Environmental Toxicology and Chemistry, Vol. 29, No. 6, pp. 1358–1366, 2010# 2010 SETAC

Printed in the USADOI: 10.1002/etc.156

INTEGRATED ASSESSMENT OF OIL POLLUTION USING BIOLOGICAL MONITORING

AND CHEMICAL FINGERPRINTING

CERI LEWIS,*y CARLOS GUITART,z CHRIS POOK,y ALAN SCARLETT,§ JAMES W. READMAN,z and TAMARA S. GALLOWAYyyHatherly Laboratories, School of Biosciences, University of Exeter, Exeter EX4 4PS, United Kingdom

zPlymouth Marine Laboratory, Plymouth, Devon PL1 3DH, United Kingdom

§School of Geography, Earth and Environmental Sciences, University of Plymouth, Plymouth, Devon PL4 8AA, United Kingdom

(Submitted 14 May 2009; Returned for Revision 21 December 2009; Accepted 11 January 2010)

* T(c.n.lew

Pub(www.

Abstract—A full assessment of the impact of oil and chemical spills at sea requires the identification of both the polluting chemicals and thebiological effects they cause. Here, a combination of chemical fingerprinting of surface oils, tissue residue analysis, and biological effectsmeasures was used to explore the relationship between spilled oil and biological impact following the grounding of the MSC Napolicontainer ship in Lyme Bay, England in January 2007. Initially, oil contamination remained restricted to a surface slick in the vicinity of thewreck, and there was no chemical evidence to link biological impairment of animals (the common limpet, Patella vulgata) on the shoreadjacent to the oil spill. Secondary oil contamination associated with salvage activities in July 2007 was also assessed. Chemical analyses ofaliphatic hydrocarbons and terpanes in shell swabs taken from limpet shells provided an unequivocal match with the fuel oil carried by theship. Corresponding chemical analysis of limpet tissues revealed increased concentrations of polycyclic aromatic hydrocarbons (PAHs)dominated by phenanthrene and C1 to C3 phenanthrenes with smaller contributions from heavier molecular weight PAHs. Concurrentecotoxicological tests indicated impairment of cellular viability (p< 0.001), reduced immune function (p< 0.001), and damage to DNA(Comet assay, p< 0.001) in these animals, whereas antioxidant defenses were elevated relative to un-oiled animals. These results illustratethe value of combining biological monitoring with chemical fingerprinting for the rapid identification of spilled oils and their sublethalimpacts on biota in situ. Environ. Toxicol. Chem. 2010;29:1358–1366. # 2010 SETAC

Keywords—Oil spill Biomarker Chemical fingerprinting Polyaromatic hydrocarbon Limpet

INTRODUCTION

It is generally recognized that recent environmental manage-ment practices have led to an improvement in the quality of theaquatic environments where they are implemented. While thishas been largely due to improved methods of treating andmonitoring wastewaters and intentional pollution discharges,a persistent danger remains to both marine and freshwaterecosystems from accidental pollution events [1], often causedby oil-related activities. Pollution nature, magnitude, and siteof occurrence can all be very different, with an unpredictableoutcome on the responses of individual organisms, andthe biodiversity and functioning of the exposed aquatic eco-system [2].

Essential activities in the risk assessment of oil and hazard-ous chemical spills include the rapid determination of themagnitude of the incident, the protection of the (aquatic)ecosystem from further harm, and the identification and mon-itoring of remediation measures to restore the impacted eco-system, as much as possible, to its original condition [3–5].Central to most ecological risk assessment frameworks, includ-ing those adapted specifically to the assessment of acute oilspills (e.g., Natural Resource Damage Assessment, MarineResource Damage Assessment), is the need to determine thenature and extent of exposure of biota to contamination and toevaluate the sublethal effects of biological consequence [6,7].At present, the main decision criteria are generally based onwater and sediment chemistry, and tissue residue analyses, todetermine the bioavailability and bioaccumulation of contam-

o whom correspondence may be addressedsi@exeter.ac.uk).

lished online 18 February 2010 in Wiley InterScienceinterscience.wiley.com).

1358

inants, combined with monitoring of community composition toquantify changes in the abundance and diversity of species [8].In addition, biomarkers may be used which relate quantifiablemolecular, cellular, or physiological changes in an exposure ortime-dependent manner to the degree of dysfunction that thecontaminant has produced in organisms [9–11]. Some degree ofanthropogenic contamination is generally present throughoutmost marine and coastal habitats, however, making the differ-entiation between chronic pollutants and those resulting fromaccidental spills highly problematic. The application of bio-markers in conjunction with chemical analyses to the assess-ment of acute oil spill situations is of particular interest, becausethey can, in principle, be used to identify which pollutants areresponsible for environmental degradation, thus enabling differ-entiation between chronic and acute contamination, to identifysites at risk and to track the progress of remedial action [12–14].

The Mediterranean Shipping Company (MSC) Napoli con-tainer ship, gross tonnage 53 K, length 270 m, ran into diffi-culties in the English Channel 80 km southwest of the Lizard inCornwall, SW England, in January 2007, and was eventuallybeached approximately 1 km offshore of Branscombe Beach,Devon (50840.62N, 3809.89W), an area which is part of a WorldHeritage site of exceptional geological interest. The ship wascarrying 41,473 tonnes of cargo including industrial, agricul-tural, and personal care chemicals. Following beaching of thevessel in Lyme Bay, spillage of fuel oil (estimated at 200tonnes) occurred and approximately 100 containers were lostoverboard, provoking fears of a major environmental event; aseries of coordinated protection measures were implemented bythe United Kingdom government [15]. Results from a detailedchemical contamination survey conducted after the beaching ofthe stricken cargo ship have been reported [16]. The dataobtained from the survey indicated localized oil contamination

Fig. 1. Study sites showing location of the MSC Napoli container ship withinLyme Bay after beaching in January 2007, the sampling site at BranscombeBeach, England, and the reference sites used for comparisons of limpetbiomarkers’ responses at Bantham and Port Quin. [Color figure can be seen inthe online version of this article, available at www.interscience.wiley.com.]

Integrated impact assessment of spilled oils on local biota Environ. Toxicol. Chem. 29, 2010 1359

from the MSC Napoli, primarily in the form of a surface slickwith contamination extending throughout an area of approx-imately 15-km radius from the wreck. Further away from thewreck, polycyclic aromatic hydrocarbon (PAH) concentrationsthroughout Lyme Bay were generally typical of those associ-ated with marine environments, with low concentrations ofpyrolytically derived PAHs. Initially, there was additionalconcern that containers carrying pesticides had been lost over-board with the potential for pesticide contamination of thesurrounding waters. Subsequent screening by programmabletemperature vaporization (PTV) gas chromatography (GC) withmass spectrometric (MS) detection (PTV-GC/MS) did notreveal the presence of any other chemical contaminants fromthe MSC Napoli cargo inventory, suggesting none were actuallyspilled.

The aim of the present study was to explore the relationshipbetween the nature and extent of acute chemical exposure ofbiota and the sublethal biological effects of that exposure for akey rocky shore species, using a real spill incident as a casestudy. The present study was conducted in parallel with adetailed chemical evaluation of waterborne contamination[16], enabling the relationship between contamination fromthe MSC Napoli and sublethal biological effects in the commonlimpet Patella vulgata, to be investigated. Limpets are proso-branch mollusks that inhabit rocky shores and are the dominantfauna around Lyme Bay. They are often chosen for ecotoxico-logical study because they are grazers which occupy an impor-tant position in maintaining the community structure of rockyshores, and are highly vulnerable to the toxicological effects ofpetroleum-derived hydrocarbons [17,18]. The biological effectstesting regime selected in the present study included biomarkersof cellular function (cell viability, phagocytic index), oxidativestress (total antioxidant status), and genetic damage (Cometassay). A biomarker of organophosphorous pesticide exposure(acetylcholinesterase activity) was also initially includedbecause, at the time of analysis, it was feared some of thepesticide cargo of the ship may have been spilled. Shorelinesampling sites were chosen to investigate the relationshipbetween hydrocarbon exposure and biological effects, to eval-uate environmental impact both immediately after the beachingof the ship (January 2007), and during the subsequent wrecksalvage operations to split the hull for removal and subsequentrecycling (July 2007).

MATERIALS AND METHODS

Sample collection and preparation

Marine organisms (the common limpet Patella vulgata)were collected from the shoreline at three sites in BranscombeBeach in Lyme Bay (Fig. 1) immediately following the beach-ing of the MSC Napoli in January 2007. Reference samples werealso collected from Port Quin, North Cornwall and Bantham,South Devon (sites relatively free from anthropogenic contam-ination according to Environment Agency 2008 data). Thewater temperature was 12� 18C and salinity was between 32and 35% at the time of collection. Individual limpets, of shellwidth 2.5 to 3.5 cm, were gently prized from the rocks avoidingdamage to the pedal foot (n¼ 20 at each site), and wereimmediately transported to the laboratory and maintained at158C in aerated water collected from their site of origin prior toanalyses within 24 h of collection. The second collection wasfrom Branscombe Bay on July 18, 2007 as the salvage operationwas under way, causing a secondary release of oil. To performthe chemical tests, organisms were cleaned manually and

visible residues of oil on the shells were scraped into microvialsusing a precleaned stainless steel spatula. For the toxicologicalbiomarker analyses, hemolymph samples were collected usinga 21-g hypodermic needle inserted into the pedal foot of thelimpet, and collected into an equal volume of physiologicalsaline (0.02M 4-[2-hydroxyethyl]-1-piperazineethanesulfonicacid [HEPES] buffer pH 7.4, 0.4M NaCl2, 0.1M MgSO4,0.01M KCl and 0.01M CaCl2). Animals were then weighedand frozen immediately at �808C prior to further analyses.

Chemical fingerprinting analyses

Full details of the analyses of water and fuel samples aredescribed in Guitart et al. [16]. In summary, following collec-tion, water samples were preserved with the addition ofdichloromethane. Unfiltered samples were spiked with deuter-ated internal standards and were liquid–liquid extracted, threetimes with dichloromethane. Extracts were treated with Na2SO4

to eliminate residual traces of water, rotary evaporated to500ml. Aliquots (20ml) were then analyzed by PTV-GC/MSin simultaneous full scan and selected ion monitoring modes. Asample of the original heavy oil carried aboard the MSC Napoliwas supplied by the UK Maritime and Coastguard Agency toenable direct forensic comparisons with water samples.

Limpet tissue analysis

The extraction of hydrocarbons from limpet tissues wasby alkaline saponification [19,20]. In brief, phenanthrene d10

(internal standard) was added to thawed limpet tissues (20 g wetwt) and refluxed for 2 h with methanol and potassium hydroxide(20 g), filtered, then solvent exchanged into hexane. Followingreduction in volume and clean-up on 5% deactivated alumina,the extracts were analyzed by GC/MS. Dry weights wereobtained from subsamples of wet tissue following drying at608C for 24 h.

The tissue extracts were subjected to GC/MS analyses usingan Agilent 6890 gas chromatograph interfaced with an Agilent5973N mass spectrometer in simultaneous full scan and selectedion monitoring modes to provide a general screen for contam-inants and to quantify PAHs, respectively. Calibration solutions

1360 Environ. Toxicol. Chem. 29, 2010 C. Lewis et al.

and analytical blanks were run concurrently with the tissueextract samples. For identification and quantification, extractswere compared with a standard quick turnaround methodPAH mix (catalog no. 47930-U; Supelco) containing 16 PAHsincluding naphthalene, acenaphthylene, acenaphthene, fluorene,phenanthrene, anthracene, fluoranthene, pyrene, benzo[a]-anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoran-thene, benzo[a]pyrene, dibenz[a,h]anthracene, indeno[1,2,3-cd]pyrene, and benzo[ghi]perylene. All solvents used were ofglass-distilled grade (Rathburns). Residues of oil removedfrom the mollusk shells were diluted with dichloromethaneand, following filtration through glass wool to eliminate anyinsoluble residues, extracts were further diluted and analyzed byPTV-GC/MS.

Biomarker analyses

Lysosomal stability. The membrane stability of limpethemocytes, known to be strongly correlated with parametersof organism fitness such as growth and fecundity [21], wasdetermined by measuring the ability of the hemocyte lysosomesto retain neutral red dye [13,22]. Samples of hemolymph (50ml)from 20 individuals per site were incubated in triplicate in flat-bottomed microtiter plates to allow a monolayer of cells toadhere to the wells. After 45 min, nonadhered cells were dis-carded and the plates washed with physiological saline. Asolution of 0.4% neutral red dye in physiological saline(200ml) was added and, after 3 h, the wells were washed andan acidified solution of 1% acetic acid, 20% ethanol added toresolubilize the dye. The plate was gently shaken for 10 minbefore reading the absorbance at 540 nm.

Genotoxicity. The Comet assay was used as a measure ofgenotoxic response of limpets to any oil exposure to determinethe level of DNA damage (measured as DNA strand breaks) inlimpet hemocytes. The Comet assay is widely used in ecotox-icological monitoring as a measure of genotoxic damage, andhas been linked to reproductive success in some marine inver-tebrate species [23–25]. All samples were first checked forcell viability using Eosin Y staining. Hemolymph, 50ml, wascentrifuged at 1,000 rpm for 4 min, and the cell concentrategently mixed with 1% low-melting point agarose (378C) anddropped onto slides previously coated with 1% normal-meltingpoint agarose. The slides were allowed to set for 10 min at 48C,and the Comet assay performed with modifications [10,23],using alkaline conditions at 58C. Briefly, 1-h lysis was followedby 45-min denaturation in electrophoresis buffer (0.3M NaOHand 1 mM ethylenediaminetetraacetic acid (EDTA), pH13) andthen electrophoresis for 30 min at 25 V and 300 mA, followedby neutralization. Cells were stained with 20 mg L�1 ethidiumbromide and examined using a fluorescent microscope withexcitation at 420 to 490 nm and emission at 520 nm. The percentof DNA in the Comet tail was quantified in 100 cells perpreparation using Kinetic COMET Software. Samples from20 individuals for each collection point were analyzed, withthe exception of the January Branscombe samples, where anumber of slides were damaged during the assay, leaving lowerreplication for this sampling point (n¼ 5).

Immune function. Any effect of oil exposure on the immunefunction of exposed limpets was determined using the phag-ocytosis assay [12,13]. Phagocytosis is the main mechanism ofimmune cellular defense in invertebrates and is known to besensitive to potential immunotoxicants [26]. The phagocytosisactivity of limpet hemocytes (n¼ 20) was determined by meas-uring the uptake of zymosan particles (from Saccharomycescerevisae) dyed with neutral red dye [27,28]. Particle uptake

by cell monolayers was estimated in microtiter plates byabsorbance at 540 nm against a standard curve prepared usingzymosan particles in the range 1.56 to 100� 107 ml�1. Theprotein concentration of each hemolymph sample from thephagocytosis assay was obtained using a commercial bicincho-ninic acid protein assay.

Acetylcholinesterase activity. The Ellman method was used,as adapted by Rickwood and Galloway [28], to assess possiblepesticide exposure by the inhibition of esterase activities inlimpet hemolymph from animals collected during the Januarysampling period. Samples of hemolymph (n¼ 20) or bufferblanks (50ml) were incubated for 5 min in microtiter plates at258C with 150ml of 5, 50-dithiobis-(2-nitrobenzoic acid),270mM in 50 mM sodium phosphate, pH 7.4. Measurementof enzyme activity was initiated by the addition of acetylth-iocholine iodide (3 mM) and the absorbance recorded at412 nm. Spontaneous substrate hydrolysis was determined inthe absence of hemolymph. The results were calculated as nmolsubstrate hydrolyzed min�1 mg�1 relative to the protein in thesample. The protein concentration of each hemolymph samplefrom the phagocytosis assay was obtained using a commercialbicinchoninic acid protein assay.

Statistical analysis

All statistical analyses were carried out in using the stat-istical software Minitab 15. Bartlett’s and Levene’s tests wereused to confirm homogeneity of variance between samples, thenone-way analysis of variances (ANOVAs) performed to com-pare samples. Post hoc identification of significant differencesidentified by ANOVA was conducted by Fisher’s least signifi-cant difference. Differences at the p< 0.05 level were consid-ered significant.

RESULTS

Chemical analysis of water

From the initial water column screening survey of the LymeBay area, concentrations of total PAHs in the surface microlayerwere as high as 57,000 ng/L in the immediate vicinity of theshipwreck (and corresponded to a distinct visible surfacesheen). Concentrations were shown to decline rapidly withdistance from the ship. In contrast, while total PAH concentra-tions in subsurface water samples were slightly elevated close tothe MSC Napoli (and in the area of Portland), throughout mostof the bay, concentrations were <6 ng/L, i.e., typical PAHconcentrations for marine coastal environments [15,16]. Quan-tification of PAHs in water samples from the three study sites issummarized in Table 1.

Chemical fingerprinting analyses

A selection of specific target analytes, including individualaliphatic hydrocarbons from C9 to C36 and isoprenoid hydro-carbons (pristane and phytane), the U.S. Environmental Pro-tection Agency’s priority parent PAHs, the petroleum-specificalkylated homologues of PAHs (i.e., C1–C2 naphthalenes,dibenzothiophenes, and phenanthrenes) were chosen to char-acterize the samples. In addition, other chemical biomarkers(i.e., terpanes and steranes) were investigated to fingerprint thesource(s).

Shell swabs taken from the three shoreline sites in January2007 contained undetectable concentrations of total PAHs(results not shown) even though, as mentioned below, signifi-cant differences were evident in biomarker responses in thelimpets at Branscombe Beach (immediately in front of the

Table 1. Seawater polycyclic aromatic hydrocarbon (PAH) measurements for the sampling sites and the surrounding areas of Lyme Bay, Englanda

Sampleb Branscombe Beach Lyme Regis Chesil Cove Lyme Bay Bantham Port Quin

Surface microlayerTotal PAH ng/L 57,071 50.7 40.5 6.8 NA NAFL/Py ratio 0.279 NA NA NA NA NA

Subsurface waterTotal PAH 30.8 6.6 14.6 1.8 NA NAFL/Py ratio 0.857 2.0 2.0 1.6 NA NA

a NA¼ not available.b Guitart et al. [16]; Hagger et al. [10].

Integrated impact assessment of spilled oils on local biota Environ. Toxicol. Chem. 29, 2010 1361

Napoli’s grounding site). Following salvage operations inJuly 2007, GC/MS chemical fingerprint comparisons revealedidentical profiles for components of oil on the shells and in thefuel oil carried by the MSC Napoli. In Table 2, the ratio of nC17/pristine and nC18/phytane is given to confirm similaritybetween the beach sample extracts and the fuel oil. Comparisonof the aliphatic fractions (m/z¼ 85, Fig. 2), shows considerableloss of more volatile n-alkanes by weathering in the <24-h timeperiod between spillage and collection of samples. The com-position of terpanes (m/z¼ 191, Fig. 3), however, reveal iden-tical profiles for ship oil, water samples, and limpet shell swabs,therefore confirming the identity and similarity of source of theshoreline oil samples and the Napoli fuel oil.

Chemical analysis of limpet tissue

Limpets from prespill Branscombe were found to containsimilar concentrations of PAHs as limpets from the referencesite at Chesil Beach, approximately 40 km further west inthe English Channel (Fig. 4), with total PAHs 0.2mg/g or lessat both sites. Following the second spill during the salvageoperation, limpets from Branscombe were found to containincreased concentrations of PAHs. The tissue burdens weredominated by phenanthrene and C1–C3 phenanthrenes withsmaller contributions from heavier molecular weight PAHs(Fig. 4).

Biomarkers

At the time of the first sample collections immediatelyfollowing the shipwreck in January 2007, significant differenceswere evident in biomarker responses in the limpets fromBranscombe Beach compared to the two reference sites(Fig. 5). The lysosomal stability (a measure of cellular health)of hemocytes was significantly reduced in limpets from Bran-scombe (mean¼ 5.60� 0.50) compared to those from the tworeference sites (means¼ 9.81� 1.01 and 10.61� 1.44, one-wayANOVA, F¼ 6.58, p¼ 0.003). Immune function, measuredas phagocytic index, was higher in limpets collected fromBranscombe (mean¼ 6.38� 10�8) compared to those fromthe reference sites (means¼ 7.67� 10�7 and 5.17¼ 10�7,one-way ANOVA, F¼ 3.83, p¼ 0.028). Limpets from Bran-

Table 2. Comparison of nC17/pristine and nC18/phytane ratiosbetween samplesa

Limpets Periwinkles Seawater Napoli engine oil

C17/pristane 5.2 3.5 5.8 4.7Mean 4.2 3.8 5.3SD 0.9 0.3 0.7C18/phytane 2.4 2.4 2.2 3.1Mean 2.6 2.5 2.0SD 0.2 0.3 0.3

a SD¼ standard deviation.

scombe were also found to have significantly higher levels ofDNA damage (two-sample t test, T¼�5.11, p< 0.001, df¼ 21)than Port Quin limpets (no data for Bantham). Althoughthere was a small significant difference between sites in theacetylcholinesterase activity measured (ANOVA, F¼ 3.29,p¼ 0.027), this was due to high activity in the Port Quinanimals. Acetylcholinesterase activity in the Branscombelimpets fell within the range of the two reference sites, suggest-ing no extensive release of pesticides from the ship’scontainers (later confirmed by analytical chemistry, data notshown).

Following salvage activities to remove the vessel’s hull andthe subsequent visual oiling of the shore at Branscombe Beachin July 2007, a highly significant further decline in cellularviability (mean neutral red retention¼ 2.73� 0.38, one-wayANOVA, F¼ 18.09, p< 0.001) was evident compared withthe January analyses. This was accompanied by a significantincrease in DNA damage (i.e., in the percentage of DNA inthe Comet tail) compared with the reference sites and earlierBranscombe measurements (one-way ANOVA, F¼ 54.82, p<0.0001) (Fig. 5). Phagocytic index was significantly depressedcompared to limpets from the two reference sites andthe January Branscombe measurements (one-way ANOVA,F¼ 7.93, p< 0.001).

DISCUSSION

The present study illustrates how a combination of chemicalfingerprinting with biomarkers of sublethal effect can be used toprovide a rapid identification of which pollutants are present atsites of environmental contamination, providing an extremelyuseful means of assessing acute oil spill situations. Biomarkerresponses when taken in isolation can be misleading whereunderlying contamination is present, however, when combinedwith detailed analytical chemistry and measured at more thanone time point, they can provide unique information regardingthe sublethal impact of a spill incident on the local biota. Thepresent study emphasizes the fact that the release of a chemicalor of spilled oil by itself does not necessarily equate to an effecton a biological receptor. Likewise, a detrimental biologicalresponse measured in the vicinity of an acute pollution sourcedoes not by itself prove that source to be the cause of thebiological effects [3]. Our combined analysis of chemicalfingerprinting, tissue concentrations, and biological markersreveal that the initial minor spill resulting from the beachingof the MSC Napoli had no impact on the limpets of the adjacentcoastline, despite their poor health status, but that the secondaryspill resulting from the salvage operation did result in PAHexposures and sublethal impacts on these animals.

We initially found strong statistical evidence of an alterationin the biomarker profile of limpets collected from the areadown-tide of the wreck site immediately after the beaching of

Fig. 2. Gas chromatography–mass spectrometry ion chromatogram profile (m/z¼ 85) for aliphatic hydrocarbons of showing the aliphatic fraction profile of (A)Napoli fuel oil; (B) limpet shell swabs, and (C) seawater from the sea-surface microlayer sampled near the shipwreck.

1362 Environ. Toxicol. Chem. 29, 2010 C. Lewis et al.

Fig. 3. Gas chromatography–mass spectrometry ion chromatogram profile (m/z¼ 191) for aliphatic hydrocarbons of showing the terpanes profile of (A) Napolifuel oil; (B) limpet shell swabs, and (C) seawater from the sea-surface microlayer sampled near the shipwreck.

Integrated impact assessment of spilled oils on local biota Environ. Toxicol. Chem. 29, 2010 1363

Fig. 4. Gas chromatography–mass spectrometry spell-out analysis of tissuepolyaromatic hydrocarbon content of limpets collected from the beach atBranscombe, England, immediately after the Napoli beached and in July atthe time of the salvage operation, together with limpets from a reference siteat Chesil Beach.

Fig. 5. Biomarker responses according to site of collection; (A) lysosomalstability (measured as neutral red retention); (B) immune function (measuredas phagocytosis activity); (C) DNA damage (measured as DNA strandbreaks). n¼ 20 for each site. Data as mean� standard error; �¼ a significantdifference from both reference sites; ��¼ a significant difference between thetwo samples collected in January and July from Branscombe, England.

1364 Environ. Toxicol. Chem. 29, 2010 C. Lewis et al.

the MSC Napoli. However, chemical analyses of tissues andshell swabs did not reveal the presence of MSC Napoli-relatedoil on the limpets. A detailed chemical characterization toscreen for any contamination of the sea-surface microlayerand subsurface waters surrounding the MSC Napoli shipwreckconducted by Guitart et al. [16], confirmed that the oil spilledduring the initial beaching of the ship was primarily present asa localized surface sheen. In assessing the ecological impactsof spilled crude oils, the compounds of most immediate tox-icological relevance, i.e., those exhibiting the highest acutetoxicity, are the PAHs. Spatial interpolation plots of the ana-lytical data for the total PAH concentrations in Lyme Bayimmediately following the grounding of the MSC Napoli inJanuary [16] revealed that the chemical signature of theheavy fuel oil was not apparent in the coastal sample extracts.Slightly elevated PAH concentrations were found in subsurfacewaters in the far west and the eastern parts of the bay (range,13.8–14.6 ng/L) which could not be related directly to the oilspill from MSC Napoli and might relate to different inputs ofPAH from the costal area (e.g., runoff, commercial and leisuremaritime activity).

Consideration of fluoranthene/pyrene (FL/Py) ratios (shownin Table 1) demonstrates that pyrogenic inputs of PAHs(derived from the combustion of organic matter and dominatedby four-to-six ringed PAHs) are dominant in the subsurfacewaters in Lyme Bay and are distinct from the petrogenic profile(dominated by low molecular weight two-to-three ringedPAHs) in the immediate vicinity of the wreck itself [16]. Majorinputs from the Exe estuary, Brixham Harbor, and PortlandHarbor (i.e., runoff, wastewater treatment works, and shippingcontamination) are all potential sources of pyrogenic PAHs andof other unspecified contaminants capable of contributingtowards the impacted biomarker profile of limpets from Bran-scombe collected in January.

Salvage of the MSC Napoli eventually required the hull to besplit, and this was achieved in July 2007 using explosivecharges that caused additional oil to be spilled. We havepresented unequivocal evidence to show that the oil collectedfrom limpet shell swabs from Branscombe Beach matches thefuel oil from the MSC Napoli. Oil fingerprinting by GC/MS hasbeen widely applied to characterize different oil samples and tofingerprint the sources. Individual aliphatic hydrocarbons, iso-

prenoid hydrocarbons (pristane and phytane), parent PAHs andpetroleum-specific alkylated homologues of PAHs (e.g., naph-thalenes, dibenzothiophenes, phenanthrenes) can all be used tocharacterize samples. For example, isoprenoid hydrocarbons(pristane and phytane) are formed during formation of oil from

Integrated impact assessment of spilled oils on local biota Environ. Toxicol. Chem. 29, 2010 1365

the phytyl side chain of chlorophyll and are considered char-acteristic for different oils [29,30]. The oil from the MSC Napoli(identified as IFO 380) is a heavy fuel oil commonly used forship’s engines. From the aliphatic fraction shown in Figure 2,this oil shows a bimodal distribution containing a high propor-tion of lighter n-alkanes, which are reduced considerably in theseawater and swab samples, despite the short time betweenspillage and shoreline collection. This is due to volatilization ofn-alkanes and other compounds. Despite this, the chromato-grams clearly show a similar profile for higher molecularweight n-alkanes. The nC17/pristane and nC18/phytane ratios(Table 1), as well as the terpane profiles in Figure 3, clearlyindicate similarity of origin of the fuel oil and the beachedresidues.

Tissue extracts from limpets collected at Branscombe Beachfollowing the release of oil from the MSC Napoli duringthe salvage operation in July also contained elevated concen-trations of PAHs compared to a control location and prespillBranscombe. These extracts were dominated by phenanthrenesand were consistent with hydrocarbon contamination from oilspillage [30]. The much smaller concentrations by some largermolecular weight hydrocarbons may have been of pyrolyticorigin. Tissue extracts with relatively high concentrationsof phenanthrenes and alkyl phenanthrenes also containedan unresolved complex mixture of hydrocarbons indicativeof weathering of the Napoli oil [15]. The toxicity of phenan-threnes to marine species is well documented [6], and recentevidence has demonstrated that aromatic unresolved complexmixture compounds also contribute towards toxic effects[20,31–32].

Our results illustrate the advantages of fingerprinting oilfrom the shells, in combination with tissue analysis wherefingerprinting for the source of a spillage is a priority. Evenwhen limpets were seen on the shore to be coated in oil(following the Sea Empress oil spill, UK, 1996), extracts fromthe internal tissues could not reliably confirm the source [17].Selective accumulation, breakdown, and excretion of contam-inant hydrocarbons can lead to variable rates of uptake anddepuration. Added to this, the variability in the presenceand concentration of hydrocarbons from past contaminationtogether with the presence of biogenic hydrocarbons such aspristanes, can all make fingerprinting of tissue residues and theresolution of individual component chemicals a complex andsubjective task [32]. Systematic application of ecological riskassessment principles to the assessment of accidental spillagesof oils and hazardous substances over the past few years hasallowed the development of a basic framework to link thesource of the spill to biological receptors of concern. The mostcritical aspects of this process, as illustrated in the present study,remains the need to determine the nature and extent of exposureof biota to contamination and to demonstrate significant, sub-lethal effects upon a range of physiological and biochemicalendpoints in exposed organisms.

The toxicity of the PAHs is generally attributed to theirability to interfere with membrane stability and function, toinduce Ah-dependent enzyme activities and to cause oxidativedamage to macromolecules after metabolic activation throughthe cytochrome P450 mixed-function oxidase system [33].Mollusks, including the limpet, differ from fish, marine birds,and marine mammals in that although cytochrome P450-typeactivities have been identified [34], there is some controversyover whether their activity is induced to the same extent as inhigher organisms [34–37]. Mollusks overall have a limitedability to metabolize petroleum hydrocarbons and, after expo-

sure to an oil spill, tend to bioaccumulate petroleum hydro-carbons [17,38]. Although the biomarkers chosen for inclusionin the present study (antioxidant status, damage to DNA,hemocyte viability, and phagocytic activity) have been previ-ously demonstrated to respond in a concentration-dependentmanner to petroleum-derived PAHs in controlled laboratorystudies [12,13], they cannot be considered to be specific for theeffects of PAHs alone, but will also reflect the impact of othercontaminants and of any environmental stress on the exposedanimals.

Underlying PAH contamination in areas surrounding acuteoil spill incidents have previously been demonstrated. Forexample, Page et al. [39] explored the sources of low-levelexposure to PAH in fish in Prince William Sound, USA, outsideof the spill path caused by the Exxon Valdez in 1989, andconcluded that measurable alterations in biomarker responseswere caused by pervasive exposure to a regional background ofPAHs, derived from variable natural and anthropogenic sour-ces. Laboratory studies in fish have suggested that biomarkerresponses to petrogenic PAHs can be quite different fromresponses to pyrogenic PAHs [33] due to differences in theinduction of cytochrome P450 activities [39]. As discussedabove, the nature of induction of mixed function oxidaseactivities in prosobranch mollusks is less clear. It is apparentfrom Figure 5 that there is some variation in the pattern ofbiological impact before and after the acute fuel oil exposure.For example, hemocyte viability was significantly depressed atBranscombe compared with the control sites at both samplingtimes. However, the phagocytic activity measured in limpetsfrom Branscombe in January showed marked stimulation com-pared to the control sites, which then became totally suppressedfollowing the acute fuel oil exposure. Biphasic responses of thismagnitude and nature are characteristic of inflammatory stim-ulation and have been reported in response to low concentra-tions of organic pollution, pesticides, and metals in a range ofinvertebrate species [40].

By combining comprehensive chemical fingerprinting tech-niques of high precision and accuracy with measurements ofbiota exposure and biomarkers of sublethal biological effect in asentinel organism of ecological importance, we have been ableto distinguish spilled oil from a preexisting chronic, backgroundPAH contamination within Lyme Bay. This provides clearinformation regarding the timing and extent of PAH exposuresfor local biota that resulted from the MSC Napoli incident andtheir resulting rapid biological response. The integration ofbiological and chemical biomarkers in this way offers potentialbenefits for responders in providing greater certainty for thedecision-making process in environmental management prac-tices, and directly addresses the policy need of improvedscreening risk assessment methodologies for ecological damageassessment.

Acknowledgement—Supported by Department of Environment, Fisheriesand Rural Affairs and Centre for Environment, Fisheries and AquacultureScience, UK, contract PO007355, and European Commission FP6 FACE-ITprogram, contract 018391. C. Guitart was supported by EIF Fellowship(CT2005-023699). We thank the Marine and Coastguard Agency for samplesand Robin Law for helpful discussions.

REFERENCES

1. Department for Environment, Food and Rural Affairs. 2005. Chartingprogress: An integrated assessment of the state of UK seas. MarineStewardship Report. London, UK.

2. McDonald BG, Debruyn AMH, Wernick BG, Patterson L, Pellerin N,Chapman PM. 2007. Design and application of a transparent and scalable

1366 Environ. Toxicol. Chem. 29, 2010 C. Lewis et al.

weight-of-evidence framework: An example from Wabamun Lake,Alberta, Canada. Integr Environ Assess Manag 3:476–483.

3. European Union. 2003. Common implementation strategy for theEuropean Water Framework Directive, Guidance Document 5. Transi-tional and Coastal Waters, Luxembourg.

4. Boehm PD, PageDS. 2007. Exposureelements in oil spill risk and naturalresource damage assessments: A review. Human and Ecological RiskAssessment 13:418–448.

5. Seager TP, Satterstrom FK, Linkov I, Tuler SP, Kay R. 2007.Typological review of environmental performance metrics (withillustrative examples for oil spill response). Integr Environ AssessManag 3:143–150.

6. National Research Council. 2003. Oil in the Sea III: Inputs, Fates andEffects. National Academy Press, Washington, DC.

7. U.S. Environmental Protection Agency. 1998. Guidelines for ecologicalrisk assessment. EPA/630/R-95/002F. Washington, DC.

8. OSPAR Commission for the Protectionof the Marine Environment of theNorth East Atlantic. 2000. Quality Status Report 2000. London, UK.

9. Peakall DB, Shugart LR. 1992. Biomarkers, Research, and Applicationin the Assessment of Environmental Health. Springer-Verlag, Berlin,Germany.

10. Hagger J, Jones M, Leonard P, Owen R, Galloway TS. 2006. Biomarkers,are there more questions than answers? Integr Environ Assess Manag2:312–329.

11. Allan IJ, Vrana B, Greenwood R, Mills GA, Roig B, Gonzalez C. 2006.A toolbox for biological and chemical monitoring requirements for theEuropean Union’s Water Framework Directive. Talanta 69:302–322.

12. Galloway TS, Sanger RC, Smith KL, Fillmann G, Ford TE, DepledgeMH. 2002. Rapid assessment of marine pollution using biomarkers andchemical immunoassays. Environ Sci Technol 36:2219–2226.

13. Galloway TS, Brown RJ, Browne MA, Dissanayake A, Lowe D, JonesMB, Depledge MH. 2004. A multi-biomarker approach to ecosystemmanagement. Environ Sci Technol 38:1723–1731.

14. Anderson J, Lee R. 2006. Use of biomarkers in oil spill risk assessment inthe marine environment. Human and Ecological Risk Assessment 12:1192–1222.

15. Readman JW, Guitart C, Frickers T, Law R. 2007. An assessment ofchemical pollution from the MSC Napoli. Mar Pollut Bull 54:501–503.

16. Guitart C, Frickers P, Horillo-Carballo J, Law R, Readman JW. 2008.Characterization of surface microlayer chemical contamination aftershipping accidents. Environ Sci Technol 42:2275–2282.

17. Simpson RD, Smith S, Pople AR. 1995. Effects of a spillage of diesel fuelon a rocky shore in the sub-Atlantic region (Macquane Island). MarPollut Bull 31:367–371.

18. Glegg GA, Hickman L, Rowland SJ. 1999. Contamination of limpets(Patella vulgata) following the Sea Empress oil spill. Mar Pollut Bull38:119–125.

19. Booth AM, Sutton PA, Lewis CA, Lewis AC, Scarlett A, Chau W,Widdows J, Rowland SJ. 2007. Unresolved complex mixtures ofaromatic hydrocarbons: Thousands of overlooked persistent, bioaccu-mulative, and toxic contaminants in mussels. Environ Sci Technol41:457–464.

20. Scarlett A, Rowland SJ, Galloway TS, Lewis AC, Booth AM. 2008.Chronic sublethal effects associated with branched alkylbenzenesbioaccumulated by mussels. Environ Toxicol Chem 27:561–567.

21. Moore MN, Allen JI, McVeigh A. 2006. Environmental prognostics: Anintegrated model supporting lysosomal stress responses as predictivebiomarkers of animal health status. Mar Environ Res 61:278–304.

22. Babich H, Borenfreund E. 1992. Neutral red assay for toxicology in vitro.In Watson RR, ed, Methods in Toxicology and Pharmacology. CRC,Boca Raton, FL, USA, pp 237–251.

23. Lewis C, Galloway TS. 2009. Reproductive consequences of paternalgenotoxin exposure in marine invertebrates. Environ Sci Technol43:928–933.

24. Lewis C, Galloway TS. 2008. Genotoxic damage in Polychaetes: A studyof species and cell-type sensitivities. Mutat Res Genet Toxicol EnvironMutat 654:69–75.

25. Jha AN. 2008. Ecotoxicological applications and significance of thecomet assay. Mutagenesis 23:207–221.

26. Galloway TS, Depledge MH. 2001. Immunotoxicity in invertebrates:Measurement and ecotoxicological relevance. Ecotoxicology 10:5–23.

27. Pipe RK, Coles JA. 1995. Environmental contaminants influencingimmune function in marine bivalve molluscs. Fish Shellfish Immunol5:581–595.

28. Rickwood C, Galloway TS. 2004. Can acetylcholinesterase beconsidered a biomarker of effect? A study of the mussel Mytilus edulisexposed to the priority pollutant Chlorfenvinphos. Aquat Toxicol 67:45–56.

29. Wang Z, Fingas MF. 2003. Development of oil hydrocarbon finger-printing and identification techniques. Mar Pollut Bull 47:423–452.

30. Wang Z, Stout SA. 2007. Oil Spill Environmental Forensics. Finger-printing and Source Identification. Academic/Elsevier, Amsterdam, TheNetherlands.

31. Booth A, Scarlett A, Lewis CA, Belt ST, Rowland SJ. 2008. Unresolvedcomplex mixtures (UCMs) of aromatic hydrocarbons: Branched alkylindanes and branched alkyl tetralins are present in UCMs andaccumulated by and toxic to, the mussel Mytilus edulis. Environ SciTechnol 42:8122–8126.

32. Scarlett A, Galloway TS, Rowland SJ. 2007. Chronic toxicity ofunresolved complex mixtures (UCM) of hydrocarbons in marinesediments. J Soils Sediments 7:200–206.

33. Goksoyr A, Forlin L. 1992. The cytochrome P-450 system in fish, aquatictoxicology and environmental monitoring. Aquat Toxicol 22:287–312.

34. Peters LD, Shaw JP, Nott M, O’Hara SCM, Livingstone DR. 1999.Development of cytochrome P450 as a biomarker of organic pollution inMytilus sp: Field studies in the UK and the Mediterranean. Biomarkers4:425–441.

35. Lemaire P, Livingstone DR. 1993. Pro-oxidant/anti-oxidant processesand organic xenobiotic interactions in marine organisms. Trends CompBiochem Physiol 1:1119–1149.

36. Rebelo M de F, Moreira JC, Morales MO, CoutinhoD, Torres JPM. 2003.Is there a CYP1A1 in marine bivalves? Expressed sequences related tocytochrome P450 in Crassostrea rhizophorae. Comments on Toxicology9:279–287.

37. Porte C, Lemaire P, Peters LD, Livingstone D. 1995. Partial purificationand properties of Cytp450 from digestive gland microsomes of thecommon mussel Mytilus edulis. Mar Environ Res 39:27–31.

38. Gianguzza A, Orecchio S. 2006. The PAH composition in limpets(Patella vulgata L.) from the coasts of Sicily. Polycyclic AromaticCompounds 26:37–57.

39. Page D, Huggett R, Stegeman J, Parker K, Woodin B, Brown J, Bence E.2004. Polycyclic aromatic hydrocarbon sources related to biomarkerlevels in fish from Prince William Sound and the Gulf of Alaska. EnvironSci Technol 38:4928–4936.

40. Galloway TS, Goven A. 2006. Wildlife immunotoxicology: Theinvertebrates. In House RV, Luebke R, Kimber I, eds, Immunotoxicologyand Immunopharmacology. CRC, Boca Raton, FL, USA, pp 365–384.