Systematic and Applied Microbiology 38 (2015) 189–197

Contents lists available at ScienceDirect

Systematic and Applied Microbiology

j ourna l h omepage: www.elsev ier .de /syapm

Bacterial diversity of polluted surface sediments in the northernAdriatic Sea

Marino Korlevic a,1, Jurica Zuckob,1, Mirjana Najdek Dragic a, Maria Blazinaa,Emina Pustijanacc, Tanja Vojvoda Zeljkod, Ranko Gacesab, Damir Baranasicb,Antonio Starcevicb, Janko Diminicb, Paul F. Longe,f, John Cullumg,Daslav Hranuelib, Sandi Orlic a,h,∗

a Centre for Marine Research, Ruder Boskovic Institute, Rovinj, Croatiab Faculty of Food Technology and Biotechnology, University of Zagreb, Pierottijeva 6, 10000 Zagreb, Croatiac University Juraj Dobrila Pula, Pula, Croatiad Division of Molecular Biology, Ruder Boskovic Institute, Zagreb, Croatiae Institute of Pharmaceutical Science, King’s College London, Franklin-Wilkins Building, Stamford Street, London SE1 9NH, UKf Department of Chemistry, King’s College London, Franklin-Wilkins Building, Stamford Street, London SE1 9NH, UKg Department of Genetics, University of Kaiserslautern, Postfach 3049, 67653 Kaiserslautern, Germanyh Division of Material Chemistry, Ruder Boskovic Institute, Zagreb, Croatia

a r t i c l e i n f o

Article history:Received 28 August 2014Received in revised form 6 March 2015Accepted 9 March 2015

Keywords:BacteriaSea sedimentsPetroleum compoundsAlkane degradationPhylogenetic classification

a b s t r a c t

Samples were collected from sea sediments at seven sites in the northern Adriatic Sea that included sixsites next to industrial complexes and one from a tourist site (recreational beach). The samples wereassayed for alkanes and polycyclic aromatic hydrocarbons. The composition of the hydrocarbon samplessuggested that industrial pollution was present in most cases. A sample from one site was also grownaerobically under crude oil enrichment in order to evaluate the response of indigenous bacterial popula-tions to crude oil exposure. Analysis of 16S rRNA gene sequences showed varying microbial biodiversitydepending on the level of pollution – ranging from low (200 detected genera) to high (1000+ genera)biodiversity, with lowest biodiversity observed in polluted samples. This indicated that there was con-siderable biodiversity in all sediment samples but it was severely restricted after exposure to crude oilselection pressure. Phylogenetic analysis of putative alkB genes showed high evolutionary diversity of theenzymes in the samples and suggested great potential for bioremediation and bioprospecting. The firstsystematic analysis of bacterial communities from sediments of the northern Adriatic Sea is presented,and it will provide a baseline assessment that may serve as a reference point for ecosystem changesand hydrocarbon degrading potential – a potential that could soon gain importance due to plans for oilexploitation in the area.

© 2015 Elsevier GmbH. All rights reserved.

Introduction

More than one third of the human population lives in coastalareas and on small islands. Anthropogenic contamination andaccidental oil spills are a major source of marine pollution thatcause huge damage to the environment, resulting in extensive and

∗ Corresponding author at: Ruder Boskovic Institute, 10000 Bijenicka 54, Zagreb,Croatia. Tel.: +385 1 457 1282.

E-mail address: sorlic@irb.hr (S. Orlic).1 Equal contribution.

long-term deterioration of coastal ecosystems. Natural oil seep-age and marine oil transport accidents provide a constant input ofpollutants to marine environments. Research on microbial alkanedegradation started a century ago, with a publication by Söhngen[42] on microorganisms responsible for the disappearance of oilspills on surface waters. Many different enzymes involved in hydro-carbon degradation have been characterized to date and almost200 bacterial, cyanobacterial, algal and fungal genera, representingmore than 500 species, are known to thrive on hydrocarbons. Crudeoil is perhaps one of the most complex mixtures of organic com-pounds on Earth that is typically high in saturated and aromatichydrocarbons [40]. Alkanes are rather inert chemical substances

http://dx.doi.org/10.1016/j.syapm.2015.03.0010723-2020/© 2015 Elsevier GmbH. All rights reserved.

190 M. Korlevic et al. / Systematic and Applied Microbiology 38 (2015) 189–197

and the first stage in their biological degradation is an activationstep. In aerobic degradation of linear alkanes, the activating enzymeis often a monooxygenase, which produces a terminal alcoholgroup. A variety of monooxygenases are known that have differ-ent substrate specificities for short (C2–C4), medium (C5–C17) andlong chains (>C18). One of the most intensively investigated groupsof enzymes is the family of AlkB integral membrane non-heme ironmonooxygenases, which are common in strains degrading mediumchain length alkanes [40].

Alkane degrading microorganisms are usually highly enrichedin samples from polluted sites, and some recently characterizedbacterial species have been shown to be highly specialized forhydrocarbon degradation. These species are called obligate hydro-carbonoclastic bacteria, and they play a key role in the removalof hydrocarbons from polluted and non-polluted environments[40,48,50].

Degradation of pollutants in contaminated environments is, inmany cases, carried out by microbial food webs rather than by asingle species. As a large proportion of environmental microorgan-isms cannot be grown in the laboratory, a proper understanding ofthe microbial communities present can only be gained by analysingthe metagenome (i.e. characterising the DNA sequences of all theorganisms present in the sample). In the last decade, functionalgenomics technologies have been introduced into research onmicrobial alkane degradation [45].

The coastal region of the Kvarner Gulf and the Istrian Peninsulaare the best-developed tourist regions in the northern Adriatic Sea.However, the oldest shipyard and oil refinery on the Croatian coastare located in the two most important cities (Pula and Rijeka). TheINA oil refinery is a medium-sized refinery located in Kvarner Baynear the city of Rijeka. The refinery processes approximately 3.5million tons of crude oil annually and manufactures a large numberof petroleum products. The Uljanik shipyard in Pula Bay (est. 1856)is the most significant shipyard in the Croatian part of the AdriaticSea.

Little is known about the distribution of polluting hydrocar-bons in the region and its effect on microbial populations. The firstanalyses of PAHs in marine sediments within Rijeka Bay started in1998 at three sampling sites offshore from the petroleum refineryfacilities, and they were extended in 1999 to three more sam-pling points in front of the repair shipyard within the same easternindustrial zone [2]. The concentrations of PAHs were consider-ably higher in the shipyard environment in comparison with thepetroleum refinery area. A declining trend of total PAHs and con-sequent toxicity indices was observed at all sites [2]. In a recentstudy by Treven [44], historical trends (2007–2012) of polycyclicaromatic hydrocarbon (PAH) pollution in surface sediments in thenorthern Adriatic Sea (Croatia) were assessed, and the results indi-cated a relatively low concentration of PAHs in marine sedimentsin the proximity of the oil refinery located in Kvarner Bay. Pro-posals to exploit oil and gas reserves in the Adriatic make it evenmore important to establish the present distribution of pollutionand to prepare for likely further pollution associated with newindustries.

For this study, seven sampling points of coastal surface sea sedi-ments were investigated. Six of the sites were subject to long-termindustrial contamination due to the proximity of major indus-trial complexes and the other site in Svezanj Cove was chosenas a “clean” point. In addition, one of the samples was subjectedto enrichment with crude oil (SF1). The hydrocarbon composi-tions of the samples were analysed by gas chromatography/massspectrometry. The composition of the bacterial populations wasdetermined by analysis of 16S rRNA sequences, and the occur-rence of the alkB gene was used as a marker for the presence ofalkane-degrading bacteria. These studies formed the basis for fur-ther metagenomic studies.

Fig. 1. Research area and sampling stations in the northern Adriatic Sea.

Materials and methods

Area description, sampling locations and processing

The coastal surface sediment samples (0–3 cm depth) werecollected from seven locations within Rijeka and Pula Bays inMay 2012. The exact locations are shown in Fig. 1. The stationsin Rijeka Bay (sandy sediments, average 60 m depth) were fac-ing a former coke plant area (B; latitude: 45.304745, longitude:14.539268), petroleum refinery facilities (I; latitude: 45.277181,longitude: 14.538066) and a tanker berth (TV; latitude: 45.276365,longitude: 14.549654). The stations in Pula Bay (clay sediments,average 15 m depth) were facing a marine gas station (BP; latitude:44.873499, longitude: 13.847572), and Uljanik repair shipyard’svent (BI; latitude: 44.869053, longitude: 13.836787) and cove(BZ; latitude: 44.870969, longitude: 13.836186). As a referencepoint, a sample was also taken from Svezanj Cove (US; lati-tude: 45.296504, longitude: 14.494250) and represented a touristicrecreational zone that should not be polluted by oil spills. ASCUBA diver collected the sediment samples manually with grav-ity core samplers. The sediment samples were freeze-dried for72 h and fully homogenized. The coastal sediment from the tankerberth station (TV) was subsampled into a sterile Erlenmeyer flaskand grown under crude oil pressure in aerobic (SF1) conditionsusing the method described by Wang et al. [47] for four monthsbefore the nucleic acids were extracted and sequenced (Table 1)(Fig. 2).

M. Korlevic et al. / Systematic and Applied Microbiology 38 (2015) 189–197 191

Tabl

e

1N

umbe

r

of

sequ

enci

ng

read

s

gene

rate

d

by

16S

rRN

A

sequ

enci

ng, u

nres

olve

d

com

plex

mix

ture

(UCM

)

conc

entr

atio

ns, e

stim

ated

OU

T

rich

ness

(* cal

cula

ted

usin

g

QIIM

E

wit

h

a

97%

iden

tity

leve

l)

and

dive

rsit

y an

d ri

chne

sses

tim

ator

s

for

surf

ace

sedi

men

t sam

ples

. The

95%

confi

denc

e

inte

rval

s

are

give

n

in

brac

kets

.

Acr

onym

Stat

ions

UCM

(!g

g−1

d.w

.)N

r

of

read

sRA

WN

r

of

read

spa

ssin

g

QC

OTU

s

per

sam

ple*

Alp

ha

dive

rsit

y

esti

mat

ers

Chao

1

obse

rved

OTU

s

Shan

non

ACE

B

Form

er

coke

plan

tar

ea13

.00

8986

6822

821

959

(946

; 972

)

789

(785

; 793

)

7.96

1071

(105

4;

1088

)

I

Petr

oleu

m

refin

ery

faci

litie

s1.

81

5513

4468

768

945

(933

; 957

)

731

(725

; 736

)

7.69

1050

(103

5;

1065

)

TV

Tank

er

bert

h

10.7

5

15,9

00

12,7

39

434

636

(603

; 669

)

401

(397

; 404

)

5.27

751

(712

; 791

)U

S

Svez

anj C

ove

9.52

10,2

36

8398

516

673

(659

; 688

)

472

(469

; 476

) 6.

12

800

(785

; 814

)BP

Mar

ine

gas

stat

ion

1469

.2

36,4

95

30,4

71

490

625

(616

; 635

)

476

(473

; 478

) 5.

37

701

(689

; 714

)BI

Repa

ir

ship

yard

Ulja

nik

(ven

t)16

7.48

14,6

47

12,3

10

244

245

(243

; 248

)

240

(238

; 241

)

4.44

256

(253

; 260

)

BZ

Repa

ir

ship

yard

Ulja

nik

(cov

e)49

.77

10,0

78

8185

926

1041

(103

6;

1045

)

923

(921

; 924

)

9.08

1148

(114

3;

1153

)

SF1

Sedi

men

t TV

inEr

lenm

eyer

flask

–gr

own

aero

bica

lly

–

7316

4261

201

203

(201

; 205

)

198

(197

; 199

)

4.36

211

(209

; 213

)

Fig. 2. Pyrogenic vs petrogenic polycyclic aromatic hydrocarbons in samples ofnorthern Adriatic Sea coastal sediments. Abbreviations used are: anthracene (AN),phenanthrene (PH), fluoranthene (FLU) and pyrene (PY).

Extraction and analysis of hydrocarbons

The homogenized sediment samples were analysed accord-ing to the MADEP-EPH-04 method [22]. Briefly, 3 × 40 mL ofdichloromethane (DCM), hexane (2:1) and 10 !L of surrogatestandard mixture (chloro-octadecane and ortho-terpenyl) solu-tion (final concentration 40 mg L−1) were added to 10 g of drysediment samples. All sediment samples were vortexed and ultra-sonicated for 60 min to extract hydrocarbons. Two partitions(1 mL each) of hydrocarbons were divided into aliphatic and aro-matic fractions by passing through a glass column containing5 g silica gel and 1 g anhydrous Na2SO4. The first fraction con-taining aliphatic hydrocarbons was eluted with 20 mL hexaneand the second containing aromatic hydrocarbons with 20 mLDCM.

Extracts (1 !L) were analysed using an Agilent 6890N GCequipped with an Agilent 5973 Network Mass Selective Detector, aHP-5MS capillary column (25 m × 0.3 mm × 0.25 !m; cross-linked5% phenylmethylsiloxane) with ultrapure helium (6.0) as carriergas. The samples were injected in the split-less mode at an injec-tion temperature of 280 ◦C. The column temperature was initiallyheld at 35 ◦C for 2 min, raised to 140 ◦C at the rate of 5 ◦C min−1,then to 300 ◦C at the rate of 10 ◦C min−1, and held at this tempera-ture for 15 min. The detector temperature was kept at 280 ◦C. Theidentity of polycyclic aromatic hydrocarbons (PAHs) in the sampleswas confirmed by the retention time and abundance of quantifi-cation/confirmation ions in the authentic aliphatic and aromatichydrocarbon standards. The concentrations of hydrocarbons wereexpressed as a dry weight (d.w.).

In the case of aliphatic hydrocarbons, the individual n-alkaneconcentrations (n-C12 to n-C34), the concentrations of the iso-prenoids pristane and phytane, the total resolved aliphatics (sumof n-alkanes and isoprenoids) and the unresolved complex mixture(UCM) were calculated for each sampling site. The UCM, compris-ing different unresolved compounds (e.g. branched alkanes andcycloalkanes) that appear as a peak below the baseline in a chro-matogram, was calculated using the average response factors foraliphatic hydrocarbons and PAHs, respectively. In the case of PAHs,

192 M. Korlevic et al. / Systematic and Applied Microbiology 38 (2015) 189–197

concentrations of the individual PAH, total PAH concentration(sum of resolved PAHs) and UCM (as described above) were calcu-lated. The ascertained PAHs were: naphthalene, acenaphthylene,acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene,pyrene, benzo(a)anthracene, chrysene, benzo(b)fluoranthene,benzo(k)fluoranthene, benzo(a)pyrene, indeno(1,2,3-c,d)pyrene,dibenzo(a,h)anthracene and benzo(g,h,i)perylene.

The ratios of anthracene (AN) to phenanthrene (PH) and flu-oranthene (FLU) to pyrene (PY) determined the PAH sources,since ratios of AN/(PH + AN) < 0.1 and FLU/(FLU + PY) < 0.4 usuallyimply a petrogenic source, whereas ratios of AN/(PH + AN) > 0.1 andFLU/(FLU + PY) > 0.5 suggest a pyrogenic source (grass, wood, or coalcombustion). If the FLU/(FLU + PY) ratio is between 0.4 and 0.5 com-bustion of petroleum origin is suggested.

The concentrations of the n-alkanes with different chain lengths,as well as the UCM, were calculated for each sample. The ratio ofUCM to resolved alkanes (U:R) greater than 2.5 is characteristic forsediments with considerable input of petroleum and combustionproducts or weathered and degraded petroleum residues.

Extraction and sequencing of DNA

Total DNA was isolated from 1 g of all sediments with theUltraClean® Soil DNA Isolation Kit (MO BIO, USA), accord-ing to the manufacturer’s instructions. The column waseluted in 60 !L 10 mM Tris. The bacterial V1-V2 16S rRNAregion was sequenced at MR DNA (Molecular Research LP;www.mrdnalab.com, Shallowater, TX, USA) using the bac-terial 16S-based tag-encoded FLX amplicon pyrosequencing(bTEFAP) method [15], Roche/454 FLX titanium instru-ment and reagents following the manufacturer’s guidelines.Primers used for target 16S rRNA sequence amplifica-tion were 27Fmod (5′-AGRGTTTGATCMTGGCTCAG-3′) and519Rmodbio (5′-GTNTTACNGCGGCKGCTG-3′). Primers usedfor target alkB gene sequence amplification were alkB-1f 5′-AAYACNGCNCAYGARCTNGGNCAYAA (coding for the peptideNTAHELGHK′) and alkB-1r 5′-GCRTGRTGRTCNGARTGNCGYTG(coding for QRHSDHHA′) [28].

Taxonomic classifications of bacterial communities

In total, 109,171 raw pyrotagged reads from eight samples(Table 1) were processed and demultiplexed using the QIIME1.8 package [9]. For removal of errors arising from 454 pyrose-quencing, the AmpliconNoise [37] algorithm was used, followedby Perseus [38] for removal of chimeras. A total of 87,654 readspassed quality control and were subsampled to match the small-est sample (SF1–4261 reads). Subsampled reads were clusteredinto operational taxonomic units (OTUs) using the UCLUST algo-rithm [17] with a default identity of 97% using QIIME’s de novoOTU picking protocol. All OTU clusters with less than 2 mem-bers (singletons) were filtered from further analysis. For eachOTU, representative sequences were aligned to the Greengenes[13] core reference set using PyNAST [8]. Taxonomy was assignedusing the Greengenes 16S rRNA gene database released in May2013 – version 13 8. Phylogenetic trees were constructed usingFastTree, as implemented in QIIME 1.8. Various alpha and betadiversity metrics were calculated using the QIIME platform. Foralpha diversity metrics, samples were rarefied from 300 to 3300reads in steps of 500, while for beta diversity the rarefaction depthwas 3300.

Phylogenetic analysis of alkB genes

For all seven sampling sites, putative alkB gene sequences wereamplified by PCR, cloned and sequenced as in [28], with the

degenerate primers targeting the central region of the gene andamplifying fragments of approximately 550 bp. The assignmentof sequences to alkB was tested using a hidden Markov model(HMM) profile based on KEGG orthologue K00496 [27]. HMManalysis was carried out using HMMER v3.0 [16]. After in silicotranslation, the sequences were tested against the HMM profile andsequences with a score greater than 200 were considered to belongto AlkB enzymes. In total, 174 out of 290 sequences passed qual-ity filtering. Additionally, metagenomic sequences were obtainedfrom one sample in which TV sediment had been incubated insterile marine water and light petroleum crude oil medium (12%final concentration) for 120 days shaking (150 rpm) at 15 ◦C (sam-ple SF1) (manuscript in preparation; North Adriatic sedimentmetagenome project), and they were searched with the HMM pro-file of KEGG ortholog K00496. Identified sequences were furtheranalysed to ensure only complete genes were recovered. A totalof 19 complete genes homologous to AlkB were identified in themetagenomic sequence and were included in phylogenetic anal-yses. The inferred AlkB protein sequences were aligned to theK00496 HMM profile using HMMER with the addition of sevenAlkB protein sequences belonging to the K00496 ortholog clus-ter. Phylogenetic trees of AlkB sequences were constructed usingthe neighbor-joining (NJ), minimum evolution (ME) and maxi-mum likelihood (ML) methods, as implemented in MEGA6 [43].For NJ and ME phylogenetic trees, the JTT model was used fordistance estimation and rate variation among sites was modelledwith a gamma distribution. For the ML tree, the Le Gascuel 2008model was used and evolutionary rate differences among sites weremodelled with a gamma distribution. For the ML method, the sub-stitution model was selected using the Model feature implementedin MEGA6. For all trees, bootstrapping was carried out with 1000replications.

Results

Sample hydrocarbon analysis

The concentrations of n-alkanes varied widely (Table 2) withmaximum values recorded at the marine gas station (BP) and withinthe older parts of the shipyard area (BI). As expected, gas chro-matography of the n-alkanes showed a peak of the unresolvedcomplex mixture (UCM) corresponding to a combination of unre-solved compounds (e.g. branched alkanes and cycloalkanes; Fig.S1), which was particularly prominent at sampling sites B, BP, BIand BZ. The concentrations of the n-alkanes with different chainlengths, as well as the UCM, were calculated for each sample (TableS1). Biogenic n-alkanes tend to have an odd number of carbon atomsand this results in a high value of the carbon preference index(CPI), which can be used to determine the degree of biogenic versuspetrogenic input [32].

The concentrations of 16 different PAHs were measured (TableS2), and the total concentration of PAHs at each site is shownin Table 2. The level of PAH pollution can be classified as fol-lows [5]: (a) low, <0.1 !g g−1; (b) moderate, 0.1–1 !g g−1; (c) high,1–5 !g g−1; and (d) very high, >5 !g g−1. Accordingly, five of thesamples had only moderate PAH pollution, whereas the BZ sam-ple had high pollution and the BP sample had very high pollution.The ratio of different PAH isomers in a sample helps to iden-tify the sources of PAH contamination in environmental samples[10,19,25,51]. Fig. 3 shows the relative concentrations of the twoC14H10 isomers anthracene (AN) and phenanthrene (PH), and thetwo C16H10 isomers fluoranthene (FLU) and pyrene (PY) at the sam-pling sites. These suggested that the sources of PAHs in the samplesfrom Pula Bay were pyrogenic and those from Bakar Bay werepetrogenic.

M. Korlevic et al. / Systematic and Applied Microbiology 38 (2015) 189–197 193

Table 2Concentrations of aliphatic (n-alkanes), unresolved complex mixture (UCM), polycyclic aromatic hydrocarbons (PAH), unresolved to resolved aliphatic ratio (U:R) and carbonpreference index (CPI) in samples of northern Adriatic Sea coastal sediments.

Station n-Alkanes(!g g−1 d.w.)

UCM (!g g−1

d.w.)U:R CPI(15–35) PAH (!g g−1 d.w.)

I 5.89 1.81 0.31 1.04 0.12US 10.17 9.52 0.94 0.99 0.12B 4.02 13.97 3.48 1.09 0.26TV 6.59 10.75 1.63 0.98 0.40BI 38.31 167.48 4.37 1.41 0.48BZ 3.21 49.77 15.50 1.52 2.74BP 43.81 1469.20 33.53 1.19 6.79

Taxonomic classifications of bacterial communities

Taxonomic analysis of subsampled reads from eight samples(seven taken directly from surface sediments and one transferredto an Erlenmeyer flask and grown in aerobic conditions under crudeoil pressure) identified 2568 OTUs. The smallest number of OTUs(201) was found in the sample enriched and grown with crude oil(SF1) and the largest number of OTUs (926) was found in the mod-erately polluted sample from Uljanik repair shipyard (BZ) (Table 1).In total, 33 bacterial phyla were identified in this analysis from allsampled sediments. The most abundant phylum was Proteobacte-ria, with almost 54% of all reads being assigned to this phylum. Thesecond most abundant phylum was Firmicutes with 18% of all reads,followed by Bacteroidetes (7.2%). These groups were the only onescontaining more than 5% of total reads, and a total of 88.8% of allreads were assigned to them. Observed overall diversity was alsoreflected at a sample level with six samples being dominated by thephylum Proteobacteria (abundance 45–79%), one dominated by Fir-micutes (74%) and one where both phyla were equally represented(∼30%). At the class level, Gammaproteobacteria were most com-mon (29.9%), followed by Clostridia (17.9%) and Alphaproteobacteria(11.8%). At the lowest taxonomical level assigned to reads in this

analysis, half of the samples showed a relatively uniform distribu-tion of reads between many genera (136–186 genera). In contrast,the other half of the samples analysed had one or two generaaccounting for more than half of the reads: in the TV sample thesewere an unknown genus from the Pelagibacteraceae family (41.3%)and the genus Shewanella (14.2%), in the BP sample an unknowngenus from the family Helicobacteraceae (56%), in the BI sample anunknown genus from the family Lachnospiraceae (57.2%) and anunknown genus from the order Clostridiales (11.5%), and, in sampleSF1, the genus Marinobacter (56.3%) and an unknown genus fromthe family Rhodobacteraceae (16.3%). The percentage of unassignedreads ranged from 3.3% to 18.5%, with the highest being in the SF1sample grown under crude oil pressure.

In total, 18 bacterial orders were found in the SF1 sample grownunder crude oil pressure, and the most abundant order in thissample was Alteromonadales (56.8% of reads). This order was alsopresent in the original TV sample (14.8%) making it the second mostabundant order in this sample. The second most abundant order inSF1 was Rhodobacterales with 17.9% of the reads. These two ordersand the unassigned reads accounted for 94% of all reads in SF1,which showed a great reduction in order richness from 77 in theoriginal TV sample to 18 in SF1.

Fig. 3. The distribution of bacterial orders present in surface sediment samples.

194 M. Korlevic et al. / Systematic and Applied Microbiology 38 (2015) 189–197

Fig. 4. Rarefaction curves indicating: (a) observed number of operational taxonomicunits (OTUs) for all sediment samples, and (b) average observed number of opera-tional taxonomic units (OTUs) for clean and contaminated samples with error barsindicating the standard deviation.

Alpha diversity

From the analysis of alpha diversity metrics, the tested sam-ples could be split into three groups based on species richness anddiversity–high diversity (B, I and BZ samples), medium diversity(TV, US and BP samples) and low diversity (BI and SF1). Rarefactioncurves for observed species and Faith’s phylogenetic diversity met-rics indicated that subsampling depth was sufficient for the low andmedium diversity groups, while for the high diversity group it wastoo low. Based on the pollution level of the samples, a decreasein diversity could be observed in samples with higher levels ofpollution (Fig. 4).

The observed number of OTUs, Chao1, ACE, and Shannondiversity indices showed differences between the tested samples(Table 1). Three sample groups based on richness and diversity esti-mators could be observed: (i) high diversity (samples B, I and BZ),medium diversity (TV, US and BP samples), and low diversity (BIand SF1).

Beta diversity

Principal coordinate analysis (PCoA) of tested samples basedon weighted and unweighted UniFrac [31] distance metrics wasobtained. PCoA based on unweighted Unifrac metrics showed

−1.0 −0.5 0.0 0.5 1.0 1. 5−1.5

−1.0

−0.5

0.0

0.5

1.0

−1.0−0.5

0.0 0.5

1.0 1.5

NMDS1

NMDS

2

NMDS

3 BI TVUS

BP

BI

BZ

SF1

Fig. 5. Community analysis based on the NMDS of unweighted UniFrac distances ofsamples.

clustering of samples from unpolluted sediments taken from sitesB, TV, I and US located in the Bakar Bay area. The remaining samplesfrom Pula Bay (two polluted and one unpolluted) and the samplegrown under crude oil pressure showed a dispersed distribution(Fig. 5). PCoA based on weighted Unifrac metrics also showed a dis-persed pattern. Similar results were obtained using the non-metricmultidimensional scaling (NMDS) method, where the results basedon weighted Unifrac distances showed dispersal of all points inthree-dimensional space, while the results based on unweightedUnifrac distances showed clustering of samples from Bakar Bay anda scattered pattern for other samples (Fig. 5). NMDS analysis basedon Euclidean distance showed a similar pattern to unweightedUnifrac metrics but it grouped all samples from unpolluted sites(Fig. S2). The difference in clustering depending on the Unifracmetric used implied similar OTUs present in unpolluted samples,although in varying abundances, while the dispersal of pollutedsamples implied more diverged communities in these samples.

Phylogenetic analysis of alkB genes

In order to obtain a preliminary estimate of the potential of themicrobial populations to degrade alkanes, putative alkB genes werecloned from all DNA samples by PCR amplification. Clones wereobtained from all sampling sites. From the original 290 sequencedclones, 174 gave good hits to an AlkB HMM-profile and were usedfor phylogenetic analysis (Fig. 6). The majority of the PCR clonesequences were assigned to two large clusters in the phylogenetictree. The first cluster contained 87 sequences with representa-tives from each of the eight sampling sites; 56 of the sequenceswere derived from the TV, BP and BZ sediments/samples with allthe sequences from sample BP belonging to the first cluster. Thesecond cluster contained 56 sequences with representatives fromsix sampling sites. The sequences in the first cluster were verysimilar to each other (the average evolutionary divergence oversequence pairs within the cluster was 0.062), while the sequencesin the second cluster were more diverged (the average evolution-ary divergence over sequence pairs within the cluster was 0.323).The two major clusters contained similar sequences (the evolution-ary divergence over sequence pairs between clusters was 0.377).The sequences from both clusters, as well as the majority of PCRclones, were in a clade containing AlkB1 from Alcanivorax borku-mensis. Of the remaining sequences, ten from the BI sampling sitewere in a clade with AlkM from two Acinetobacter species. In con-trast, the majority of complete AlkB sequences extracted fromthe metagenomic sequence obtained from samples grown undercrude oil pressure showed considerable evolutionary divergencefrom the sequences of PCR clones and the representatives of theK00496 ortholog group included in the tree. In fact, 13 of the 19metagenome sequences clustered outside the clade containing thePCR clone sequences and representatives of K00496. This suggested

M. Korlevic et al. / Systematic and Applied Microbiology 38 (2015) 189–197 195

Fig. 6. Phylogenetic tree of 193 putative AlkB sequences together with seven knownAlkB sequences belonging to the KEGG orthologue K00496. The AlkB sequencesobtained in this study are labeled with the sites of collection. Clades of similarsequences have been collapsed..

that the primers used for PCR amplification did not amplify allgenes representing the entire diversity of AlkB enzymes in naturalhabitats, as already shown by Jurelevicius et al. [26].

Discussion

Samples were taken from seven sites in the northern AdriaticSea; six of the sites had been exposed to significant industrial pol-lution, whereas a reference site was in the tourist area of SvezanjCove. When n-alkanes were assayed, two sites had particularly highlevels: a marine gas station (BP) and a sample from a repair shipyard(BI) (Table 2). The levels in the reference site (US) were compa-rable with those at two industrial sites: a petroleum refinery (I)and a tanker berth (TV). Many of the sites gave high values forthe ratio of unresolved to resolved components (U:R) during gaschromatography. Five of the sites had U:R ratios greater than 2.5,whereas the sampling sites with low levels of alkanes (samplesfrom Bakar Bay) had lower ratios (Table 2). The sources of PAHs inPula Bay were pyrogenic and we suggest that heat could have influ-enced these compounds during the shipbuilding process. Biological

processes produce predominately n-alkanes with an odd number ofcarbon atoms, and CPI values are proportional to their odd length toeven length ratio. The calculated CPI values (Table 2) were relativelylow, indicating the presence of n-alkanes from petroleum productsand/or from partial thermal alteration (i.e. incomplete combustion)of petroleum or recent biological materials [33].

The levels of PAHs in the samples were also determined(Table 2). Five of the sites had moderate levels, whereas therepair shipyard site had high levels (BZ), and the marine gas sta-tion site (BP) had very high levels. These values were withinthe ranges previously found in the sediments from Rijeka Bay(0.053–12.532 !g g−1 d.w. [2]). The predominance of low andmedium molecular weight PAHs (with 2, 3 and 4 rings) in sed-iments from this region reflected the presence of significantcombustion products from low temperature pyrolytic processesand/or petrogenic sources [11,21]. The ratios of different PAH iso-mers has been used to identify different sources that contributeto PAHs in environmental samples [10,19,25,51]. The ratios ofanthracene (AN) were compared to phenanthrene (PH), and flu-oranthene (FLU) to pyrene (PY). Ratios of AN/(PH + AN) <0.1 andFLU/(FLU + PY) < 0.4 usually imply a petrogenic source, whereasratios of AN/(PH + AN) > 0.1 and FLU/(FLU + PY) > 0.5 suggest a pyro-genic source (grass, wood, or coal combustion). If the FLU/(FLU + PY)ratio is between 0.4 and 0.5, combustion of petroleum origin issuggested. In three samples (BP, BI, BZ), the principal source waspyrogenic, whereas two of the samples (B and TV) had a petroleumcombustion source (Table 3). The values at the other two sites (Iand US) suggested that possible sources of PAHs could be of mixedorigin.

In the past few years, different molecular approaches, mainlyinvolving 16S rRNA gene analyses, such as DGGE [3,4,14,24,30] or T-RFLP [18,23,36], clone libraries [30,35], phylo/geochip [6], and NGSand “omics” approaches [29], have been used to describe the micro-bial communities established in polluted coastal environments. Thebacterial populations in the North Adriatic sediment samples wereanalysed using 454 pyrosequencing of the 16S rRNA gene. Gamma-and Alphaproteobacteria mainly dominated bacterial communitiesin the North Adriatic sediment samples. The ability of Alpha- andGammaproteobacteria to utilize aliphatic and aromatic compoundshas previously been established [1,29,40]. Gammaproteobacteriahave been found to be one of the most abundant bacterial groupsin marine sediments [1,29]. Previous research identified microbialcommunities dominated by members of Gamma- and Alphapro-teobacteria as key players in oil-degradation, with members ofGammaproteobacteria contributing to the early stages of oil hydro-carbon degradation (oxidizing more reactive components such asn-alkanes), and members of Alphaproteobacteria contributing to thelater stages of degradation (oxidizing more recalcitrant compoundssuch as PAHs). In this research, Gamma- and Alphaproteobacte-ria dominated the sediment communities, with more than 40% ofall reads classified into respective classes [29]. The influence ofmassive crude oil contamination on the microbial population ofcoastal sediments was investigated in the Cies Islands 18 and 53months after the Prestige tanker sank off the NW coast of Spain [1],and it was shown that Gammaproteobacteria, Deltaproteobacteriaand Acidobacteria were dominant after 18 months. These results,showing the great plasticity of bacterial community structures andsuggesting that they were constantly adapting to changing envi-ronmental factors, highlight the importance of considering theimpact of environmental parameters on microorganisms whenstudying oil degradation in coastal marine sediments. Membersof the Deltaproteobacteria often play a key role in the anaerobicdegradation of organic matter (especially Desulfobacterales) andthey were identified in our study only in one sample from Pula Bay(BZ) with a high pyrotag content. It is interesting to note that thissample was the most polluted in terms of PAHs. Païssé et al. [36]

196 M. Korlevic et al. / Systematic and Applied Microbiology 38 (2015) 189–197

demonstrated in the Berre Lagoon (France) that bacterial diversitywas associated with the oil contamination gradient. They showedthat the adaptation of the bacterial community was not charac-terized by the dominance of oil-degrading bacteria, but instead apredominance of bacterial populations associated to the sulphurcycle was observed. Other in situ studies have highlighted the pres-ence or the role of sulphur cycle microorganisms in oil-pollutedcoastal marine sediments, with a focus on sulphate-reducing bac-teria [1,41]. Epsilonproteobacteria and Clostridia were detected withhigher pyrotag frequencies in two samples from Pula Bay (BP and BI,respectively). In these two samples (BP and BI), faecal bacterial indi-cators Lachnospiraceae and Helicobacteraceae were determined inalmost 60% of the pyrotags, respectively [34,46]. In other samples,Firmicutes were assigned to the Tepidibacter identified in pollutedsamples [52]. At the family level, the highest diversity could bedetected in samples BZ, B and I, which could have been exposedover a long period of time to a very diverse mixture of hydrocarboncompounds (i.e. to a diversity of potential carbon sources).

Enrichment culture has been used to tackle the impact of crudeoil and petroleum hydrocarbon compounds on bacterial communi-ties. The addition of a crude oil (SF1) resulted in a shift in bacterialcommunity structure. The dominant related pyrotag in the SF1sample (Marinobacter; Fig. 3) was not detected in the inoculat-ing sample (TV). Marinobacter is a genus of Proteobacteria foundin seawater, and it contains 31 species [20]. A number of Mari-nobacter species have already been described in several studiesas being present in areas contaminated with oil [7]. They maybe potential proxies for the biomonitoring of oil hydrocarbonsin ecosystems. Crude oil selection has resulted in domination bypreviously rare organisms. Other studies have linked the additionof different hydrocarbons with the functional response in micro-bial communities [1,6,29]. Following the dynamic modificationsof microbial communities in slurries, fingerprinting techniquesand 16S rRNA gene library analyses have shown the ecologicalsuccession of microbial populations in response to hydrocarboncompounds in polluted marine harbour sediments [12,39,49].

A preliminary assessment of the ability of samples to degradehydrocarbons was made by cloning putative alkB genes after PCRamplification. The enzyme AlkB was chosen because it is wellcharacterised and has been used in many studies [28]. PCR usingdegenerate primers often produces artefacts, so the sequences werescreened using an HMM-profile for AlkB. This resulted in retentionof 174/290 clones. Most of the rejected clones gave poor align-ment with the profile, and dropping the cut-off score from 200 to100 only resulted in 11 more clones being retained. Phylogeneticanalysis showed that over half the clones were in a single cladeof similar sequences present at all seven collecting sites (Fig. 4).The other sequences showed higher taxonomic diversity, suggest-ing that such samples might be good sources of novel enzymes.The use of only one alkB primer is not sufficient to detect all thealkB genes in cultivable and uncultivable bacteria, and for deepercharacterization and distribution of alkane-degrading bacteria indifferent environments it is necessary to use a set of primers or ametagenomic approach [26].

This study established a baseline for future studies of pollu-tion in the northern Adriatic Sea. Environmental pollution couldbecome a significant problem in this area due to possible futureoil and gas exploration, and more intensive discharge of agri-cultural, industrial and municipal wastewaters. Little is knownabout the function of even the most dominant microbial groupsin Adriatic Sea sediments, which limits our understanding of themicrobial metabolic potential. Bioremediation strategies need toemploy appropriate consortia of microorganisms and they need tobe monitored for both hydrocarbon concentrations and the com-position of the microorganism population. Microorganism samplesare also a promising source of novel enzymes that can be detected

by bioprospecting of metagenomes. It can be concluded that oilcontamination had a profound impact on the abundance and com-munity composition of indigenous bacteria in the North AdriaticSea, and the evidence of the current study points to members ofthe Gammaproteobacteria (Marinobacter) as possible key players inoil degradation.

Funding

This work was supported by grant 09/5 (to D.H.) from the Croa-tian Science Foundation, Republic of Croatia and by a cooperationgrant from the German Academic Exchange Service (DAAD) andthe Ministry of Science, Education and Sports, Republic of Croatia(to D.H. and J.C.). The study was also supported by King’s College,London (to P.F.L.).

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.syapm.2015.03.001.

References

[1] Acosta-González, A., Rosselló-Móra, R., Marqués, S. (2013) Characterization ofthe anaerobic microbial community in oil-polluted subtidal sediments: aro-matic biodegradation potential after the Prestige oil spill. Environ. Microbiol.15, 77–92, http://dx.doi.org/10.1111/j.1462-2920.2012.02782.x.

[2] Alebic-Juretic, A. (2011) Polycyclic aromatic hydrocarbons in marine sedimentsfrom the Rijeka Bay area, Northern Adriatic, Croatia, 1998–2006. Mar. Pollut.Bull. 62, 863–869, http://dx.doi.org/10.1016/j.marpolbul.2011.01.035.

[3] Al-Thukair, A.A., Abed, R.M.M., Mohamed, L. (2007) Microbial commu-nity of cyanobacteria mats in the intertidal zone of oil-polluted coastof Saudi Arabia. Mar. Pollut. Bull. 54, 173–179, http://dx.doi.org/10.1016/j.marpolbul.2006.08.043.

[4] Andrade, L.L., Leite, D.C.A., Ferreira, E.M., Ferreira, L.Q., Paula, G.R.,Maguire, M.J., Hubert, C.R., Peixoto, R.S., Dominigues, R.M., Rosado, A.S.(2012) Microbial diversity and anaerobic hydrocarbon degradation poten-tial in an oil-contaminated mangrove sediment. BMC Microbiol. 12, 186,http://dx.doi.org/10.1186/1471-2180-12-186.

[5] Baumard, P., Budzinski, H., Garrigues, P. (1998) Polycyclic aromatic hydrocar-bons in sediments and mussels of the western Mediterranean Sea. Environ.Toxicol. Chem. 17, 765–776, http://dx.doi.org/10.1016/j.jhazmat.2009.04.084.

[6] Beazley, M.J., Martinez, R.J., Rajan, S., Powell, J., Piceno, Y.M., Tom, L.M., Ander-sen, G.L., Hazen, T.C., Van Nostrand, J.D., Zhou, J., Mortazavi, B., Sobecky,P.A. (2012) Microbial community analysis of a coastal salt marsh affectedby the Deepwater Horizon oil spill. PLoS ONE 7, e41305, http://dx.doi.org/10.1371/journal.pone.0041305.

[7] Brito, E.M.S., Guyoneaud, R.M., Goni-Urriza, M., Ranchou-Peyruse, A., Ver-baere, A., Crapez, M.A.C., Wasserman, J.C.S.A., Duran, R. (2006) Characterizationof hydrocarbonoclastic bacterial communities from mangrove sedimentsin Guanabara Bay. Brazil. Res. Microbiol. 157, 752–762, http://dx.doi.org/10.1016/j.resmic.2006.03.005.

[8] Caporaso, J.G., Bittinger, K., Bushman, F.D., DeSantis, T.Z., Andersen, G.L., Knight,R. (2010) PyNAST: a flexible tool for aligning sequences to a template alignment.Bioinformatics 26, 266–267, http://dx.doi.org/10.1093/bioinformatics/btp636.

[9] Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello,E.K., Fierer, N., Gonzalez Pena, A., Goodrich, J.K., Gordon, J.I., Huttley, G.A., Kelley,S.T., Knights, D., Koenig, J.E., Ley, R.E., Lozupone, C.A., McDonald, D., Muegge,B.D., Pirrung, M., Reeder, J., Sevinsky, J.R., Turnbaugh, P.J., Walters, W.A., Wid-mann, J., Yatsunenko, T., Zaneveld, J., Knight, R. (2010) QIIME allows analysisof high-throughput community sequencing data. Nat. Methods 7, 335–336,http://dx.doi.org/10.1038/nmeth.f.303.

[10] Chen, C.W., Chen, C.F. (2011) Distribution, origin and potential toxicologi-cal significance of polycyclic aromatic hydrocarbons (PAHs) in sediments ofKaohsiung Harbor, Taiwan. Mar. Pollut. Bull. 63, 417–423, http://dx.doi.org/10.1016/j.marpolbul.2011.04.047.

[11] Chen, C.W., Chen, C.F., Dong, C.D., Tu, Y.T. (2012) Composition and sourceapportionment of PAHs in sediments at river mouths and channel in Kaoh-siung Harbor, Taiwan. J. Environ. Monit. 14, 105–115, http://dx.doi.org/10.1039/c1em10557d.

[12] Dell’Anno, A., Beolchini, F., Rocchetti, L., Luna, G.M., Danavaro, R. (2012) Highbacterial biodiversity increases degradation performance of hydrocarbons dur-ing bioremediation of contaminated harbor marine sediments. Environ. Pollut.167, 85–92, http://dx.doi.org/10.1016/j.envpol.2012.03.043.

[13] DeSantis, T.Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie, E.L., Keller, K., Huber,T., Dalevi, D., Hu, P., Andersen, G.L. (2006) Greengenes, a chimera-checked

M. Korlevic et al. / Systematic and Applied Microbiology 38 (2015) 189–197 197

16S rRNA gene database and workbench compatible with ARB. Appl. Environ.Microb. 72, 5069–5072, http://dx.doi.org/10.1128/AEM.03006-05.

[14] Dias, A.C.F., Dini-Andreote, F., Taketani, R.G., Tsai, S.M., Azevedo, J.L., De Melo,I.S., Andreote, F.M. (2011) Archaeal communities in the sediments of threecontrasting mangroves. J. Soils Sediments 11, 1466–1476, http://dx.doi.org/10.1007/s11368-011-0423-7.

[15] Dowd, S.E., Sun, Y., Wolcott, R.D., Domingo, A., Carroll, J.A. (2008) Bacterialtag-encoded FLX amplicon pyrosequencing (bTEFAP) for microbiome stud-ies: bacterial diversity in the ileum of newly weaned Salmonella-infected pigs.Foodborne Pathog. Dis. 5, 459–472, http://dx.doi.org/10.1089/fpd.2008.0107.

[16] Eddy, S.R. (2009) A new generation of homology search tools basedon probabilistic inference. Genome Inform. 23, 205–211, http://dx.doi.org/10.1142/9781848165632 0019.

[17] Edgar, R.C. (2010) Search and clustering orders of magnitude fasterthan BLAST. Bioinformatics 26, 2460–2461, http://dx.doi.org/10.1093/bioinformatics/btq461.

[18] Edlund, A., Jansson, J.K. (2006) Changes in active bacterial communities beforeand after dredging of highly polluted Baltic Sea sediments. Appl. Environ.Microbiol. 72, 6800–6807, http://dx.doi.org/10.1128/AEM. 00971-06.

[19] Fang, M.D., Lee, C.L., Yu, C.S. (2003) Distribution and source recognitionof polycyclic aromatic hydrocarbons in the sediments of Hsin-Ta Har-bour and adjacent coastal areas, Taiwan. Mar. Pollut. Bull. 46, 941–953,http://dx.doi.org/10.1016/S0025-326X(03)00099-7.

[20] Gao, W., Cui, Z., Li, Q., Xu, G., Jia, X., Zheng, L. (2013) Marinobacter nanhaiticussp. nov., polycyclic aromatic hydrocarbon-degrading bacterium isolated fromthe sediment of the South China Sea. Antonie van Leeuwenhoek 103, 485–491,http://dx.doi.org/10.1007/s10482-012-9830-z.

[21] Garcia-Falcon, M.S., Soto-Gonzalez, B., Simal-Gandara, J. (2006) Evolution of theconcentrations of polycyclic aromatic hydrocarbons in burnt woodland soils.Environ. Sci. Technol. 40, 759–763, http://dx.doi.org/10.1021/es051803v.

[22] Herzfelder, E.R. 2004 Method for the determination of extractable petroleumhydrocarbons. In: Massachusetts Department of Environmental ProtectionMADEP-EPH-04, Massachusetts Department of Environmental Protection,⟨http://www.mass.gov/dep/cleanup/laws/eph0504.pdf⟩.

[23] Iannelli, R., Bianchi, V., Macci, C., Peruzzi, E., Chiellini, C., Petroni, G., Mas-ciandaro, G. (2012) Assessment of pollution impact on biological activity andstructure of seabed bacterial communities in the Port of Livorno (Italy). Sci.Total Environ. 426, 56–64, http://dx.doi.org/10.1016/j.scitotenv.2012.03.033.

[24] Isaac, P., Sanchez, L.A., Bourguignon, N., Cabral, M.E., Ferrero, M.A.(2013) Indigenous PAH-degrading bacteria from oil-polluted sediments inCaleta Cordova, Patagonia Argentina. Int. Biodeter. Biodegr. 82, 207–214,http://dx.doi.org/10.1016/j.ibiod.2013.03.009.

[25] Jiang, J.J., Lee, C.L., Fang, M.D., Liu, J.T. (2009) Polycyclic aromatic hydrocarbonsin coastal sediments of southwest Taiwan: an appraisal of diagnostic ratios insource recognition. Mar. Pollut. Bull. 58, 752–760, http://dx.doi.org/10.1016/j.marpolbul.2008.12.017.

[26] Jurelevicius, D., Alvarez, V.M., Peixoto, R., Rosado, A.S., Seldin, L. (2013) Theuse of a combination of alkB primers to better characterize the distribution ofalkane-degrading bacteria. PLoS ONE 8 (6), e66565, http://dx.doi.org/10.1371/journal.pone.0066565.

[27] Kanehisa, M., Goto, S., Sato, Y., Furumichi, M., Tanabe, M. (2012) KEGG for inte-gration and interpretation of large-scale molecular data sets. Nucleic Acids Res.40 (Database issue), D109–D114, http://dx.doi.org/10.1093/nar/gkr988.

[28] Kloos, K., Munch, J.C., Schloter, M. (2006) A new method for the detec-tion of alkane-monooxygenase homologous genes (alkB) in soils basedon PCR-hybridization. J. Microbiol. Methods 66, 486–496, http://dx.doi.org/10.1016/j.mimet.2006.01.014.

[29] Kostka, J.E., Prakash, O., Overholt, W.A., Green, S.J., Freyer, G., Canion, A., Del-gardio, J., Norton, N., Hazen, T.C., Huettel, M. (2011) Hydrocarbon-degradingbacteria and the bacterial community response in Gulf of Mexico beach sandsimpacted by the Deepwater Horizon oil spill. Appl. Environ. Microbiol. 77,7962–7974, http://dx.doi.org/10.1128/AEM.05402-11.

[30] LaMontagne, M.G., Leifer, I., Bergmann, S., Van De Werfhorst, L.C., Holden, P.A.(2004) Bacterial diversity in marine hydrocarbon seep sediments. Environ.Microbiol. 6, 799–808, http://dx.doi.org/10.1111/j.1462-2920.2004.00613.x.

[31] Lozupone, C., Knight, R. (2005) UniFrac: a new phylogenetic method forcomparing microbial communities. Appl. Environ. Microbiol. 71, 8228–8235,http://dx.doi.org/10.1128/AEM.71.12.8228-8235.2005.

[32] Mazurek, M.A., Cass, G.R., Simoneit, B.R.T. (1989) Interpretation of high-resolution gas chromatography and high-resolution gas chromatography/massspectrometry data acquired from atmospheric organic aerosol samples. AerosolSci. Technol. 10, 408–420, http://dx.doi.org/10.1080/02786828908959280.

[33] Mazurek, M.A., Simoneit, B.R.T. (1984) Characterization of biogenic andpetroleum-derived organic matter in aerosols over remote, rural andurban areas. In: Keith, L.H. (Ed.), Identification and Analysis of OrganicPollutants in Air, Butterworth Publishers, Ann Arbor, MI, London,pp. 353–370.

[34] Meehan, C.J., Beiko, R.G. (2014) A phylogenomic view of ecological special-ization in the Lachnospiraceae, a family of digestive tract-associated bacteria.Genome Biol. Evol. 6, 703–713, http://dx.doi.org/10.1093/gbe/evu050.

[35] Miralles, G., Nérini, D., Manté, C., Acquaviva, M., Doumenq, P., Michotey,V., Nazaret, S., Bertrand, J.C., Cuny, P. (2007) Effects of spilled oil on bac-terial communities of Mediterranean coastal anoxic sediments chronicallysubjected to oil hydrocarbon contamination. Microb. Ecol. 54, 646–661,http://dx.doi.org/10.1007/s00248-007-9221-6.

[36] Paissé, S., Goni-Urriza, M., Stadler, T., Budzinski, H., Duran, R. (2012) Ring-hydroxylating dioxygenase (RHD) expression in a microbial communityduring the early response to oil pollution. FEMS Microbiol. Ecol. 80, 77–86,http://dx.doi.org/10.1111/j.1574-6941.2011.01270.x.

[37] Quince, C., Lanzén, A., Curtis, T.P., Davenport, R.J., Hall, N., Head, I.M.,Read, L.F., Sloan, W.T. (2009) Accurate determination of microbial diversityfrom 454 pyrosequencing data. Nat. Methods 6, 639–641, http://dx.doi.org/10.1038/nmeth.1361.

[38] Quince, C., Lanzen, A., Davenport, R.J., Turnbaugh, P.J. (2011) Removing noisefrom pyrosequenced amplicons. BMC Bioinform. 12, 38, http://dx.doi.org/10.1186/1471-2105-12-38.

[39] Rocchetti, L., Beolchini, F., Hallberg, K.B., Johnson, D.B., Dell’Anno, A. (2012)Effects of prokaryotic diversity changes on hydrocarbon degradation ratesand metal partitioning during bioremediation of contaminated anoxicmarine sediments. Mar. Pollut. Bull. 64, 1688–1698, http://dx.doi.org/10.1016/j.marpolbul.2012.05.038.

[40] Rojo, F. (2009) Degradation of alkanes by bacteria. Environ. Microbiol. 11,2477–2490, http://dx.doi.org/10.1111/j.1462-2920.2009.01948.x.

[41] Rosano-Hernandez, M.C., Ramirez-Saad, H., Fernandez-Linares, L. (2012)Petroleum-influenced beach sediments of the Campeche Bank, Mexico: diver-sity and bacterial community structure assessment. J. Environ. Manage. 95,S325–S331, http://dx.doi.org/10.1016/j.jenvman.2011.06.046.

[42] Söhngen, N.L. (1913) Benzin, Petroleum, Paraffinöl und Paraffin als Kohlenstoff-und Energiequelle für Mikroben. Zbl. Bacteriol. Parasitek Abt. II. 37,595–609.

[43] Tamura, K., Stecher, G., Peterson, D., Filipski, A., Kumar, S. (2013) MEGA6:molecular evolutionary genetics analysis version 6.0. Mol. Biol. Evol. 30,2725–2729, http://dx.doi.org/10.1093/molbev/mst197.

[44] Traven, L. (2013) Sources, trends and ecotoxicological risks of PAH pollution insurface sediments from the northern Adriatic Sea (Croatia). Mar. Pollut. Bull.77, 445–450, http://dx.doi.org/10.1016/j.marpolbul.2013.08.043.

[45] Vilchez-Vargas, R., Junca, H., Pieper, D.H. (2010) Metabolic networks, microbialecology and ‘omics’ technologies: towards understanding in situ biodegrada-tion processes. Environ. Microbiol. 12, 3089–3104, http://dx.doi.org/10.1111/j.1462-2920.2010.02340.x.

[46] Vitousek, P.M., Mooney, H.A., Lubchenco, J., Melillo, J.M. (1997) Human domi-nation of Earth’s ecosystems. Science 277, 494–499, http://dx.doi.org/10.1126/science.277.5325.494.

[47] Wang, L., Wang, W., Lai, Q., Shao, Z. (2010) Gene diversity of CYP153Aand AlkB alkane hydroxylases in oil-degrading bacteria isolated from theAtlantic Ocean. Environ. Microbiol. 12, 1230–1242, http://dx.doi.org/10.1111/j.1462-2920.2010.02165.x.

[48] Wang, W., Shao, Z. (2013) Enzymes and genes involved in aerobic alkane degra-dation. Front. Microbiol. 4, 116, http://dx.doi.org/10.3389/fmicb.2013.00116.

[49] Wang, Y.F., Tam, N.F.Y. (2011) Microbial community dynamics and biodegra-dation of polycyclic aromatic hydrocarbons in polluted marine sedimentsin Hong Kong. Mar. Pollut. Bull. 63, 424–430, http://dx.doi.org/10.1016/j.marpolbul.2011.04.046.

[50] Yakimov, M.M., Timmis, K.N., Golyshin, P.N. (2007) Obligate oil-degradingmarine bacteria. Curr. Opin. Biotechnol. 18, 257–266, http://dx.doi.org/10.1016/j.copbio.2007.04.006.

[51] Yunker, M.B., Macdonald, R.W., Vingarzan, R., Mitchell, R.H., Goyette, D.,Sylvestre, S. (2002) PAHs in the Fraser River Basin: a critical appraisal of PAHratios as indicators of PAH source and composition. Org. Geochem. 33, 489–515,http://dx.doi.org/10.1016/S0146-6380(02)00002-5.

[52] Zhao, J.S., Manno, D., Hawari, J. (2007) Abundance and diversity ofoctahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX)-metabolizing bacteriain UXO-contaminated marine sediments. FEMS Microbiol. Ecol. 59, 706–717,http://dx.doi.org/10.1111/j.1574-6941.2006.00248.x.

Further reading

www.mrdnalab.com, last accessed August 27th, 2014.http://www.mass.gov/dep/cleanup/laws/eph0504.pdf, last

accessed August 27th, 2014.