Computers & Electronics

A multidisciplinary weight of evidence approach for classifying polluted sediments: Integrating sediment chemistry, bioavailability, biomarkers responses and bioassays

A multidisciplinary weight of evidence approach for classifying polluted sediments: Integrating sediment chemistry, bioavailability, biomarkers responses and bioassays
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  A multidisciplinary weight of evidence approach for environmentalrisk assessment at the Costa Concordiawreck: Integrative indices fromMussel Watch Francesco Regoli a , * , David Pellegrini b , Anna Maria Cicero c , Marco Nigro d ,Maura Benedetti a , Stefania Gorbi a , Daniele Fattorini a , Giuseppe D ’ Errico a ,Marta Di Carlo a , Alessandro Nardi a , Andrea Gaion b , Alice Scuderi b , Silvia Giuliani b ,Giulia Romanelli c , Daniela Berto e , Benedetta Trabucco c , Patrizia Guidi d ,Margherita Bernardeschi d , Vittoria Scarcelli d , Giada Frenzilli d a Dipartimento di Scienze della Vita e dell ’   Ambiente, Università Politecnica delle Marche, Ancona, Italy b ISPRA, Istituto Superiore per la Protezione e la Ricerca Ambientale, Livorno, Italy c ISPRA, Istituto Superiore per la Protezione e la Ricerca Ambientale, Roma, Italy d Dipartimento di Medicina Clinica e Sperimentale, Università di Pisa, Italy e ISPRA, Istituto Superiore per la Protezione e la Ricerca Ambientale, Chioggia, Italy a r t i c l e i n f o  Article history: Received 31 July 2013Received in revised form18 September 2013Accepted 21 September 2013 Keywords: Costa ConcordiaMusselsBioavailabilityBiomarkersWeight of evidenceEnvironmental risk assessment a b s t r a c t A complex framework of chemical, biological and oceanographic activities was immediately activatedafter the Costa Concordia shipwreck, to assess possible contamination events and the environmentalimpact during both emergency and wreck removal operations. In the present paper, we describe theresults obtained with caged mussels,  Mytilus galloprovincialis , chosen as bioindicator organisms to detectvariations of bioavailability and the early onset of molecular and cellular effects (biomarkers). Seventranslocation experiments were carried out during the  fi rst year from the incident, with organismsdeployed at 2 depths in 3 different sites. After 4 e 6 weeks, tissue concentrations were measured for themain classes of potentially released chemicals (trace metals, polycyclic aromatic hydrocarbons, volatileand aliphatic hydrocarbons, polychlorinated biphenyls, halogenated pesticides, organotin compounds,brominated  fl ame retardants, anionic surfactants); a wide battery of biomarkers covered responsesindicative of exposure, detoxi fi cation, oxidative stress, cell damage and genotoxic effects. Resultsexcluded serious contamination events or a consistent increase of environmental pollution althoughsome episodic spills with reversible effects were detected. Data were elaborated within a quantitativeweight of evidence (WOE) model which provided synthetic hazard indices for each typology of data,before their overall integration in an environmental risk index, which generally ranged from slight tomoderate. The proposed WOE model was con fi rmed a useful tool to summarize large datasets of complexdata in integrative indices, and to simplify the interpretation for stakeholders and decision makers, thussupporting a more comprehensive process of   “ site-oriented ”  management decisions.   2013 Elsevier Ltd. All rights reserved. 1. Introduction On the night of 13 January 2012, the Costa Concordia ship, with3.206 passengers and 1.023 crew members on board, collided witha submerged rock close to the Giglio Island (Tuscany, Italy),causing 32 victims. A coincidence of winds and currents preventedthe ship from sinking in the deep waters, and the partiallysubmerged wreck laid on a rock ledge at the entrance of Giglioharbour.Immediate fears of an ecological disaster were raised from thepossibility of the wreck to slip at a much greater depth, concomi-tantly with the risk of oil pollution and loss of other chemicals thatcould have devastated this highly relevant marine area. An emer-gency plan was rapidly activated after the incident to remove fuelsand oil were from the ship, and to monitor the possible impact of such activities.Acomplexframeworkofchemical,biologicalandoceanographicinvestigations was coordinated by the National Civil Protection and *  Corresponding author. Tel.:  þ 39 (0)712204613; fax:  þ 39 (0)712204609. E-mail address: (F. Regoli). Contents lists available at ScienceDirect Marine Environmental Research journal homepage: 0141-1136/$  e  see front matter    2013 Elsevier Ltd. All rights reserved. Marine Environmental Research xxx (2013) 1 e 13 Please cite this article in press as: Regoli, F., et al., A multidisciplinary weight of evidence approach for environmental risk assessment at theCosta Concordia wreck: Integrative indices from Mussel Watch, Marine Environmental Research (2013), j.marenvres.2013.09.016  the Italian Institute for Environmental Protection and Research(ISPRA); main activities included chemical and ecotoxicologicalanalyses on water column and sediments, bioaccumulation andsublethal effects (biomarkers) in several invertebrates and  fi shspecies, evolution of   Posidonia oceanica  seagrass meadows, densityand distribution patterns of the endangered species  Pinna nobilis ,assessment of   fi sh assemblages by visual census, benthic cartog-raphy on ecologically relevant coastal habitats.The present workdescribes the experimental design and resultsobtained during one year of an active Mussel Watch investigation.Mussels,  Mytilus galloprovincialis , were caged in proximity of theCosta Concordia wreck to detect environmental changes, boththrough the analysis of chemicals bioaccumulation and changes of molecular and cellular responses (biomarkers). A similar approachis widely recognized as an important component of environmentalrisk assessment to evaluate bioavailability and early toxicologicaleffects of anthropogenic pollutants released into the marine envi-ronment (McCarty et al., 2002): in this respect, biomarkers oftenrepresent the  fi rst warning signals of environmental disturbance,even in the absence of acutely toxic responses at organism orpopulation levels (Gallowayet al., 2004; Haggeret al., 2008; Regoli,2000; Regoli et al., 2011; Viarengo et al., 2007).The use of caged mussels facilitates the investigation in areaswhere native organisms are absent, also reducing the in fl uence of genetic/population differences, seasonal variability or adaptivephenomena, all factors which can attenuate the capability todiscriminate between different levels of anthropogenic impact(Bocchetti et al., 2008a; Brooks et al., 2012; Marigómez et al.,2013b; Nigro et al., 2006; Zorita et al., 2007). Considering thetolerance of these organisms to environmental conditions andhandling, their elevated capability to accumulate both inorganicand organic pollutants and the responsiveness of molecular andcellular responses, caged mussels can typically provide a time-integrated assessment of bioavailability and early biological ef-fects over a 4 e 6 weeks translocation period (Gorbi et al., 2008;Viarengo et al., 2007).Duringthe fi rstyearfromtheCostaConcordiaincident,atotalof 7 translocation experiments were carried out during both theemergency phase related to removal of oil and fuels, and the op-erations of the  “ Parbuckling project ”  aimed to re fl oat and towawaythe wreck. Several classes of pollutants potentially released duringsuch activities were analysed in tissues of caged mussels, i.e. tracemetals, polycyclic aromatic hydrocarbons (PAHs), volatile andaliphatic hydrocarbons (C6 e C10 and C10 e C40), polychlorinatedbiphenyls (PCBs), organo-chlorinated pesticides (OCPs), organotincompounds, brominated  fl ame retardants (BFRs) and anionicsurfactants.These measurements were integrated with a multi-biomarkerapproach, based on a wide array of molecular and subcellularresponses re fl ecting speci fi c mechanisms of chemicals exposureor detoxi fi cation, cellular pathways of oxidative response toreactive oxygen species (ROS) and oxidative stress, onset of different levels of cellular unbalance and genotoxic effects.Metallothioneins-like proteins, acetylcholinesterase and Acyl CoAoxidase (AOX) were measured as typical responses to metals andseveral forms of organic pollutants. Metallothioneins, physiolog-ically involved in intracellular homeostasis of essential metals, aretypically induced in organisms exposed to increased environ-mental concentrations of these elements (Viarengo et al., 2007); the activity of acetylcholinesterase, a crucial enzyme in theneurotransmission of vertebrates and invertebrates, is inhibitedby organophosphate or carbammate pesticides, and modulated byother chemicals and bioactive compounds (Gorbi et al., 2008,2012). Peroxisomes are involved in lipid metabolism throughoxidative reactions, and their proliferation in mussels exposed toorganic xenobiotics is re fl ected by increased number and volumeof organelles, and the induction of enzymes involved in fatty acidoxidation, such as AOX (Cajaraville et al., 2003; Orbea andCajaraville, 2006).Since a typical pathway of chemical toxicity is mediated by theenhancement of intracellular generation of reactive oxygen species(ROS), variations of antioxidant defences are rapidly activated inMediterranean mussels to counteract enhanced prooxidant chem-icalchallenge(Frenzillietal.,2004;Giulianietal.,2013;Marigómezet al., 2013b; Regoli, 1998, 2000; Regoli and Giuliani, 2013; Regoliand Principato, 1995; Viarengo et al., 2007). In the present study,oxidative effects in caged mussels were investigated throughthe activities of catalase, glutathione S-transferases, glutathionereductase, glutathione peroxidases, levels of glutathione; theseresults were integrated with the total oxyradical scavenging ca-pacity (TOSC) towardperoxylradicals, ROO  , and hydroxylradicals,HO   (Regoli and Winston, 1999; Regoli et al., 1998). Compared toindividualantioxidants,variationsofTOSChavea greaterbiologicalrelevance and prognostic value, being an impaired capability toneutralize ROS associated tothe onset of various forms of oxidativedamages like lysosomal dysfunctions, lipid peroxidation and gen-otoxic effects (Camus et al., 2003; Frenzilli et al., 2001, 2004; Gorbiand Regoli, 2003; Meyer et al., 2003; Moore et al., 2006; Regoli,2000; Regoli et al., 2004).Toxic effects on lysosomal systems are generally included in allthe ecotoxicological studies with mussels (Moore et al., 2004,2006). These organelles are responsible of fundamental processesin cell physiology, food digestion, intracellular turnover, immunefunction, sequestration and excretion of harmful compounds(Moore et al., 2006); however, they are also highly sensitive to several contaminants which can act both directly on the mem-branes and indirectly, through generation of ROS or signallingpathways (Regoli, 1992, 2000; Moore et al., 2006; Viarengo et al.,2007). In this study, lysosomal impairment was measured asmembrane stability, while content of malondialdehyde, accumu-lation of lipofuscin and neutral lipids in tertiary lysosomes wereinvestigated as further indices of lipid peroxidation processes andxenobiotic-mediated lipidosis (Regoli, 1992). The presence of agenotoxic risk around the wreck was  fi nally evaluated in cagedmussels as both DNA strand breaks, a potentially repairable effectcaused by chemicals or enhanced prooxidant challenge, and asfrequency of micronuclei, an irreversible genetic damage of chro-matin breakage or aneuploidy during cell division (Frenzilli et al.,2009; Nigro et al., 2006).Multidisciplinary studies which combine chemical and bio-logical measurements, represent an added value to monitoringand management protocols, and also recent European Directivesrecommend the use of different quality indicators to evaluate theenvironmental status of aquatic ecosystems (Chapman, 2007;Chapman et al., 2013; Galloway et al., 2004; Hagger et al., 2008;Marigómez et al., 2013a). However, the combination of multipletypologies of investigations (or lines of evidence, LOEs) is oftenhampered by the lack of standardized procedures for the inter-pretation and the integration of complex datasets of heteroge-neous results, which typically require various expert judgements(Benedetti et al., 2012; Dagnino et al., 2008; Linkov et al., 2009;Piva et al., 2011; Semenzin et al., 2008); additional critical is-sues are represented by development of qualitative and quanti-tative evaluations, the synthetic classi fi cation and, last but notleast, the communication of risk to stakeholders (Benedetti et al.,2012; Linkov et al., 2009; Marigómez et al., 2013a). In this respect,a quantitative  “ Weight Of Evidence ” , WOE model (Sediqualsoft)has been recently developed to integrate and differently weightdata from various lines of evidence, which actually include sedi-ment chemistry, bioavailability of chemicals in key bioindicator F. Regoli et al. / Marine Environmental Research xxx (2013) 1 e 13 2 Please cite this article in press as: Regoli, F., et al., A multidisciplinary weight of evidence approach for environmental risk assessment at theCosta Concordia wreck: Integrative indices from Mussel Watch, Marine Environmental Research (2013), j.marenvres.2013.09.016  species, sublethal effects measured through biomarkers, toxico-logical effects at organism level assessed by standardized labo-ratory bioassays (Piva et al., 2011): different typologies of data are initially evaluated within single modules by logical  fl owchartsand mathematical algorithms which provide speci fi c hazardindices for each of considered line of evidence, before their dif-ferential weighting and integration in a quantitative WOE eval-uation (Piva et al., 2011). The Sediqualsoft model has been previously applied to different multidisciplinary studies for thecharacterization and classi fi cation of sediment quality, and theassessment of environmental hazards in coastal areas andbrackish environments (Benedetti et al., 2012, 2013; Piva et al.,2011).In this study, the bioaccumulation and biomarkers results of theMussel Watch have been elaborated within the WOE model as theinitial stepof a moredetailed environmental risk assessment in theareawhichwillincludetheintegrationofadditionalLOEslikethoseon bioaccumulation and biomarkers in native  fi sh and in-vertebrates, chemical and ecotoxicological characterization of sediments and seawater, changes of benthic communities andseagrass meadows. The overall aimwas to further demonstrate theversatility of the WOE approach to summarize huge amounts of complex datasets from different LOEs: the elaborated syntheticindices were expected to discriminate with elevated sensitivity thedifferent environmental scenarios around the Costa Concordiawreck, and provide scienti fi cally sound information which mightbeeasily interpreted alsobynon-expert stakeholdersand decision-makers. 2. Materials and methods  2.1. Experimental design and Mussel Watch experiments Mussels,  M. galloprovincialis  (5.5    0.5 cm shell length), weresampledfromthereferencesiteofPortonovo(Ancona,Adriaticsea)at a depth of 8 m, immediately transported to Giglio Island anddeployed within 24 h. Three caging sites were selected (Fig. 1),respectively north of Costa Concordia wreck in proximity of thestern (42  22 0 04.80 00 N, 10  55 0 16.80 00 E), in front of Giglio Portomarina (42  21 0 37.20 00 N, 10  55 0 22.50 00 E) and in a reference sitelocated in front of the Caldane beach, approximately 2 nauticalmiles south of the wreck (42  20 0 42.00 00 N, 10  55 0 27.00 00 E). Toevaluate possible differences along the water column, at each siteorganisms were caged at 2 different depths, approximately 1.5 mfrom both the surface and the bottom (30 e 45 m). Translocationexperiments have been carried on February 2012 (two weeks afterthe incident, T1), March (T2), April (T3), May (T4), June (T5),October (T6) and December 2012 (T7): organisms have beenrecovered after 4 e 6 weeks from the deployment.At the end of translocation periods, for every site and depth,15pools, each constituted by whole tissues of at least 10 organismswere dissected and stored at   20   C for chemical analyses. Forbiomarker measurements, digestive glands were rapidly removedfrom 30 specimens, pooled in 10 samples (each with tissues of 3specimens), frozen in liquid nitrogen and maintained at   80   C;small pieces of digestive glands were rapidly excised from 5 mus-sels for histological analyses, placed on cork chucks, frozen in  n - Fig. 1.  Localization of translocation sites close to the Costa Concordia wreck, in front of the marina of Giglio Porto, and at the Caldane beach chosen as reference site. F. Regoli et al. / Marine Environmental Research xxx (2013) 1 e 13  3 Please cite this article in press as: Regoli, F., et al., A multidisciplinary weight of evidence approach for environmental risk assessment at theCosta Concordia wreck: Integrative indices from Mussel Watch, Marine Environmental Research (2013), j.marenvres.2013.09.016  hexane precooled to   70   C in liquid nitrogen, and maintainedat  80  C; hemolymph was withdrawnfromadductor muscle of 15specimens, divided in 5 aliquots, each with  fl uids of 3 individuals,and immediately frozen in liquid nitrogen; genotoxic effects weremeasured in branchial cells isolated from gills of 5 specimens.  2.2. Chemical analyses Volatile (C6 e C10) and aliphatic (C10 e C40) hydrocarbons,polycyclic aromatic hydrocarbons (PAHs), polychlorinated bi-phenyls (PCBs), organo-halogenated pesticides (OCPs), organotincompounds (OSn), brominated  fl ame retardants (BFRs), tracemetals(As,Cd,Cr,Cu, Hg,Ni,Pb,V,Zn)andtotalanionicsurfactants(TASs) were analysed in the whole body mass of mussels by con-ventional procedures based on gas-chromatography with  fl ameionization detector (FID), electron capture detector (ECD), massdetector (MS), high performance liquid chromatography (HPLC)with diode array (DAD) and  fl uorimetric detection, inductivelycoupled plasma (ICP) and atomic absorption spectrophotometry(AAS). Details on analytical methods and procedures for qualityassurance/quality control are given in Supplementary Material(SM1).  2.3. Biomarkers analyses Biomarkers in tissues of caged mussels were measured throughstandardized protocols (Bocchetti et al., 2008a) which included: spectrophotometric determination of acetylcholinesterase (AChE)in gills and hemolymph, and of metallothioneins (MTs), Acyl-CoAoxidase (AOX), antioxidants defences (catalase, glutathione S-transferases, GST, Se-dependent and sum of Se dependent and Se-independent glutathione peroxidases, GPx-H 2 O 2  and GPx-CHP,glutathione reductase, GR, glutathione, GSH), malondialdehyde(MDA) in digestive glands; gas-chromatographic assay of totalantioxidant capacity (TOSC) toward peroxyl radicals (ROO $ ) andhydroxyl radicals (HO $ ) in digestive glands; histochemical quanti- fi cation of lysosomal membrane stability, lipofuscin and neutrallipids on digestive gland cryostat sections; electrophoretic andcytogenetic analysis of DNA integrity in branchial cells and inhaemocytes. Detailed protocols are given in SM1.  2.4. Statistical analyses and weight of evidence (WOE) elaboration Analysis of variance (ANOVA) was applied to chemical data andbiomarker responses to test the signi fi cance of the differences be-tween areas (level of signi fi cance at the 95% of con fi dence interval, a  ¼  0.05); homogeneity of variance was tested by Cochran C, and  post-hoc   comparisons (Student Newman e Keuls) were used tocompare means of values. The data on DNA damage (comet andMNs assays) were analysed by the non-parametric Kruskal e Wallistest.Anon-metricMultiDimensionalScaling(nMDS)wasperformedas multivariate ordination analysis in which variables describing amultidimensional space are scaled on a two-dimension plot basedon their similarity, thus maximizing the distance among points(Clarke and Gorley, 2001). The metaMDS function from the VeganR-package (Oksanen et al., 2011) was applied on bioaccumulationand biomarker responses measured in mussels caged for all the 7sampling periods (T1-T7) at the 3 sites (Reference, Marina andWreck), and 2 depths (super fi cial and bottom). Clusteranalysis wasusedtoidentifydiscretegroupsandtheirdissimilarity;Bray e Curtisdissimilarity between groups was calculated in R and visualizedthrough cluster plot with the UPGMA methods (Vegan R-Package).Data on bioaccumulation and variations of biomarkers werefurther elaborated within the quantitative WOE model,Sediqualsoft, which summarizes hazard indices for various LOEs inspeci fi c modules, before their overall integration: whole calcula-tions, detailed  fl ow-charts, rationale for weights, thresholds andexpert judgements have been fully given elsewhere (Benedettiet al., 2012; Piva et al., 2011).Brie fl y, the elaboration of bioavailability hazard within LOE2 isbased on the initial calculation, for each pollutant, of the Ratio toReference (RTR), i.e. the increase of concentration compared tocontrol organisms; from this value, a RTRw is obtained by theapplication of a correction factor ( w ) which depends on the sta-tistical signi fi cance of the difference and the typology of chemical(if   “ non priority ” ,  “ priority ”  or  “ priority and hazardous ”  accordingto EC Directive, 2008/105). Based on expert judgement, the modelassigns the RTRw of each parameter to 1 of 5 classes of effect(Absent, Slight, Moderate, Major, Severe); the calculation of thecumulative Hazard Quotient for bioavailability (HQ  BA ) does notconsider parameters with RTRw  < 1.3 (effect Absent, i.e.concentrations    control value for a priority and hazardous pol-lutants), calculates the average for those with RTRw ranging be-tween 1.3 and 2.6 (Slight, i.e. up to 2 fold increase compared tocontrols), and adds the summation ( S ) of all those with RTRw  2.6(i.e. effects Moderate, Major and Severe): HQ  BA  ¼ P  jn ¼ 1 RTR  W  ð n Þ 1 : 3  RTR  W  < 2 : 6  j  þ X kn ¼ 1 RTR  W  ð n Þ RTR  W   2 : 6 The level of cumulative HQ  BA  is summarized in one class of hazard for bioavailability, from Absent to Severe, depending on thedistribution of analyzed chemicals within the different classes of effect (Benedetti et al., 2012; Piva et al., 2011). The module for elaboration of biomarkers results (LOE3) con-tains a large selection of responses among those more widely usedby scienti fi c community in different bioindicator organisms (Pivaet al., 2011); according to species and tissue, each biomarker, hasa  “ weight ”  based on the relevance of biological endpoint, and a “ threshold ”  for changes of biological signi fi cance which considerboth inductions and/or inhibitions of various responses. For everyanalysed biomarker, the measured variation is compared to thethreshold, then corrected for the weight of the response and thestatistical signi fi cance of the difference compared to controls. Eachbiomarker response is assigned by the model to 1 of 5 classes of effect (from Absent to Severe); the calculation of the Hazard Quo-tient for biomarkers (HQ  BM ) does not consider the contribution of responses with an effect lower or equal to threshold (Absent orSlight), calculates the average for those with an effect up to two-fold compared to the threshold (Moderate) and adds the summa-tion ( S ) for the responses more than 2 fold greater than therespective threshold, i.e. Major or Severe (Piva et al., 2011): HQ  BM  ¼ P N  j ¼ 1 Effect W  ð  j Þ 1 < Effect ð  j Þ 2 numbiomark 1 < Effect ð  j Þ 2 þ X M k ¼ 1 Effect W  ð k Þ Effect ð  j Þ > 2 ! According to variations measured for various biomarkers, themodel summarizes the level of cumulative HQ  BM  in one of   fi veclasses of hazard for biomarkers, from Absent to Severe (Piva et al.,2011).The elaborations of results from individual LOEs are  fi nally in-tegrated within a classical weight of evidence approach which,after normalization of indices to a common scale, gives a differentweighttovariouslinesofevidence;thecalculatedWOElevelofriskis  fi nally assigned to 1 of 5 classes of risk from Absent to Severe(Piva et al., 2011). F. Regoli et al. / Marine Environmental Research xxx (2013) 1 e 13 4 Please cite this article in press as: Regoli, F., et al., A multidisciplinary weight of evidence approach for environmental risk assessment at theCosta Concordia wreck: Integrative indices from Mussel Watch, Marine Environmental Research (2013), j.marenvres.2013.09.016  Fig. 2.  Bioaccumulation of chemicals in tissues of mussels caged at the Costa Concordia wreck. Data are given as percentage variations compared to organisms deployed at thereference site (mean values    standard deviations); different letters indicate statistical differences between means of values as indicated by ANOVA and  post-hoc   comparison. Please cite this article in press as: Regoli, F., et al., A multidisciplinary weight of evidence approach for environmental risk assessment at theCosta Concordia wreck: Integrative indices from Mussel Watch, Marine Environmental Research (2013), j.marenvres.2013.09.016
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks