Biodegradation of dispersed oil in seawater is not inhibited by a commercial oil spill dispersant

Marine Pollution Bulletin xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Biodegradation of dispersed oil in seawater is not inhibited by a commercial oil spill dispersant Odd G. Brakstada,⁎, Deni Ribicicb, Anika Winklerc, Roman Netzera a b c

SINTEF Ocean, Dept. Environmental Technology, Brattørkaia 17C, 7010 Trondheim, Norway The Norwegian University of Science and Technology, Dept. Cancer Research and Molecular Medicine, 7491 Trondheim, Norway Bielefeld University, Centre for Biotechnology (CeBiTec), 33501 Bielefeld, Germany

A R T I C L E I N F O

A B S T R A C T

Keywords: Oil Dispersants Dispersibility Biodegradation Seawater

Chemical dispersants are well-established as oil spill response tools. Several studies have emphasized their positive effects on oil biodegradation, but recent studies have claimed that dispersants may actually inhibit the oil biodegradation process. In this study, biodegradation of oil dispersions in natural seawater at low temperature (5 °C) was compared, using oil without dispersant, and oil premixed with different concentrations of Slickgone NS, a widely used oil spill dispersant in Europe. Saturates (nC10-nC36 alkanes), naphthalenes and 2- to 5-ring polycyclic aromatic hydrocarbons (PAH) were biotransformed at comparable rates in all dispersions, both with and without dispersant. Microbial communities differed primarily between samples with or without oil, and they were not significantly affected by increasing dispersant concentrations. Our data therefore showed that a common oil spill dispersant did not inhibit biodegradation of oil at dispersant concentrations relevant for response operations.

1. Introduction Chemical dispersants, consisting of mixtures of surfactants and solvents (Place et al., 2016; Prince, 2015), are used in marine oil spill response operations to reduce the surface tension of the oil by creating a more hydrophilic oil surface (Lessard and DeMarco, 2000). When oil on the sea surface is efficiently treated with dispersants, wave actions will generate dispersions of small oil droplets with near-to neutral buoyancies in the seawater, removing most of the oil from the sea surface (Tkalich and Chan, 2002), and generating dispersions with median oil droplet sizes typically between 20 and 50 μm (Brakstad et al., 2014; Lunel, 1993). Common stockpiled dispersants like Corexit 9500 and Slickgone NS, have shown comparable acute toxicities to marine copepods (Calanus finmarchicus), with LC50-values of 21–24 mg/ L, being considerably less toxic than the dispersed oil itself, which showed LC50 of 0.5–0.8 mg/L to the same copepod (Hansen et al., 2012; Hansen et al., 2014). Dispersants were also injected near the wellhead during the Deepwater Horizon oil spill in the Gulf of Mexico, resulting in reduced oil surfacing, and generating a deep sea plume of small oil droplets at 900–1300 m depth (Camilli et al., 2010). The expected goals of the dispersant treatment were to prevent large slicks in the area with many ships were gathered to stop the leak, and to prevent the oil from

⁎

impacting the shoreline (Atlas and Hazen, 2011). By encouraging the formation of small oil droplets, dispersants increase the surface-to-volume ratio of the oil, and enhance oil biodegradation. Faster n-alkane biodegradation was measured in chemically than physically dispersed Macondo oil in a flume basin study at breaking wave conditions, in accordance with smaller oil droplets generated in the chemically (median size < 20 μm) than the physically (median size appr. 80 μm) dispersed oil after 5 h of incubation (Brakstad et al., 2014). Colonized oil droplets eventually generate ‘flocs’ of biodegraded oil, microbial cells and extracellular polymers (Bælum et al., 2012; Hazen et al., 2010; Macnaughton et al., 2003). Several laboratory studies have shown positive effects of dispersants on oil biodegradation rates in seawater or with enrichment cultures at different conditions (Brakstad et al., 2014; Bælum et al., 2012; Hazen et al., 2010; McFarlin et al., 2014; Prince et al., 2013; Siron et al., 1995; Techtmann et al., 2017; Venosa and Holder, 2007). However, other studies have indicated negligible or uncertain effects of dispersants on oil compound biodegradation rates (Lindstrom and Braddock, 2002; Macnaughton et al., 2003; Swannell and Daniel, 1999). Data from recent studies even suggested that dispersants may suppress the activity of oil-degrading microbes (Kleindienst et al., 2015; Rahsepar et al., 2016). Once the oil is chemically dispersed, rapid dilution will occur in the water column. Environmental concentrations of dispersed oil over

Corresponding author. E-mail address: [email protected] (O.G. Brakstad).

http://dx.doi.org/10.1016/j.marpolbul.2017.10.030 Received 8 June 2017; Received in revised form 10 October 2017; Accepted 11 October 2017 0025-326X/ © 2017 Elsevier Ltd. All rights reserved.

Please cite this article as: Brakstad, O.G., Marine Pollution Bulletin (2017), http://dx.doi.org/10.1016/j.marpolbul.2017.10.030

Marine Pollution Bulletin xxx (xxxx) xxx–xxx

O.G. Brakstad et al.

Beckman Coulter Inc., Brea, CA, U.S.A.), fitted with either 100 μm or 280 μm apertures. These apertures were used for measurements of droplets within a diameter range of 2–60 μm or 5.6–100 μm, respectively. Oil only and DOR 1:100 were measured with 280 μm aperture, while DORs 1:25 and 1:10 were measured with 100 μm aperture. Sterile-filtered (0.22 μm) seawater was used as electrolyte in the Coulter Counter. Droplet concentrations were determined from volume concentrations (μm3/ml) and recalculated (mg/L), based on the density of the fresh oil (0.834 g/cm3).

the period of biodegradation is expected to be below 1 mg/L (Lee et al., 2013). It is therefore important to account for dilutions in biodegradation studies with chemically dispersed oil, by running experiments at low oil concentrations, although compromises may be made to enable measurements of oil compounds biodegradation. The objective of this study was to determine if increasing dispersant-to-oil ratios (DORs) of a common oil spill dispersant inhibited biodegradation and microbial communities rates of hydrocarbons at low oil concentrations (2–3 mg/L) in natural unamended seawater. At this concentration, biodegradation of nC10- to nC36-alkanes and 2- to 4-ring aromatic compounds was possible to measure. Slickgone NS is an approved dispersant in several European countries, including the use as a secondary tool in oil spill operations on the Norwegian Continental Shelf, and biodegradation studies were performed at a temperature relevant for this area (5 °C).

2.3.2. Chemical analyses Samples of dispersions and seawater blanks were solvent-solvent extracted (dichloromethane) for measurements of semivolatile organic compounds by gas chromatographic methods. The flask glass walls were also rinsed with DCM after removal of dispersions to extract material attached to the glass walls. Extracts of dispersions and glass walls were pooled. Total extractable organic carbon (TEOC) was analyzed by GF-FID, while quantification of 96 individual targeted compounds (nC10-nC36 alkanes, decalins, phenols, 2- to 5-ring polycyclic aromatic hydrocarbons (PAH) and 17α(H),21β(H)-Hopane) was performed by GC–MS analyses, as previously described (Brakstad et al., 2014). In the GC–MS analyses, response values for individual target analytes were determined, and based on a signal-to-noise ratio of > 10, the lower limit of detections (LOD) varied from 0.001 μg/L to 0.01 μg/L for individual oil compounds. Target analytes were normalized against 17α(H),21β(H)-Hopane (Prince et al., 1994; Wang et al., 1998).

2. Materials and methods 2.1. Oil, dispersant and seawater Fresh Statfjord C paraffinic oil (batch 2014–0081), provided from Statoil ASA, Mongstad, Norway, was pre-heated at 50 °C for 30 min to melt the wax in the oil, and cooled to room temperature. The dispersant Slickgone NS was provided from K. Todnem AS, Sandnes, Norway. The Statfjord oil was pre-mixed with Slickgone in DORs of 1:100, 1:25 and 1:10. Seawater was collected via a pipeline system at 80 m depth (below thermocline, salinity 34‰) in a Norwegian fjord (Trondheimsfjord; 63°26′N, 10°23′E), outside the harbour area of Trondheim. The seawater temperatures was 5.9 °C at the time of collection. The seawater was acclimated overnight to 5 °C before start of biodegradation experiment.

2.3.3. Analyses of total cell concentrations and most probable number determinations Total prokaryote concentrations (epifluorescence microscopy) and most probable number (MPN) determinations of culturable heterotrophic prokaryotes were performed as previously described (Brakstad et al., 2008).

2.2. Biodegradation experiment The three pre-mixed dispersions, and the oil without dispersant, were prepared in natural unfiltered seawater at room temperature with an oil droplet generator, as previously described (Brakstad et al. 2015a; Nordtug et al. 2011). Oil droplet concentrations of 200 mg/L and droplet sizes of 10 μm were pre-set for the droplet generator system. Based on oil droplet concentration measurements (Coulter Counter; see below), each stock dispersion was diluted in natural unfiltered seawater (5 °C) to reach a final nominal concentration of 2 mg/L oil droplets. This oil concentration did not require additional mineral nutrient amendment, as previously shown (Brakstad et al. 2015a; Prince et al. 2013). The dispersions were distributed in baked (450 °C) and autoclaved flasks (2 L; Schott), completely filled and capped without headspace or air bubbles, and flasks were mounted on a carousel system with slow continuous rotation (0.75 r.p.m.), as previously described (Brakstad et al. 2015a). Dispersant in seawater without oil (0.2 mg/L; corresponding to concentration in DOR 1:10) was run through the oil droplet generator and distributed on carousels similarly to the oil dispersions. The carousel system was placed in a temperature-controlled room at 5 °C for 64 days in the dark, and flasks were sacrificed for analyses after 30 min incubation (0 days) in triplicate (oil without dispersant and premixed oil in DOR 1:25), duplicate (premixed oil in 1:100), or as single samples (premixed oil in DOR 1:10). Flasks were then sacrificed for analyses after 7, 14, 21, 28 and 64 days of incubation, as described for 0-day samples. Flasks, completely filled with unfiltered seawater without oil or dispersant, were incubated at the same conditions and used as seawater blanks.

2.3.4. 16 S rRNA gene amplicon sequencing Seawater samples without oil and oil dispersions (approximately 500 mL) were filtered through 0.22 μm filters (Millipore), and DNA was extracted from filters by employing FastDNA Spin kit for soil (MP Biomedicals), according to the manufacturer's instructions. DNA quantification was performed by a Qubit 3.0 fluorometer (Thermo Fisher Scientific Waltham, MA, U.S.A.), with dsDNA High Sensitivity kit (ThermoFisher Scientific, MA, U.S.A.). Microbial community compositions in samples were analyzed by 16S rRNA gene amplicon sequencing, using a primer combination targeting the V3 and V4 hypervariable regions of bacterial 16S rRNA genes and 2 × KAPA HiFi HotStart ReadyMix (Klindworth et al., 2013). Both forward and reverse primers were tagged with Illumina adapter overhang, and amplicons generated by PCR (Eppendorf Mastercycler), were isolated using PCR clean-up spin columns (Qiagen; QIAquick Gel Extraction Kit), and thereafter used in index PCR with Illumina NEXTERA XT indexes (Nextera XT Index Primer 1 (N7xx) and 2 (S5xx)) and 2× KAPA HiFi HotStart ReadyMix. Indexed 16S rRNA amplicons were isolated using PCR clean-up spin columns (Qiagen; QIAquick Gel Extraction Kit), and thereafter verified and quantified using Qubit 3.0. All DNA indexed amplicons were diluted to 4 nM in Tris-HCl (10 mM, pH 8.5) and pooled for generation of a library for sequencing. DNA library and Illumina positive control PhiX samples were thereafter denatured and incubated at 96 °C for 2 min, before running in an Illumina MiSeq sequencer, according to Illumina's 16S rRNA sample preparation guideline. Raw pair-end reads were assembled with fastq-join in QIIME 1.9.1. Assembled sequences were demultiplexed and quality filtered to remove low quality reads (Phred score < 20; −q 19). UCHIME was employed for chimera detection on assembled quality filtered reads. Operational Taxonomic Units (OTUs) were determined by clustering assembled sequences on 97% nucleotide identity, using UCLUST with open reference clustering option. Representative sequences were

2.3. Analyses 2.3.1. Oil droplet size and concentrations Concentrations and size distributions of oil droplets were determined by Coulter Counter measurements (Beckman Multisizer 4; 2

Marine Pollution Bulletin xxx (xxxx) xxx–xxx

O.G. Brakstad et al.

The oil droplet concentrations were reduced in the dispersions during the biodegradation experiment, faster in the dispersions with oil only (no dispersant) than in the dispersions with Slickgone (Fig. S1, Supplementary Material). After 14 days of incubation, droplet concentrations were 50% of the start concentrations in the dispersions without Slickgone, compared to 75–87% in the premixed dispersions. After 28 days the corresponding values were 22% in the dispersions without dispersants, and 40–58% in the premixed dispersions, increasing from DOR 1:100 to 1:10 (Fig. S1). The reductions in oil droplet concentrations were related to emergences of visible oil ‘flocs’ in the treatments. Also the median oil droplet sizes in the dispersions were reduced, ranging from 6.3–8.1 μm after 14 days and 4.1–9.8 μm after 28 days (results not shown). Some of the oil will also attach to the glass walls of the flasks, and recent studies have shown that the material attached to the glass wall at the end of the experiments represents approximately 10–15% of the original TEOC concentrations, when dispersions of approximately 2 mg/L dispersed oil are used (Brakstad et al., 2015a). As previously described, the oil compounds will gradually attach to the glass walls, preferentially larger alkylated 3-ring, 4- to 5-ring PAH, decalines and n-alkanes (Brakstad et al., 2015a). However, despite ‘floc’ formation and glass wall attachment, we have experienced that the initial oil droplet size distribution is important for the oil compound biodegradation (Brakstad et al., 2015a).

aligned with PyNAST and taxonomy assignment was performed with RDP classifier based on SILVA-113 database. In order to visualize differences in taxonomical composition, relative abundances for OTUs on each sampling point were calculated. For the purpose of statistical analysis of OTUs, DESeq2 (Love et al., 2014), an R package, was used to standardize the counts between samples rather than rarifying to the number of reads present in the sample with smallest number of reads (McMurdie and Holmes, 2014). Diversity and statistical analysis was performed with the Phyloseq package v.1.12.2 (McMurdie and Holmes, 2013) in R-studio v.3.2.1. Next, the raw reads of each OTU were converted to relative abundances and taxa were agglomerated to the desired taxonomic level for the purpose of OTU comparison across incubation days and treatments. 2.3.5. Other analyses Dissolved oxygen (DO) and water temperatures were determined by a DO meter (YSI, Inc., Yellow Springs, OH, U.S.A.). 2.4. Data treatment and statistics Non-linear regression analyses were performed by the option “onephase decay” in GraphPad Prism vs. 6.0 (GraphPad Software Inc., La Jolla, CA, U.S.A.). Half-lives were determined from rate coefficients (t1/2 = 0.693/k1). Since sterilized controls were not included we are unable to categorically exclude abiotic losses, but none of the analytes considered here have significant volatilities that might have led to evaporative losses during the experiments or subsequent oil-extraction and analysis. One-way Anova analyses were performed in GraphPad Prism vs. 6. Principal component analyses (PCoA) plots of microbial community dynamics were calculated based on Weighted-UniFrac distance metric in R-studio v.3.2.1 using Phyloseq package v.1.12.2, taking into account the relative abundance of species/taxa shared between samples (Lozupone et al., 2007).

3.2. Oxygen consumption During the first 21-day period of the experiment, when most of the n-alkane and PAH biotransformation appeared (Fig. 1), DO saturations in oil dispersions ranged from 47 to 57% (4.7–5.7 mg/L DO), compared to 88–89% (8.7–8.8 mg/L DO) in samples without oil (Fig. S2, Supplementary Material). Thus, no DO limitation therefore appeared in this period. At the end of the experiment (64 days), DO saturation was reduced to 17–43% (1.7–4.3 mg/L DO) in oil dispersions, compared to 88–89% (8.7–8.8 mg/L DO) in samples without oil (Fig. S2). The DO saturation was highest in oil dispersions not premixed with Slickgone (43 ± 4%), while DO in premixed dispersions averaged 26 ± 9% (ranging from 17 to 35%) after day 64. The lowest DO saturation was measured in the oil dispersion with DOR 1:25 (Fig. S2), probably explained by its higher initial concentration (Table 1).

3. Results and discussions 3.1. Oil droplet sizes and concentrations Initial median oil droplet sizes, prepared in the oil droplet generator, decreased with increasing dispersant concentrations. The droplet sizes were larger in dispersions not premixed with Slickgone (Table 1), in agreement with previous studies using the oil droplet generator (Nordtug et al., 2011). However, oil droplet sizes were within ranges expected for dispersant-treated oil (Brakstad et al., 2014; Lunel, 1993). The initial droplet concentrations ranged between 2.1 and 3.5 mg/L (Table 1), as determined by Coulter Counter analyses, and in agreement with the nominal concentrations of 2 mg/L oil. TEOC concentrations, determined by GC-FID analyses, were in agreement with the Coulter Counter analyses, except in dispersions of oil only (Table 1).

3.3. Biotransformation of targeted oil compounds Since flasks used in the experiment were completely filled with dispersion (no headspace), no evaporation should appear, and biodegradation was expected to be the sole depletion process (Brakstad et al., 2015a). Biotransformation of n-alkanes (nC10-nC36) and 2- to 5-ring PAH was determined after normalization against the persistent biomarker 17α(H),21β(H)-Hopane (Prince et al., 1994; Wang et al., 1998). The transformation of total nC10-nC36 alkanes and 2- to 5-ring PAH were comparable in all dispersions, with or without Slickgone, and no differences in n-alkane and PAH biotransformations were measured in the oil dispersions at the end of the experiment (Fig. 1). Biotransformation of targeted n-alkanes showed faster depletion of nC10-nC24 alkanes than nC25-nC36 alkanes (Fig. S3, Supplementary Material), and only small differences in n-alkane biodegradation were measured between the dispersions without Slickgone and with different DORs. Differences in biotransformation of 2- to 4-ring PAH became apparent between C0- to C2- and C3- to C4- alkyl-substituted PAH during the biodegradation period, and with only minor differences between dispersions with and without Slickgone. (Fig. S4, Supplementary material). The 4-ring PAH (pyrenes/fluoranthenes and chrysenes) showed partial biotransformation after 28 days, and were mainly biotransformed by ≥50% after 64 days in all dispersions. Typically, biotransformation was related to the alkyl-substitution of the 2- to 4-ring aromatics (Fig. S4). Biotransformation rates were determined for targeted oil

Table 1 Oil droplet concentrations and median oil droplet sizes (Coulter Counter analyses), and TEOC (GC-FID analyses), at the start of the experiment with dispersed Statfjord oil (30 min after final dilutions in seawater). Treatment

Oil only DOR 1:100 DOR 1:25 DOR 1:10A Dispersant only Seawater A

Oil concentrations (mg/L ± SD) Droplets (Coulter Counter)

TEOC (GC-FID)

2.42 ± 0.07 2.11 ± 0.05 3.50 ± 0.89 2.1 nd nd

1.33 ± 1.99 ± 3.47 ± 2.01 0.02 ± < 0.01

0.17 0.01 0.46 0.01

Median oil droplet sizes (μm)

13.8 ± 6.3 9.0 ± 3.8 6.4 ± 2.6 6.6 nd nd

Only one replicate.

3

Marine Pollution Bulletin xxx (xxxx) xxx–xxx

% of day 0 (normalized against 17α(H),21 β(H)-Hopane

% of day 0 (normalized against 17α (H),21 β(H)-Hopane

O.G. Brakstad et al.

Fig. 1. Depletion of nC10-nC36 alkanes (A) and 2- to 5-ring PAH (B) in dispersions with Statfjord crude oil. The groups were normalized against 17α(H),21β(H)-Hopane and depletion determined as percentage of normalized ratios at day 0.

A 120

No Dispersant DOR 1:100 DOR 1:25 DOR 1:10

100 80 60 40 20 0 0

20

40

60

80

60

80

Days

B 120 100 80 60 40 20 0 0

20

40

Days

oil concentration (Table 1) in this dispersion. It should also be noted that the nC30-nC36 alkanes constituted only 6.5% and the 4-ring PAH 0.5% by weight of the oil compounds measured by the GC–MS method. The biotransformation rates determined in this study were in agreement with those measured elsewhere (Brakstad et al., 2015a; Bælum et al., 2012; McFarlin et al., 2014; Prince and Butler, 2014; Prince et al., 2016; Prince et al., 2013; Siron et al., 1995; Techtmann et al., 2017; Venosa and Holder, 2007). No inhibition of oil compound biotransformation rates was therefore determined here, even at a DOR of 1:10. Our data do not support recent results, claiming that dispersants may suppress biodegradation of oil compounds in cold natural seawater, using water-accommodated fractions (WAFs) prepared with or without dispersant (Kleindienst et al., 2015). It should be noted, that also in this WAF study, only small differences were measured between oil compound biodegradation in treatments with or without dispersant (Kleindienst et al., 2015), and these differences would have little environmental relevance (Prince et al., 2016).

compounds and used for calculations of half-lives (Brakstad et al., 2015a; Wang et al., 2016). The half-lives of the nC10-nC36 alkanes increased by chain-length in all treatments, ranging from 1.4–2.1 days for nC11 to 29.6–45.2 days for nC36 (Fig. 2A; Table S1, Supplementary material). Comparison of the half-lives between the treatments by oneway Anova resulted in a p-value of 0.599, showing no significant differences between the data sets. The half-lives of selected 2- to 4-ring PAH increased both by higher aromatic ring-numbers and by increased alkyl-substitution (Fig. 2B), as shown also in the depletion curves (Fig. S4, Supplementary material). For example, naphthalene showed halflives of 6.1 to 10.1 days, while half-lives of C4-naphthalene ranged between 10.7 and 12.5 days. Corresponding data for chrysenes were 36.6–73.1 days for chrysene and 62.2 to > 100 days for C3-chrysene (Table S2, Supplementary material). Also for the PAH there were no statistically significant differences between these data sets (p = 0.237). Our data did therefore not indicate any inhibitory effects of the dispersant on oil biotransformation. The dispersions with DOR 1:25 experienced moderately slower biodegradation of the larger n-alkanes (nC30-nC36) and 4-ring PAH (fluoranthenes and chrysenes), with higher concentrations of these compounds left in the DOR 1:25 than in the other oil dispersions (Fig. S3 and S4). This may be attributed to the low DO concentrations at the end of the experiment (Fig. S2), presumably caused by the high initial

3.4. Microbial concentrations and community successions Total cell counts and counts of viable heterotrophic prokaryotes (MPN-analyses) increased in all dispersions, as well as in samples with dispersant alone. Peak levels were reached after 14–21 days, with 4

Marine Pollution Bulletin xxx (xxxx) xxx–xxx

O.G. Brakstad et al.

A Half-life (days)

60

Oil alone DOR 1:100 DOR1:25 DOR 1:10

40

20

nC -1 0 nC -1 1 nC -1 2 nC -1 3 nC -1 4 nC -1 5 nC -1 6 nC -1 7 nC -1 8 nC -1 9 nC -2 0 nC -2 1 nC -2 2 nC -2 3 nC -2 4 nC -2 5 nC -2 6 nC -2 7 nC -2 8 nC -2 9 nC -3 0 nC -3 1 nC -3 2 nC -3 3 nC -3 4 nC -3 5 nC -3 6

0

B

>

Half-life (days)

100

>

>

80 60 40 20

Naphthalenes

Fluorenes

Phenanthrenes/ Anthracenes

Dibenzothiophenes

Fl C Py 1Fl r C /Py 2Fl r C /Py 3Fl r /P yr C h C r 1ch C r 2ch C r 3ch r

F C 1F C 2F C 3F P/ A C 1P C 2P C 3P C 4P D B T C 1D B C 2- T D C BT 3D C BT 4D B T

N C 1N C 2N C 3N C 4N

0

Fluoranthenes/ Pyrenes

Chrysenes

Fig. 2. Half-lives of n-alkanes (A) and 2- to 4-ring PAH (B) in dispersed Statfjord crude oil determined from first-order rate coefficients, see also Tables S2–S3. The results are averages of replicate samples (see Figs. S3 and S4). Compounds with half-life > 100 days are marked with “ > ”.

(64 days), Polaribacter also became abundant in samples without oil, probably associated with heterotrophic activities of these bacteria in cold seawater (Nikrad et al., 2012). Comparison of bacterial communities by PCoA showed 4 clusters (Fig. 5). One of these (cluster A) was associated with the samples from the start of the experiment, both with and without oil. Incubated samples with seawater only and seawater with dispersant from days 7 to 28 were also located in cluster A, but no incubated oil-containing samples. After day 64, samples without oil were located in another cluster (cluster B). The clusters C and D included all the communities in the oil-containing samples, and no samples without oil were present here. Cluster C were probably related to the high abundances of Oleispira at day 7, while cluster D were associated with Polaribacter and Cycloclasticus. The increased abundance of Polaribacter in samples without oil after 64 days may have resulted in cluster B moving in the direction of cluster D. These results are in agreement with data from recent oil biodegradation studies at 5 °C, showing that there was no significant differences between microbial communities in physically dispersed oil and in oil dispersions treated with Corexit 9500 (DOR of 1:25), while the differences between oil dispersions and samples with Corexit without oil were significantly different (Techtmann et al., 2017). Also in those studies, members of Oleispira were abundant in all oil dispersions, while Cycloclasticus became abundant during PAH degradation. None of these genera were abundant in samples with dispersant only (Techtmann et al., 2017).

subsequent declines (Fig. 3). The microbial growth period from days 0 to 21 corresponded well with the rapid oil degradation during this period. No inhibitions of microbial concentrations were measured at any of the DORs compared to the oil alone (no dispersant). Microbial community analyses showed that the samples at the start of the study were predominated by Gammaproteobacteria of the genera Colwellia and Oleispira (Fig. 4). Large fractions of microbes in these samples were not identified to any genus. After 7 days of incubation, distinct community differences were observed between samples without oil (seawater blanks and seawater with dispersant only) and with oil dispersions (oil alone and dispersions at the different DORs). Colwellia was still predominant in the samples without oil, while both Oleispira and Colwellia were abundant the oil-containing samples (Fig. 4). High abundances of Colwellia were determined in samples of seawater without oil also after 14–28 days of incubation, while Oleispira was mainly succeeded by Polaribacter and Cycloclasticus in all oil-containing samples. Colwellia is abundant in cold uncontaminated or oilpolluted seawater (Brakstad et al., 2015b; Bælum et al., 2012; Hazen et al., 2010; Prabagaran et al., 2007; Redmond and Valentine, 2012), while Oleispira, Polaribacter and Cycloclasticus are associated with oilcontaminated marine environments (Brakstad et al., 2008; Dubinsky et al., 2013; Harayama et al., 2004; Prabagaran et al., 2007). Members of Oleispira and Cycloclasticus are considered to be obligate hydrocarbonoclastic bacteria (Yakimov et al., 2007). Oleispira and Polaribacter are associated with degradation of saturated and unsaturated hydrocarbons, harbouring alkB genes for alkane monooxygenases (Guibert et al., 2016; Kube et al., 2013). Cycloclasticus is able to transform a variety of monoaromatic hydrocarbons and PAH (Geiselbrecht et al., 1998; Kasai et al., 2003; Wang et al., 2008), in agreement with the slower PAH than n-alkane degradation (Fig. 1; Fig. 4; Fig. S3; Fig. S4). Towards the end of the biodegradation period

4. Conclusions In conclusion, our results indicate that oil biotransformation with a commercially available dispersant at environmentally relevant concentrations (Lee et al., 2013; Li et al., 2009) is essentially unaffected by 5

Marine Pollution Bulletin xxx (xxxx) xxx–xxx

O.G. Brakstad et al.

A 7

Log10 Cells/ml

6 5

Oil alone DOR 1:100 DOR1:25 DOR 1:10 Dispersant alone Seawater

4 3 2 1 0

20

40

60

80

Days

B

Fig. 5. PCoA plot of microbial community clustering during oil biodegradation. The colours describe the treatments (as explained in Fig. 4), and the numbers the days of incubation. The clustering and their letters are explained in the text.

7

Log10 MPN/ml

6

diluted, the droplets do not recoalesce. At these low concentrations there is no reason to expect dispersant to stimulate or suppress oil compound biodegradation, and the dispersant components may serve as an additional carbon source for microbial activity. While our study was performed with Slickgone NS, a commercial dispersant approved as an oil spill response tool in several European countries, most other biodegradation studies have been performed with Corexit 9500. Surfactant composition is known in Corexit (Place et al., 2016), while no such information is available for Slickgone. However, comparable acute toxicities to marine copepods have been reported for the dispersants (Hansen et al., 2014).

5 4 3 2 1 0

20

40

60

80

Days Fig. 3. Microbial concentrations (Log10) in experiment with Statfjord crude oil, measured as total counts by epifluorescence microscopy (A) and MPN-concentrations of heterotrophic prokaryotes (B).

Acknowledgements

the presence of the dispersant within DOR ranges relevant for oil spill response operations. Neither biotransformation, microbial concentrations, nor microbial communities were negatively affected by the increasing dispersant concentrations in the experiment, and typically hydrocarbonoclastic microbes were abundant in all oil dispersions. These results are fully consistent with the hypothesis that the role of the dispersant is to facilitate oil dispersion by lowering the interfacial tension between oil and water so that less energy is required to form droplets. Once the oil is dispersed and allowed to become sufficiently

This study was supported by the Norwegian Research Council Petromaks2 program (contract #228271-E30) and the oil companies Statoil Petroleum AS, ExxonMobil Exploration and Production Norway AS, AkerBP ASA, TOTAL E & P Norge ASA, and ConocoPhillips Skandinavia AS. We will thank Thor-Arne Pettersen, Inger Steinsvik, Marianne Unaas Rønsberg and Inger K. Almås, Marianne Aas, Marianne A. Molid, Bror Johansen, and Daniel F. Krause for chemical analyses and technical assistance.

Fig. 4. Relative abundances of microbial genera representing > 1% of the total community populations. The results are shown at the start of the degradation period (0), and after 7, 14, 21, 28 and 64 days of incubation. Samples represent seawater without oil and dispersant (Blank), dispersant in seawater without oil (Dispersant), dispersed oil premixed with Slickgone (DOR 1:10, 1:25 and 1:100) and dispersed oil not premixed with dispersant (Oil).

6

Marine Pollution Bulletin xxx (xxxx) xxx–xxx

O.G. Brakstad et al.

Lindstrom, J.E., Braddock, J.F., 2002. Biodegradation of petroleum hydrocarbons at low temperature in the presence of the dispersant Corexit 9500. Mar. Pollut. Bull. 44, 739–747. Love, M.I., Huber, W., Anders, S., 2014. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550. Lozupone, C.A., Hamady, M., Kelley, S.T., Knight, R., 2007. Quantitative and qualitative β diversity measures lead to different insights into factors that structure microbial communities. Appl. Environ. Microbiol. 73, 1576–1585. Lunel, T., 1993. Dispersion: Oil Droplet Size Measurements at Sea, International Oil Spill Conference. American Petroleum Institute, pp. 794–795. Macnaughton, S.J., Swannell, R., Daniel, F., Bristow, L., 2003. Biodegradation of dispersed Forties crude and Alaskan North Slope oils in microcosms under simulated marine conditions. Spill Sci. Technol. Bull. 8, 179–186. McFarlin, K.M., Prince, R.C., Perkins, R., Leigh, M.B., 2014. Biodegradation of dispersed oil in arctic seawater at-1 C. PLoS One 9, e84297. McMurdie, P.J., Holmes, S., 2013. Phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One 8, e61217. McMurdie, P.J., Holmes, S., 2014. Waste not, want not: why rarefying microbiome data is inadmissible. PLoS Comput. Biol. 10, e1003531. Nikrad, M.P., Cottrell, M.T., Kirchman, D.L., 2012. Abundance and single-cell activity of heterotrophic bacterial groups in the western Arctic Ocean in summer and winter. Appl. Environ. Microbiol. 78, 2402–2409. Nordtug, T., Olsen, A.J., Altin, D., Meier, S., Overrein, I., Hansen, B.H., Johansen, Ø., 2011. Method for generating parameterized ecotoxicity data of dispersed oil for use in environmental modelling. Mar. Pollut. Bull. 62, 2106–2113. Place, B.J., Perkins, M.J., Sinclair, E., Barsamian, A.L., Blakemore, P.R., Field, J.A., 2016. Trace analysis of surfactants in Corexit oil dispersant formulations and seawater. Deep Sea Res. Part II: Top. Studies Oceanogr. 129, 273–281. Prabagaran, S.R., Manorama, R., Delille, D., Shivaji, S., 2007. Predominance of Roseobacter, Sulfitobacter, Glaciecola and Psychrobacter in seawater collected off Ushuaia, Argentina, sub-Antarctica. FEMS Microbiol. Ecol. 59, 342–355. Prince, R.C., 2015. Oil spill dispersants: boon or bane? Environ. Sci. Technol. 49, 6376–6384. Prince, R.C., Butler, J.D., 2014. A protocol for assessing the effectiveness of oil spill dispersants in stimulating the biodegradation of oil. Environ. Sci. Pollut. Res. 21, 9506–9510. Prince, R.C., Elmendorf, D.L., Lute, J.R., Hsu, C.S., Haith, C.E., Senius, J.D., Dechert, G.J., Douglas, G.S., Butler, E.L., 1994. 17. alpha.(H)-21. beta.(H)-hopane as a conserved internal marker for estimating the biodegradation of crude oil. Environ. Sci. Technol. 28, 142–145. Prince, R.C., McFarlin, K.M., Butler, J.D., Febbo, E.J., Wang, F.C., Nedwed, T.J., 2013. The primary biodegradation of dispersed crude oil in the sea. Chemosphere 90, 521–526. Prince, R.C., Butler, J.D., Redman, A.D., 2016. The rate of crude oil biodegradation in the sea. Environ. Sci. Technol. 51, 1278–1284. Rahsepar, S., Smit, M.P., Murk, A.J., Rijnaarts, H.H., Langenhoff, A.A., 2016. Chemical dispersants: oil biodegradation friend or foe? Mar. Pollut. Bull. 108, 113–119. Redmond, M.C., Valentine, D.L., 2012. Natural gas and temperature structured a microbial community response to the Deepwater Horizon oil spill. Proc. Natl. Acad. Sci. 109, 20292–20297. Siron, R., Pelletier, E., Brochu, C., 1995. Environmental factors influencing the biodegradation of petroleum hydrocarbons in cold seawater. Arch. Environ. Contam. Toxicol. 28, 406–416. Swannell, R.P., Daniel, F., 1999. Effect of Dispersants on Oil Biodegradation Under Simulated Marine Conditions, International Oil Spill Conference. American Petroleum Institute, pp. 169–176. Techtmann, S.M., Zhuang, M., Campo, P., Holder, E., Elk, M., Hazene, T.C., Conmy, R., Santo Domingo, J.W., 2017. Corexit 9500 enhances oil biodegradation and changes active bacterial community structure of oil-enriched microcosms. Appl. Environ. Microbiol. 83, e03462-16. Tkalich, P., Chan, E.S., 2002. Vertical mixing of oil droplets by breaking waves. Mar. Pollut. Bull. 44, 1219–1229. Venosa, A., Holder, E., 2007. Biodegradability of dispersed crude oil at two different temperatures. Mar. Pollut. Bull. 54, 545–553. Wang, Z., Fingas, M., Blenkinsopp, S., Sergy, G., Landriault, M., Sigouin, L., Lambert, P., 1998. Study of the 25-year-old Nipisi oil spill: persistence of oil residues and comparisons between surface and subsurface sediments. Environ. Sci. Technol. 32, 2222–2232. Wang, B., Lai, Q., Cui, Z., Tan, T., Shao, Z., 2008. A pyrene-degrading consortium from deep-sea sediment of the West Pacific and its key member Cycloclasticus sp. P1. Environ. Microbiol. 10, 1948–1963. Wang, J., Sandoval, K., Ding, Y., Stoeckel, D., Minard-Smith, A., Andersen, G., Dubinsky, E.A., Atlas, R., Gardinali, P., 2016. Biodegradation of dispersed Macondo crude oil by indigenous Gulf of Mexico microbial communities. Sci. Tot. Environ. 557, 453–468. Yakimov, M.M., Timmis, K.N., Golyshin, P.N., 2007. Obligate oil-degrading marine bacteria. Curr. Opin. Biotechnol. 18, 257–266.

Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.marpolbul.2017.10.030. References Atlas, R.M., Hazen, T.C., 2011. Oil biodegradation and bioremediation: a tale of the two worst spills in U.S. history. Environ. Sci. Technol. 45, 6709–6715. Bælum, J., Borglin, S., Chakraborty, R., Fortney, J.L., Lamendella, R., Mason, O.U., Auer, M., Zemla, M., Bill, M., Conrad, M.E., Malfatti, S.A., Tringe, S.G., Holman, H.-Y., Hazen, T.C., Jansson, J.K., 2012. Deep-sea bacteria enriched by oil and dispersant from the Deepwater Horizon spill. Environ. Microbiol. 14, 2405–2416. Brakstad, O.G., Nonstad, I., Faksness, L.-G., Brandvik, P.J., 2008. Responses of microbial communities in Arctic Sea Ice after contamination by crude petroleum oil. Microb. Ecol. 55, 540–552. Brakstad, O.G., Daling, P.S., Faksness, L.-G., Almås, I.K., Vang, S.-H., Syslak, L., Leirvik, F., 2014. Depletion and biodegradation of hydrocarbons in dispersions and emulsions of the Macondo 252 oil generated in an oil-on-seawater mesocosm flume basin. Mar. Pollut. Bull. 84, 125–134. Brakstad, O.G., Nordtug, T., Throne-Holst, M., 2015a. Biodegradation of dispersed Macondo oil in seawater at low temperature and different oil droplet sizes. Mar. Pollut. Bull. 93, 144–152. Brakstad, O.G., Throne-Holst, M., Netzer, R., Stoeckel, D.M., Atlas, R.M., 2015b. Microbial communities related to biodegradation of dispersed Macondo oil at low seawater temperature with Norwegian coastal seawater. Microb. Biotechnol. 8, 989–998. Camilli, R., Reddy, C.M., Yoerger, D.R., Van Mooy, B.A., Jakuba, M.V., Kinsey, J.C., McIntyre, C.P., Sylva, S.P., Maloney, J.V., 2010. Tracking hydrocarbon plume transport and biodegradation at Deepwater Horizon. Science 330, 201–204. Dubinsky, E.A., Conrad, M.E., Chakraborty, R., Bill, M., Borglin, S.E., Hollibaugh, J.T., Mason, O.U., Piceno, M., Reid, F.C., Stringfellow, W.T., Tom, L.M., Hazen, T.C., Andersen, G.L., 2013. Succession of hydrocarbon-degrading bacteria in the aftermath of the deepwater horizon oil spill in the Gulf of Mexico. Environ. Sci. Technol. 47, 10860–10867. Geiselbrecht, A.D., Hedlund, B.P., Tichi, M.A., Staley, J.T., 1998. Isolation of marine polycyclic aromatic hydrocarbon (PAH)-degrading Cycloclasticus strains from the Gulf of Mexico and comparison of their PAH degradation ability with that of Puget Sound Cycloclasticus strains. Appl. Environ. Microbiol. 64, 4703–4710. Guibert, L.M., Loviso, C.L., Borglin, S., Jansson, J.K., Dionisi, H.M., Lozada, M., 2016. Diverse bacterial groups contribute to the alkane degradation potential of chronically polluted subantarctic coastal sediments. Microb. Ecol. 71, 100–112. Hansen, B.H., Altin, D., Olsen, A.J., Nordtug, T., 2012. Acute toxicity of naturally and chemically dispersed oil on the filter-feeding copepod Calanus finmarchicus. Ecotoxicol. Environ. Saf. 86, 38–46. Hansen, B.H., Altin, D., Bonaunet, K., Øverjordet, I.B., 2014. Acute toxicity of eight oil spill response chemicals to temperate, boreal, and arctic species. J. Toxicol. Environ. Health A 77, 495–505. Harayama, S., Kasai, Y., Hara, A., 2004. Microbial communities in oil-contaminated seawater. Curr. Opin. Biotechnol. 15, 205–214. Hazen, T.C., Dubinsky, E.A., DeSantis, T.Z., Andersen, G.L., Piceno, Y.M., Singh, N., Jansson, J.K., Probst, A., Borglin, S.E., Fortney, J.L., 2010. Deep-sea oil plume enriches indigenous oil-degrading bacteria. Science 330, 204–208. Kasai, Y., Shindo, K., Harayama, S., Misawa, N., 2003. Molecular characterization and substrate preference of a polycyclic aromatic hydrocarbon dioxygenase from Cycloclasticus sp. strain A5. Appl. Environ. Microbiol. 69, 6688–6697. Kleindienst, S., Seidel, M., Ziervogel, K., Grim, S., Loftis, K., Harrison, S., Malkin, S.Y., Perkins, M.J., Field, J., Sogin, M.L., 2015. Chemical dispersants can suppress the activity of natural oil-degrading microorganisms. Proc. Nat. Acad. Sci. 112, 14900–14905. Klindworth, A., Pruesse, E., Schweer, T., Peplies, J., Quast, C., Horn, M., Glöckner, F.O., 2013. Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucl. Acids Res. 41 (e1-e1). Kube, M., Chernikova, T.N., Al-Ramahi, Y., Beloqui, A., Lopez-Cortez, N., Guazzaroni, M.E., et al., 2013. Genome sequence and functional genomic analysis of the oil-degrading bacterium Oleispira antarctica. Nat. Commun. 4, 2156. Lee, K., Nedwed, T., Prince, R.C., Palandro, D., 2013. Lab tests on the biodegradation of chemically dispersed oil should consider the rapid dilution that occurs at sea. Mar. Pollut. Bull. 73, 314–318. Lessard, R.R., DeMarco, G., 2000. The significance of oil spill dispersants. Spill Sci. Technol. Bull. 6, 59–68. Li, Z., Lee, K., King, T., Boufadel, M.C., Venosa, A.D., 2009. Evaluating crude oil chemical dispersion efficacy in a flow-through wave tank under regular non-breaking wave and breaking wave conditions. Mar. Pollut. Bull. 58, 735–744.

7