocarbon degradation combined with highly efficient transport systems for nutrients and trace elements suggest this bacterium to be the most effective primary colonizer of oil–water interfaces (Schneiker et al., 2006; Yakimov et al., 2005, 2007). Because of its high competitiveness and its ubiquity, A. borkumensis is considered to be an excellent organism for bioremediation and oil spill mitigation in marine ecosystems (Head et al., 2006; McKew et al., 2007; Gertler et al., 2009).
The first experimental trials on oil bioremediation in marine systems were conducted in the 1970s (Atlas & Bartha, 1972; Cerniglia & Perry, 1973), following the oil spill of the oil tanker Torrey Canyon in the English Channel in 1967. Both biostimulation (addition of nutrients) and bioaugmentation (addition of microbial inocula) have been evaluated in numerous lab studies and refined in field trials performed after the Exxon Valdez oil spill in 1991 (Lindstrom et al., 1991). Indeed, beached oil from the Torrey Canyon (1967), Amoco Cadiz (1978), Exxon Valdez (1991), Nakhodka (1998), and Prestige (2002) caused severe damages to shorelines and required huge costs for shoreline clean-ups. The majority of experimental trials, therefore, focused on different types of beaches and shores (Swannell et al., 1996; MacNaughton et al., 1999; Kasai et al., 2002b; Röling, 2004; Cappello et al., 2007; Jiménez et al., 2007) and only a few experiments were dedicated to the offshore oil mitigation (for a review, see Swannell et al., 1996; Brakstad & Lodeng, 2005). Culture-independent methods were applied to detect microorganisms involved in hydrocarbon degradation and showed major differences between oil-degrading consortia in the seawater and those in beaches. Experiments conducted on beaches showed the abundance of terrestrial bacteria such as Rhodococcus spp. (MacNaughton et al., 1999) and Pseudomonas spp. (Röling, 2004) as well as marine oil-degrading bacteria such as Alcanivorax spp. (Röling, 2004). In contrast, open water studies in both microcosm and mesocosm scales displayed the prevalence of OHCB belonging to the Gammaproteobacteria, Alphaproteobacteria and members of the Cytophaga–Flavobacterium group (CFB) (Kasai et al., 2002b; Brakstad & Lodeng, 2005; Capello et al., 2005; Jiménez et al., 2007; Gertler et al., 2009). While the primary degraders of hydrocarbons can be easily identified (Jiménez et al., 2007), the function of many bacteria within the consortia remains unclear. Therefore, it is common to perform oil bioremediation with consortia rather than single organisms (for a review, see Swannell et al., 1996).
Yet, the practical limitation of bioaugmentation for offshore oil spills is the rapid dispersal of bacteria (Tagger et al., 1983). A potential solution to this problem can be found in the application of oil sorbent material. Numerous products of different origins and composition are currently available for purchase: i.e. polyethylene or polypropylene cloth (Wei et al., 2003), or plant fibres (Pasila, 2004; Suni et al., 2004, 2007). Both types of sorbents immobilize and absorb hydrocarbons, increasing the bioavailability of oil. A mesocosm study comparing a sorbent based on cotton grass with a synthetic sorbent proved the superiority of the plant fibres for absorbing diesel oil in brackish seawater (Suni et al., 2007). Although bioaugmentation with a microbial consortium from oil-contaminated soil significantly increased the hydrocarbon degradation in the same study, the fate of these microorganisms was not studied in depth.
Little is known about the fate of bacteria in the presence of sorbents or immobilized oil in the seawater. In the present work, we performed a mesocosm experiment to survey the fate of oil-degrading microorganisms in the presence, and on the surface, of oil sorbents in combination with biostimulation. This study provides a new insight into the dynamics of populations in microbial food webs in experimental oil booms and into perspectives of application of oil sorbents in oil bioremediation.
Materials and methods
- Top of page
- Materials and methods
- Supporting Information
Two experimental polyvinylchloride basins of 0.8 m × 0.9 m × 1.1 m each were set up at Biologische Anstalt Helgoland. The first basin was filled with 500 L of freshly collected seawater from Helgoland Roads on 17 July 2006. An experimental circular oil boom with a length of 2 m and a diameter of 0.2 m was prepared using a polypropylene tissue (Hellmann-Tech, Lehrte, Germany) as a hull material, which was filled with 10 kg of X-Oil® oil sorbent material (Hellmann-Tech), 4 kg of slow-release fertilizer Osmocote Pro® 18+10+11+2MgO+TE [9% (w/w) NO3−, 8% (w/w) NH4+, 1% (w/w) urea, 8% (w/w) P2O5, 11% (w/w) K2O, 2% (w/w) MgO, 0.24% (w/w) Cu, 0.2% (w/w) Fe, 0.03% (w/w) Mn, 0.009% (w/w) B, 0.008% (w/w) Mo, and 0.008% (w/w) Zn] (Scotts International B.V., Geldermalsen, the Netherlands), 10 polyethylene flotation bottles of 250 mL each (Kautex, Bonn, Germany) and 300 g of inoculum. Additionally, two floating cushions of 0.6 m × 0.7 m × 0.1 m each were constructed using 4 kg X-Oil®, 1.5 kg of Osmocote Pro® fertilizer, nine polyethylene bottles and the polypropylene hull material described above for application in a second mesocosm. The polyethylene bottles served as lifting bodies for the boom and cushions.
Both the experimental boom and the primary mesocosm basin (MC I) were spiked with a total of 2.5 L (0.5% v/v) of heat-sterilized and preheated Bunker C heavy fuel oil IFO 380 (Shell, Hamburg, Germany), containing 0.025% (v/v) squalene (2,6,10,15,19,23-hexamethyl-2,6,10,14,18,22-tetracosahexaene; Merck, Darmstadt, Germany) as an internal standard. The basin was aerated using compressed air from a central compressor system, sterilized tubing and four sterilized air stones and a 0.2-μm membrane filter (pore size, Sartorius, Göttingen, Germany), respectively. Temperature and concentration of dissolved oxygen were measured daily using an OxyScan Graphic electronic oxygen electrode (UMS, Meinersen, Germany) and are shown in Supporting Information, Fig. S2.
Samples from the seawater and the oil sorbent were taken twice a week over a period of 2 months. Two hundred and fifty litres of the content of MC I were transferred to a second basin (MC II) and filled up with seawater to 500 L each after 28 days. MC II was used for wastewater treatment. After 56 days of experiment, wastewater from both MC I and MC II was filtered through a sand filter. The presence of residual oil and its toxicity was tested with both indicator papers for oil residues (Macherey-Nagel, Düren, Germany) and an ARTOXKit M (Strategic Diagnostics Inc., Newark, DE) (Ronco et al., 1993). Briefly, eggs of Artemia franciscana were incubated in saline water supplied with the kit. Nauplius larvae were collected after 48 h of incubation at room temperature. Tenfold dilutions of the mesocosm water of both MC I and II were prepared in 24-well plates. Four wells were used as parallels for each dilution. Saline medium delivered with the kit was used as a blank, whereas a 10% solution of ethanol (v/v) in saline medium was applied as a positive control.
Inoculum of immobilized A. borkumensis biomass
The inoculum consisted of 250 g of X-Oil® oil sorbent material that was ethanol-sterilized and spiked with 50 g of heat-sterilized Bunker C heavy fuel oil (IFO 380, Shell). The oil-spiked sorbent was incubated in 10 L of ONR 7a medium (Dyksterhouse et al., 1995) at room temperature and under a constant oxygen supply via sterile tubing and a 0.2-μm-membrane air filter (Sartorius) for 14 days after the addition of 104 cells mL−1 of A. borkumensis SK2 (DSM 11573). Development of biofilms of A. borkumensis was monitored microscopically.
Extraction of nucleic acids
The biomass was collected by filtration of 200 mL of the water samples through a bottle-top filtration system (Nalgene, Hereford, UK) on the surface of 0.2-μm nitrocellulose filters (Sartorius). Extraction of total DNA was conducted using phenol–chloroform extraction (Wichels et al., 2004) for water samples, whereas DNA extraction of X-Oil® was performed with the DNA FastSpin Kit for Soil (Qbiogene). Extraction of total RNA was conducted using the RNA FastSpin Kit (Qbiogene).
Purification of total RNA and reverse transcription
Ten microlitres of the RNA extracts were diluted to a total volume of 500 μL using diethylpyrocarbonate-treated water (Ambion, Austin). Five hundred microlitres of acidic phenol–chloroform solution (5 : 1, pH 4.7; Ambion) was added and centrifuged for 10 min at 16 000 g in a microcentrifuge. The aqueous phases were transferred, treated with 50 μL of RNAse-free 3 M sodium acetate solution and precipitated with 1100 μL of ethanol at −20 °C overnight. The resulting extracts were tested by electrophoresis on 1.5% agarose gels.
Reverse transcription was conducted using the Reverse Transcription System Kit (Promega, Madison) according to the manufacturer's manual. Ten nanograms of RNA extracts were taken into each reaction.
PCR amplification of total DNA and cDNA
For amplification of DNA for a denaturing gradient gel electrophoresis (DGGE) fingerprinting analysis, a ‘touchdown’ PCR using the primers 341f-GC (5′-CGC CCG CCG CGC CCC GCG CCC GGC CCG CCG CCC CCG CCC CCC TAC GGG AGG CAG CAG CCT ACG GGA GGC AGC AG-3′) and 907r (5′-CCT ACG GGA GGC AGC AG-3′) (Muyzer et al., 1993) was performed as described previously (Sapp et al., 2007) to amplify PCR products of c. 566 bp. The PCR products were inspected on 1.2% agarose gels before DGGE. DNA extracts of 50–100 ng were taken into each reaction, whereas 10–50 ng of cDNA were applied for PCR amplification.
DGGE fingerprinting analysis, band excision and reamplification
DGGE was conducted according to the protocol of Muyzer et al. (1993) using a BioRad D-Code System (BioRad, Hercules). DGGE gels contained 6% (w/v) polyacrylamide denaturing gradient gels with linear gradients from 15% to 55% of denaturing agents [where 100% denaturant is 7 M urea and 40% (v/v) formamide] for analysis and primary excision of bands, respectively, and 20–50% gradients of denaturing agents [where 100% denaturant is 7 M urea and 40% (v/v) formamide] for secondary excision of reamplified DGGE bands. According to the signal intensity of each PCR product on agarose gels, 10–50 ng of each PCR product was loaded on the gels and run in 0.5 × TAE buffer at 100 V, 200 mA at 60 °C for 16 h. The resulting gels were stained with SYBR-Gold (Invitrogen, Carlsbad) diluted 1 : 10 000 in 0.5 × TAE buffer, and visualized using a gel documentation system (BioRad).
Extraction of DGGE bands, purification and sequencing of band DNA
For band excision, gels were transferred to a Dark Reader UV table (Clare Chemicals Ltd, Dolores) and excised using a scalpel and forceps. DNA was extracted as described previously (Wichels et al., 2004). Reamplification of the extracted DNA was performed using a modified protocol of Sambrook and Russel as described in Gertler et al. (2009). Fifty nanograms of the resulting DNA extracts were used as a template for the PCR reaction, applying the parameters described above. Using 20–50% denaturing gradients, PCR products were tested for purity in a second DGGE. In case of the presence of multiple bands in a sample, extraction and purification was repeated. Sequencing of purified PCR products was conducted as described previously. Sequences were compared with those deposited in the GenBank using the blast algorithm (http://www.ncbi.nlm.nih.gov/BLAST, Altschul et al., 1997). Results of the comparison are displayed in Tables S1–S3. The DNA sequences obtained in this study were deposited in the EMBL Data Library under the accession numbers FM210354–FM210408.
Phylogenetic analysis of DNA and cDNA sequences resulting from DGGE gels
For sequence assembly, the bioedit software (version 2.3–2.5) was used. A number of relevant 16S rRNA gene sequences from the microorganisms with validly published names were selected from GenBank as references. Multiple alignment of this set of sequences was conducted by clustalw software (http://www.ebi.ac.uk/clustalw/). The resulting alignment was refined by bioedit software and analysed with phylip 3.5 (Felsenstein, 1989) and mega 4.0 (Tamura et al., 2007). Maximum likelihood and neighbour-joining treeing algorithms were applied. Bootstrapping analysis was conducted using 100 sample trees for phylip 3.5 and 1000 trees for mega 4.0 phylogeny analysis, respectively. All these methods delivered phylogenetic trees with similar topologies.
Determination of total respiration
The intensity of oxygen consumption or biological oxygen demand in each sample was tested using a modified version of the Winkler method (Grasshoff et al., 1999). Briefly, 1.1 L of each mesocosm water sample was collected in a total of 18 60-mL Winkler bottles. Three bottles were examined for the determination of the initial oxygen concentration. On each of the following 5 days, further three bottles were examined using the same method.
Briefly, 0.25 mL of an 80% (w/v) MnCl2·4H2O solution and 0.25 mL of a 48% (w/v) NaOH, 15% (w/v) KI and 2% NaN3 solution were added to each bottle, mixed by inverting vigorously and incubated for a minimum of 60 min in the dark at room temperature. For acidification, 1 mL of a 19% (v/v) HCl solution was applied, the mixture was transferred into a glass vial and 1 mL of a 1% (w/v) starch solution added. The resulting blue colour was used as an indicator for a titration using 0.01 M NaS2O4 solution (Titrisol, Merck). Concentrations of dissolved oxygen were calculated as described before (Grasshoff et al., 1999). Decay curves and their corresponding slopes indicating the intensity of respiration were calculated using the prizm software, version 4.00 (Graph Pad Software Inc., http://www.graphpad.com) and displayed as the K-value in a separate diagram.
Photometric measurements of concentrations of mineral nutrients
Concentrations of dissolved ammonium, nitrate, nitrite and phosphate were determined by means of photometry as described previously (Grasshoff et al., 1999). For each sampling point, 500 mL of mesocosm water was filtered as described above. For determination of nutrients, measurements in triplicate were made using 15 mL of each filtrate. All measurements were conducted with a Lambda 12 photometer (Perkin Elmer, Waltham).
- Top of page
- Materials and methods
- Supporting Information
After 6 days of the experiment, an increasing oil emulsification was observed. The oil slick floating within the oil boom showed increased viscosity within the first 2 weeks of experiment (Fig. S1). Emulsified oil initially consisted of oil droplets of 100 μm up to several millimetres in diameter, and bacteria were visualized to be attached to the surface of oil droplets. The density of emulsified oil in MC I increased between days 8 and 21 and seemed to disappear in the following 7 days.
Following the splitting and refilling of MC I with fresh seawater on the day 28, a second increase of emulsification appeared from days 29 to 35, although the amount of biofilm aggregates decreased steadily after this point of time. Similar observations were made for the MC II. Both basins showed significant amounts of foam during the second half of the experiment. Oil slicks in MC I were reduced to small patches of 2–5 cm diameter, which disappeared by day 56. Biofilm aggregates could not be seen in basin II, whereas basin I contained less than five biofilm aggregates per litre of mesocosm fluid at the end of the experiment. Both mesocosms showed a strong abundance of marine protozoa, i.e. flagellates, ciliates and amoeba. No mortality effects upon Nauplius larvae were caused by any of the mesocosm water dilutions or the pure samples. No hydrocarbon contamination was detected at the end of the experiment in MC I or II.
Four different sample sets were taken from both DNA and RNA extracted from the water body and the surface of oil sorbent material X-Oil®. Three single bands from prior experiments have been used as length markers. One of these bands contained DNA of A. borkumensis SK2 for a quick identification of this organism in the DGGE band patterns. The other bands corresponded to uncultured bacteria. Fingerprints generated from DNA samples extracted from the water body of MC I showed a relatively high diversity, and no band correlating to A. borkumensis SK2 appeared over the whole course of the experiment (Fig. 1). Microbial communities showed high identities on days 10–18 and 42–56. Transitional communities with a higher number of bands could be detected on days 21 and 24. Band D14 displayed a 91% similarity to Sulfitobacter litoralis. Two other prominent bands, R13 and R24, present from days 10 to 56, proved to have a 94% identity to Yeosuana aromativorans, a well-known hydrocarbon-degrading member of CFB. Another DNA sequence with a 98% similarity to the hydrocarbon-degrading organism Lutibacterium annuloederans was affiliated to band R28. This band exclusively appeared on days 24 and 28.
Because of the severe autofluorescence of heavy fuel oil, FISH could not successfully be applied under the given experimental conditions. Therefore, the active fraction of bacteria was assessed through the extraction of small subunit (SSU) rRNA and its analysis via cDNA fingerprinting (Fig. 3). Fingerprinting gels of DNA samples collected from the surface of oil sorbent can be seen in Fig. 2. In contrast to the planktonic fraction, a band corresponding to A. borkumensis SK2 could be seen over the whole course of the experiment. The fingerprinting patterns of cDNA in Fig. 3 showed a significantly higher number of bands than the complementary dataset in Fig. 1. Microbial communities displayed significant changes from days 0 to 14, but very similar patterns subsequently. Again, it appeared that A. borkumensis SK2 is present in the sample set over the whole duration of the experiment. Sequences similar to both Y. aromativorans and L. annuloederans were detected. However, those similar to Y. aromativorans could be detected from day 14 up to the end of the experiment, whereas those similar to L. annuloederans were observed sporadically. DNA sequences similar to another organism connected to the degradation of hydrocarbons in prior studies, Phaeobacter inhibens, appeared among the band patterns on days 3–10. Bands R35 and R39, which were increasing in intensity from days 28 to 56, displayed high similarities to Roseobacter spp. The activity of the well-known bacterial predator Bacteriovorax spp. could as well be detected according to bands R28 and R38, indicating predation on bacterial biomass in this stage of the experiment.
Phylogenetic analysis of 16S rRNA gene and 16S rRNA sequences
After the analysis of all three datasets, the majority of sequences retrieved could be assigned to three major groups: the Alphaproteobacteria, the Gammaproteobacteria and the CFB. Figure 4 shows an overview of all sequences determined in this experiment in comparison with the closest culturable relatives with validly published names. Sequences of both Alpha- and Gammaproteobacteria displayed a rather low diversity with a few prominent phylotypes. SSU rRNA signatures of hydrocarbon-degrading microorganisms such as A. borkumensis, Y. aromativorans (Kwon et al., 2005) and L. annuloederans (Chung & King, 2001) were detected in both DNA and RNA samples. A few members of Deltaproteobacteria, such as Bacteriovorax spp., and Betaproteobacteria [an organism related to Pusillimonas noertemannii, DGGE band no. R33 (Stolz et al., 2005; Hilyard et al., 2008)] have also been found among the DNA sequences analysed. The last-mentioned bacterium is a degrader of aromatic compounds such as benzoic acid, a prominent intermediate metabolite of many pathways of the aromatic hydrocarbon degradation.
The majority of Alphaproteobacteria belonged to the genus Roseobacter or Sulfitobacter. These organisms were widespread in experiments that challenged seawater communities with high amounts of nutrients. They are primary colonizers of marine aggregates or ‘opportunistic’ bacteria that benefit from sudden availability of complex or monomeric carbon and energy sources in seawater (Eilers et al., 2000; Brakstad & Lodeng, 2005).
Another sequence detected in this study showed a 94% identity to that of Thalassospira lucentensis. Although within the genus Thalassospira, only Thalassospira tepidivorans possesses the ability to degrade certain aromatic hydrocarbons (Kodama et al., 2008), other members of this genus that lack this ability have been both isolated and detected in several oil-polluted environments and might consume intermediates of the marine oil degradation (López-López et al., 2002; López et al., 2006; Liu et al., 2007).
A close-up insight of Gammaproteobacteria demonstrated the importance of A. borkumensis. Although the Gammaproteobacteria comprise the vast majority of obligate hydrocarbon-degrading marine bacteria, only this species could be identified in the fingerprinting analysis. The phylogenetic analysis of CFB, in contrast, showed a broad variety of phylotypes similar to bacteria commonly found in seawater and some with a loose affiliation to Y. aromativorans.
Total respiration and nutrient concentrations
Total respiration takes into account the cumulative metabolism of both prokaryotes and eukaryotes, and indicates the intensity of the highly oxygen-demanding process of aerobic oil degradation. Total respiration for MC I and MC II is displayed in Fig. 5. Two peaks of intense respiration 200–300% above the average total respiration during the rest of the experiment have been detected on days 3 and 30 in the progression of seawater influx in MC I. Total respiration decreased quickly by >60% between days 3 and 7 and steadily decreased until day 18, followed by a slight increase in respiration for 7 days and a subsequent drop to a minimum on day 28. After the addition of fresh seawater and an increase in respiration by 1200%, the oxygen demand reduced again in the same way and remained constant until the end of the experiment. MC II displayed a similar behaviour as MC I between days 28 and 56 of the experiment.
Over the whole course of the experiment, a significant increase in both nitrate and ammonium could be seen in MC I (Fig. 6) and MC II (Fig. 7). Ammonium concentrations in MC I increased steadily to a maximum of 24 mmol L−1 and were reduced only after the refilling with fresh seawater on day 28. Nitrate concentration increased comparably, achieving maximum levels of 18 mmol L−1, but showed a slight decrease on day 7 as well as a slightly slower increase in the following days. The concentrations of both nutrients fluctuated strongly in the second half of the experiment. The progression of phosphate concentrations was equally unsteady and oscillated between 100 and 300 μmol L−1. Nitrite appeared in MC I from days 3 to 56 and showed maximum concentrations of 240 μmol L−1 on day 14 and 320 μmol mL−1 on day 24, respectively. MC II displayed unsteady concentrations of ammonium, nitrate and phosphate. Nitrite appeared shortly after the start of this mesocosm experiment at a concentration of 700 μmol L−1 and rapidly decreased afterwards (Fig. 6). The production of nitrite in an aerobic mesocosm is rather surprising. However, both mesocosms were clearly eutrophicated due to confinement of seawater in a batch style experiment. It can be assumed that nutrient concentrations may be lower in an actual offshore application, although this should be verified in a field study.
- Top of page
- Materials and methods
- Supporting Information
Oil biodegradation using experimental boom
The prototype constructed consisted of a number of components that have been tested individually for their effectiveness, but have never before been combined into a single application. The oil sorbent X-Oil® has already been used in previous microcosm experiments (Gertler et al., 2009) and was proved neutral on the formation of oil-degrading consortia. In fact, many oil sorbents have been tested for absorbing oil in aqueous solutions. Suni et al. (2004, 2007) proved that plant fibres of cotton grass (Eriophorum sp.), an easily accessible material from a sustainable source, very efficiently removes oil when used for filtration purposes and bioremediation. Many other plastic-polymer-based materials are currently in use and are commercially available. However, the interaction of oil sorbent surfaces and marine bacteria has been investigated in very few cases (Suni et al., 2007). The sorbent material X-Oil® applied in this experiment performed both the task of absorbing oil and providing a matrix for oil-degrading microorganisms. In fact, both processes are linked, because the oil bound to the sorbent surface acts as an energy and carbon source for oil-degrading microorganisms such as A. borkumensis. In order to achieve a sustainable growth of oil-degrading microorganisms, a slow-release fertilizer that was proven to be effective in prior mesocosm experiments (Xu et al., 2005) was applied. The results of our study showed that only a very small proportion of both ammonium and nitrate were utilized. In fact, eutrophication of both mesocosm basins with nutrients in the range of a few millimolars was observed. As a result of this eutrophication and high respiration rates, a significant decrease of pH and production of highly toxic nitrite occurred. The dosage of nutrients is clearly a major problem for offshore marine bioremediation, and the use of hydrophilic slow-release fertilizers might possibly not suit these applications. Although it can be assumed that nutrients can be retained within the oil sorbent matrix, the nutrients supplied by slow-release fertilizers can be effectively tested only in the open sea. Alternatively, oleophilic fertilizers such as guano- or uric acid-based fertilizers (Koren et al., 2003) might solve this problem due to their poor water solubility, hydrophobicity and ability to adhere to oil.
Both oxygen availability and supply are further key factors of bioremediation. Although oxygen is not limited in seawater, it may limit microbial oxidation of hydrocarbons within the oil booms presented in this study, due to the lack of oxygen permeability within the oil sorbent (Fig. S2), resulting in a localized production of nitrite within barrier microniches. Denitrification and nitrite production within microbial assemblages has been described before for flocs of sewage sludge (Schramm et al., 1999). Microprofiles of microbial flocs show decreasing oxygen levels and nitrite production within a 300-μm perimeter around the floc's surface. Both factors intensify below the floc's surface. Nitrite production in this study occurred at oxygen concentrations of c. 0.4 mg L−1 (Schramm et al., 1999). Although oxygen for the oxidation of hydrocarbons is provided in the boom for a certain period, a continuous supply is imperative for any successful future oil mitigation technique. This could be provided by the addition of chemical oxygen sources in the booms (Urrestarazu & Mazuela, 2005), electrical ventilation pumps or electrokinetical enhancement (Suni et al., 2004), if a ship-bourne or harbour-based source of electricity is available. Furthermore, extrapolating this to a real offshore situation, where nitrite diffusion occurs in an open system, the oxygen limitation would not necessarily be a problem.
Containment of oil, the primary task of the boom, was performed with undaunted success. Water samples showed no abundance of emulsified oil at the end of the experiment and no toxicity to nauplius larvae was detected, although nutrient levels indicated eutrophication of the mesocosms. Abundance and high levels of obligate oil-degrading microorganisms were detected within the boom and proven through the analysis of DNA and RNA samples from the whole community.
Role of A. borkumensis as a key organism for oil degradation in the mesocosms
Because of its global distribution and high adaptation, the gammaproteobacterium A. borkumensis clearly stands out among the cluster of marine hydrocarbonoclastic bacteria. Its unique metabolic (van Beilen et al., 2004) and genomic features (Schneiker et al., 2006), its competitiveness (Kasai et al., 2002b; Gertler et al., 2009) and its ubiquity (Yakimov et al., 2005, 2007) have been analysed extensively, and unequivocally recommend this organism for biodegradation purposes. Indeed, A. borkumensis has been introduced into the experimental system as an immobilized biomass that could be detected on the oil sorbent surface throughout the whole experiment. Surprisingly, though, this observation could not be made in the water body of the mesocosm. A comparison of DNA and RNA samples revealed the abundance of this microorganism on both oil sorbent and within the seawater in the experimental basin. While this observation indicates a high abundance of A. borkumensis on the sorbent surface, it also indicates low abundance and high activity in the mesocosm fluid. As flocks of emulsified oil and bacterial biomass have been detected in this fluid, this phenomenon could be due to either a depletion of oil within the flocks of biofilm or by a consumption of the biofilms by secondary consumers (Head et al., 2006). Indeed, protozoa were detected in high numbers in the vicinity of or on the surface of the emulsified oil, obviously affecting the biofilms. The presence of these mostly bacteriovorous heterotrophic Eukarya, therefore, might indicate grazing of bacterial cell mass or biofilms and emulsified oil as a whole.
Composition and dynamics of microbial communities
Although A. borkumensis affected the population composition in many micro- or mesocosm experiments, this study showed the abundance and appearance of a multitude of bacterial species. This is in contrast to a prior study that used a similar composition of experimental parameters but a microcosm design (Gertler et al., 2009). Reasons for this might include the elimination of ‘cultivation effects’ (Eilers et al., 2000) due to the larger scale of the experiment. Nevertheless, similar species belonging mostly to Alphaproteobacteria, Gammaproteobacteria and CFB were detected in both studies. Especially, prominent DNA sequences showed similarity to Roseobacter spp. and Sulfitobacter spp., which are well adapted to the availability of high levels of nutrients as well as complex or monomeric substances in seawater and tend to outgrow oligotrophic competitors among the natural communities in seawater (Eilers et al., 2000; Brakstad & Lodeng, 2005). Such features make these members of the Alphaproteobacteria the primary or early colonizers of any organic substrate or surface in marine environments. DNA-fingerprinting profiles showed the formation of several different consortia and at least two transitions over the course of the experiment, which may be caused by the limitation of oil as the major carbon and energy source, the dilution of MC I at day 28 or the availability of products resulting from the primary oil degradation (Head et al., 2006) for secondary consumers.
Three organisms with possible capabilities to degrade aromatic hydrocarbons were detected by DNA fingerprinting: Bands containing sequences similar to Y. aromativorans, (Kwon et al., 2005) a member of Flavobacteriaceae, were detected from day 14 to the end of the experiment in both RNA and DNA samples from the oil sorbents surface and the water body of MC I. Sequences similar to L. annuloederans (Chung & King, 2001) were detected in RNA samples at a late stage of the experiment, indicating the activity of multiple organisms degrading aromatic compounds. DNA sequences similar to T. lucentensis, a very enigmatic organism, were found both in this and a prior microcosm study conducted before (Gertler et al., 2009). Although many members of the genus Thalassospira were detected in oil-polluted freshwater and seawater habitats (López-López et al., 2002; López et al., 2006; Liu et al., 2007), the recently discovered species Thalassospira tepidiphila is the only member of the family yet known to degrade certain aromatic hydrocarbons (Kodama et al., 2008). The participation of multiple microorganisms in the biodegradation of a complex mixture of hydrocarbons such as heavy fuel oil has been simulated in a recent series of microcosms containing hydrocarbon mixtures (e.g. fuel oil) as well as single hydrocarbons. Four different species, A. borkumensis, Cycloclasticus pugettii, P. inhibens and Thalassolituus oleivorans, were proven to degrade specific classes of hydrocarbons in the same seawater sample (McKew et al., 2007). Although both A. borkumensis and P. inhibens were detected in our study, C. pugettii did not appear in either a microcosm (Gertler et al., 2009) or the mesocosm experiment using North Sea water, in contrast to many other studies performed in other locations (Kasai et al., 2002b; Yakimov et al., 2005, 2007; McKew et al., 2007). Reasons for this surprising observation could be the lack of aromatic hydrocarbons of industrial origin and the excessive abundance of polyphenolic compounds in algal biomass at this specific sampling site. However, the absence of specialized consumers of aromatic hydrocarbons such as C. pugettii is further proof of the functional redundancy of microbial communities in seawater (Griffiths, 2000; Giller et al., 2004).
DNA-fingerprinting analysis of RNA extracted from the water body of MC I furthermore displayed the activity of a Bacteriovorax species between days 21 and 35. This obligate parasite commonly lives in brackish or salt water and poses a threat to fast-growing populations of Gram-negative bacteria. Bacteriovorax species have been detected in two earlier studies involving marine microbial hydrocarbon degradation (Kasai et al., 2002a; Cappello et al., 2007), particularly in the initial phases of a mesocosm study (Cappello et al., 2007), although they could not be detected in the later stages of the same experiments. Bacteriovorax species naturally occur in marine assemblages of bacteria. They require a certain minimum of bacterial abundance to multiply (Kelley et al., 1997) and benefit from high concentrations of prey (Keya & Alexander, 1975): a biofilm offers this high concentration of prey immobilized in an extracellular matrix. Two studies document the activities of Bdellovibrio and Bacteriovorax strains on biofilms of different bacterial species (Núñez et al., 2003; Kadouri & O'Toole, 2005). Penetration of biofilms and predation of embedded microorganisms was demonstrated in both studies. Bdellovibrio species are reported to use a form of gliding motility (Núñez et al., 2003) when accessing biofilms and are capable of penetrating biofilms with a thickness of up to 30 μm (Kadouri & O'Toole, 2005).
However, the rRNA gene of Bacteriovorax could be detected only in the water body of the mesocosm in this study. A possible explanation for this could be the high nutrient or low oxygen concentrations inside the oil boom, toxicity of oil hydrocarbons or the concurrence of protozoan grazers. The effects of protozoan grazing on the microbial communities, therefore, are currently under investigation.
Microbial ecology of oil-degrading microbial communities in booms and water body
Although the research in the field of marine oil degradation has been intensively conducted for more than five decades, in practice, there is still a lack of a feasible technique for offshore oil spills. One of the obvious reasons for this is the complexity of marine microbial communities. Secondly, the high dissimilarity of experimental setups and procedures combined with the widespread distribution of sampling points all over the world gravely influenced by the ‘biogeography’ of oil-degrading microorganisms (Yakimov et al., 2005, 2007). Marine bioremediation and oil degradation have been studied on many different scales, ranging from large field studies (Lindstrom et al., 1991) to numerous microcosm experiments (Bachoon et al., 2001; Sei et al., 2003; McKew et al., 2007). However, the lack of comparability of these studies is a major problem, because microcosm studies tend to rapidly exhaust both oxygen and mineral nutrients, and thus promote the growth of bacteria (e.g. strains of Alteromonas, Vibrio or Colwellia (Eilers et al., 2000), which do not occur naturally in the comparable situations in open sea. Scaling-up experiments to mesocosms with sizes of several hundred to thousand litres (Cappello et al., 2007) is a potential solution to this problem. Future projects on biodegradation levels should therefore involve identical mesocosm experiments investigating both microbial oil-degrading communities and eukaryotic/prokaryotic grazers of these microorganisms.
- Top of page
- Materials and methods
- Supporting Information
We would like to thank Antje Wichels, Karl-Walther Klings, Kristine Carstens, Britta Knefelkamp and Alexandra Kraberg from Biologische Anstalt Helgoland for their help in conducting this work. We are indebted to Melanie Sapp for her kind assistance. Furthermore, we would like to thank Joachim Hellmann from Hellmann-Tech and Shell Marine Fuels, Hamburg, for providing materials and oil samples. K.N.T. acknowledges the Fonds der Chemischen Industrie for their generous support. This work was funded by GenoMik initiative of German Ministry for Science and Education (BMBF). The authors would like to thank Mark Malpass for proofreading the manuscript.
- Top of page
- Materials and methods
- Supporting Information
- 1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389–3402. , , , , , & (
- 1972) Degradation and mineralization of petroleum by two bacteria isolated from coastal waters. Biotechnol Bioeng 14: 297–308. & (
- 2001) Microbial community assessment in oil-impacted salt marsh sediment microcosms by traditional and nucleic acid-based indices. J Microbiol Meth 30: 37–49. , & (
- 2005) Microbial diversity during biodegradation of crude oil in seawater from the North Sea. Microb Ecol 49: 94–103. & (
- 2007) Predominant growth of Alcanivorax during experiments on ‘oil spill bioremediation’. in mesocosms. Microbiol Res 162: 185–190. , , , & (
- 1973) Crude oil degradation by microorganisms isolated from the marine environment. Z Allg Mikrobiol 13: 299–306. & (
- 2001) Isolation, characterization, and polyaromatic hydrocarbon degradation potential of aerobic bacteria from marine macrofaunal burrow sediments and description of Lutibacterium anuloederans gen. nov., sp. nov., and Cycloclasticus spirillensus sp. nov. Appl Environ Microb 67: 5585–5592. & (
- 1995) Cycloclasticus pugetii gen. nov., sp. nov., an aromatic hydrocarbon-degrading bacterium from marine sediments. Int J Syst Bacteriol 45: 116–123. , , , & (
- 2000) Succession of pelagic marine bacteria during enrichment: a close look at cultivation-induced shifts. Appl Environ Microb 66: 4634–4640. , & (
- 1989) PHYLIP – phylogeny inference package (version 3.2). Cladistics 5: 164–166. (
- 2009) Populations of heavy fuel oil-degrading marine microbial community in presence of oil sorbent materials. J Appl Microbiol, DOI:DOI: 10.1111/j.1365-2672.2009.04245.x. , , , & (
- 2004) Biodiversity effects on ecosystem functioning: emerging issues and their experimental test in aquatic environments. Oikos 104: 423–436. , & (
- 1999) Methods of Seawater Analysis. Wiley-VCH, New York. , , & (
- 2000) Ecosystem response of pasture soil communities to fumigation-induced microbial diversity reductions: an examination of the biodiversity–ecosystem function relationship. Oikos 90: 279–294. (
- 2003) Genome sequence completed of Alcanivorax borkumensis, a hydrocarbon-degrading bacterium that plays a global role in oil removal from marine systems. J Biotechnol 19: 215–220. , , et al. (
- 2006) Marine microorganisms make a meal of oil. Nat Rev Microbiol 4: 173–182. , & (
- 2008) Enrichment, isolation, and phylogenetic identification of polycyclic aromatic hydrocarbon-degrading bacteria from elizabeth river sediments. Appl Environ Microb 74: 1176–1182. , , & (
- 2007) The prestige oil spill: bacterial community dynamics during a field biostimulation assay. Appl Microbiol Biot 77: 935–945. , , , & (
- 2005) Susceptibility of biofilms to Bdellovibrio bacteriovorus attack. Appl Environ Microb 71: 4044–4051. & (
- 2002a) Bacteria belonging to the genus cycloclasticus play a primary role in the degradation of aromatic hydrocarbons released in a marine environment. Appl Environ Microb 68: 5625–5633. , & (
- 2002b) Predominant growth of Alcanivorax strains in oil-contaminated and nutrient-supplemented sea water. Environ Microbiol 4: 141–147. , , , , & (
- 1997) Effects of temperature, salinity, and substrate on the colonization of surfaces in situ by aquatic bdellovibrios. Appl Environ Microb 63: 84–90. , , & (
- 1975) Regulation of parasitism by host density: the Bdellovibrio–Rhizobium interrelationship. Soil Biol Biochem 7: 231–237. & (
- 2008) Thalassospira tepidiphila sp. nov., a polycyclic aromatic hydrocarbon-degrading bacterium isolated from seawater. Int J Syst Evol Micr 58: 711–715. , , , & (
- 2003) Petroleum pollution bioremediation using water-insoluble uric acid as the nitrogen source. Appl Environ Microb 69: 6337–6339. , , & (
- 2005) Yeosuana aromativorans gen. nov., sp. nov., a mesophilic marine bacterium belonging to the family Flavobacteriaceae, isolated from estuarine sediment of the South Sea, Korea. Int J Syst Evol Micr 56: 727–732. , , , & (
- 1991) Microbial populations and hydrocarbon biodegradation potentials in fertilized shoreline sediments affected by the T/V Exxon Valdez oil spill. Appl Environ Microb 57: 2514–2522. , , , , , & (
- 2007) Thalassospira xiamenensis sp. nov. and Thalassospira profundimaris sp. nov. Int J Syst Evol Micr 57: 316–320. , , , & (
- 2006) Phylogenetic analysis of a biofilm bacterial population in a water pipeline in the Gulf of Mexico. FEMS Microbiol Ecol 58: 145–154. , , , & (
- 2002) Thalassospira lucentensis gen. nov., sp. nov., a new marine member of the α-Proteobacteria. Int J Syst Evol Micr 52: 1277–1283. , , , , , & (
- 1999) Microbial population changes during bioremediation of an experimental oil spill. Appl Environ Microb 65: 3566–3574. , , , , & (
- 2007) Determining the identity and roles of oil-metabolizing marine bacteria from the Thames estuary, UK. Environ Microbiol 9: 165–176. , , , & (
- 1993) Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl Environ Microb 59: 695–700. , & (
- 2003) Investigations into the life cycle of the bacterial predator Bdellovibrio bacteriovorus 109J at an interface by atomic force microscopy. Biophys J 84: 3379–3388. , , , & (
- 2004) A biological oil adsorption filter. Mar Pollut Bull 49: 1006–1012. (
- 2004) Bacterial community dynamics and hydrocarbon degradation during a field-scale evaluation of bioremediation on a mudflat beach contaminated with buried oil. Appl Environ Microb 70: 2603–2613. (
- 1993) Pollution studies in the Rio Santiago basin, tributary of the Rio de la Plata estuary: preliminary risk assessment evaluation. B Environ Contam Tox 51: 657–664. , & (
- 2007) Bacterial community dynamics during the winter–spring transition in the North Sea. FEMS Microbiol Ecol 59: 622–637. , , & (
- 2006) Genome sequence of the ubiquitous hydrocarbon degrading marine bacterium Alcanivorax borkumensis. Nat Biotechnol 24: 997–1004. , , et al. (
- 1999) On the occurrence of anoxic microniches, denitrification, and sulfate reduction in aerated activated sludge. Appl Environ Microb 65: 4189–4196. , , , , , , , & (
- 2003) Monitoring of alkane-degrading bacteria in a sea-water microcosm during crude oil degradation by polymerase chain reaction based on alkane-catabolic genes. Environ Microbiol 5: 517–522. , , , & (
- 2005) Pusillimonas noertemannii gen. nov., sp. nov., a new member of the family Alcaligenaceae that degrades substituted salicylates. Int J Syst Evol Micr 55: 1077–1081. , , , & (
- 2004) Use of a by-product of peat excavation, cotton grass fibre, as a sorbent for oil-spills. Mar Pollut Bull 49: 916–921. , , , & (
- 2007) Removal by sorption and in situ biodegradation of oil spills limits damage to marine biota: a laboratory simulation. Ambio 36: 173–179. , , , , , , , & (
- 1996) Field evaluations of marine oil spill bioremediation. Microbiol Rev 60: 342–365. , & (
- 1983) Effect of microbial seeding of crude oil in seawater in a model system. Mar Biol 78: 13–20. , , , & (
- 2007) MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 24: 1596–1599. , , & (
- 2005) Effect of slow-release oxygen supply by fertigation on horticultural crops under soilless culture. Sci Hortic 106: 484–490. & (
- 2004) Characterization of two alkane hydroxylase genes from the marine hydrocarbonoclastic bacterium Alcanivorax borkumensis. Environ Microbiol 6: 264–273. , , , , , & (
- 2003) Evaluation of nonwoven polypropylene oil sorbents in marine oil-spill recovery. Mar Pollut Bull 46: 780–783. , , & (
- 2004) Bacterial diversity in toxic Alexandrium tamarense bloom off the Orkney Isles and the Firth of Forth. Helgoland Mar Res 58: 93–103. , , , , & (
- 2005) Bioremediation of oil-contaminated sediments on an intertidal shoreline using a slow-release fertilizer and chitosan. Mar Pollut Bull 51: 1062–1070. , , & (
- 1998) Alcanivorax borkumensis gen. nov., sp. nov., a new, hydrocarbon-degrading and surfactant producing marine bacterium. Int J Syst Bacteriol 48: 339–348. , , , , , & (
- 2005) Natural microbial diversity in superficial sediments of Milazzo Harbor (Sicily) and community successions during microcosm enrichment with various hydrocarbons. Environ Microbiol 7: 1426–1441. , , , , , , , & (
- 2007) Obligate oil degrading bacteria. Curr Opin Biotech 18: 257–266. , & (
- Top of page
- Materials and methods
- Supporting Information
Fig. S1. Experimental mesocosm basins MC I and MC II. Left figure shows Mesocosm basin I at day 10 of the experiment, whereas right figure shows Mesocosm basin II at day 30 of the experiment. Oil boom in MC I is stained in a black colour due to oil spiking. An oil slick can be seen in the upper right corner of MC I as well as within the ring shaped oil boom.
Fig. S2. Oxygen concentration in the water body and the oil boom (X-Oil®) in MC I over the course of the experiment. All measurements were conducted using an OxyScan Graphic electrode (UMS, Meinersen, Germany). Data points represent the mean of three measurements. Open diamonds represent the temperatures measured in the mesocosms' water body, whereas open triangles represent those measured in the oil boom. Full circles represent those measured in the mesocosms' water body, whereas black squares represent the oxygen concentrations measured in the oil boom. Due to extensive amounts of oil in the booms, no measurements were possible between days 0 and 8. Oxygen levels remained at 9±0.5 mg L-1 in the water body and below 1 mg L-1 in the booms after day 33.
Table S1. Sequence identities of DGGE bands that appear in Figs 1 and 4.
Table S2. Sequence identities of DGGE bands that appear in Figs 2 and 4.
Table S3. Sequence identities of DGGE bands that appear in Figs 3 and 4.
Please note: Wiley-Blackwell is 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.
|FEM_693_sm_suppinfofigstables.doc||4682K||Supporting info item|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.