Sediment integrative assessment of the Bay of Cádiz (Spain): An ecotoxicological and chemical approach

Description
Sediment integrative assessment of the Bay of Cádiz (Spain): An ecotoxicological and chemical approach

Please download to get full document.

View again

of 11
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.
Information
Category:

Others

Publish on:

Views: 76 | Pages: 11

Extension: PDF | Download: 0

Share
Transcript
  Sediment integrative assessment of the Bay of Cádiz (Spain): An ecotoxicological andchemical approach Cristiano V.M. Araújo ⁎ , Fernando R. Diz, Irene Laiz, Luís M. Lubián, Julián Blasco, Ignacio Moreno-Garrido Instituto de Ciencias Marinas de Andalucía (ICMAN), Campus Universitario Río San Pedro s/n, 11510, Puerto Real, Cádiz, Spain a b s t r a c ta r t i c l e i n f o  Article history: Received 7 November 2008Accepted 6 February 2009Available online 24 March 2009 Keywords:Cylindrotheca closterium GranulometryLeadLiquid nitrogenMultivariate analysis Tisbe battagliai This study consisted of the sediment toxicity assessment of the Bay of Cádiz based on two endpoints: growthinhibition for  Cylindrotheca closterium  (benthic microalgae) and fecundity inhibition for  Tisbe battagliai (harpacticoid copepod). A new methodology to eliminate (but not as storage technique) the autochthonousbiota present in the sediment samples by immersing them in liquid nitrogen ( − 196 °C) was also assessed.Sediment toxicity data showed different toxicity levels for both organisms. In general,  T  .  battagliai  was moresensitive; however a good correlation ( r  =0.75;  p  b 0.05) between sediment toxicity results for both specieswas found. Data in pore water (pH, redox potential, and toxicity for microalgae and copepod) and sediment(pH, redox potential, organic carbon, and metal concentrations) demonstrated that ultra-freezing did notalter sample characteristics; thus, this technique can be adopted as a pre-treatment in whole-sedimenttoxicity tests in order to avoid misleading results due to presence of autochthonous biota. Multivariatestatistical analysis such as cluster and principal component analysis using chemical and ecotoxicological datawere employed. Silt and organic matter percentage and lead concentration were found to be the factors thatexplain about 77% of sediment toxicity in the Bay of Cádiz. Assay methodology determined in this study forboth assayed species is considered adequate to be used in sediment toxicity monitoring programs. Resultsobtained using both species show that the Bay of Cádiz can be considered a moderately polluted zone.© 2009 Elsevier Ltd. All rights reserved. 1. Introduction Coastal areas are fundamental to marine life and to maintainbiodiversity (Prósperi and Nascimento, 2006). In these zones,sediment quality is vital for the health of the system. Hence,sediments are considered an important fraction of the ecosystemmainly because they are a potential source of contaminants; none-theless they can also act as sink of contaminants and nutrientsdepending on their physicochemical properties, and on the geochem-ical and hydrodynamic conditions (Malueg et al.,1986; DelValls et al.,2004).Assedimentisaheterogeneousmedium,organismslivinginoron the surface may be exposed to certain substances dissolved in theoverlying water or in the pore water, or by direct contact with thosesubstances adsorbed to sediment particles (Macken et al., 2008).Sedimentoften sequestershigh concentrations of metal mixtures thatare not usually re 󿬂 ective of concentrations in the overlying waters(Tessier and Campbell,1987), especially in littoral systems affected bydifferent contamination sources (DelValls et al., 2002). Sediment bioassays have proven to be a powerful tool in studying sediment-related toxicity in coastal and estuarine systems (Costa et al., 1998;Hack et al., 2008), mainly because the toxicological characterizationsfrom elutriate or interstitial water are not always good predictors of sediment toxicity (Guzzella, 1998). Nevertheless, the lack of standar-dized methods made routine testing of sediment quality problematic.Therefore, assays in which test organisms are exposed to whole-sediment may be encouraged, particularly if there is a suspicion of sediment contamination (Pereira et al., 2000; Roman et al., 2007). Although the Bay of Cádiz is a very important area for industry, thepopulation and the environment, it has suffered pollution problemsrelated to ship, offshore, car and aerospace manufacturing (Carrascoet al., 2003). The bay is a very shallow (less than 3 m) semi-enclosedsystem in which contamination processes are driven by the mechan-ism of chemical transfer across the water – sediment interface(DelValls et al., 1998). These characteristics associated to extensiveand intensive marine aquaculture, industrialization, urban dischargein 󿬂 uence, and the signi 󿬁 cant increase in metal concentrationregistered in the second half of the 20th century make it a zone thatdeserves special attention as pointed out by several authors (Gómez-Parra et al., 1984; Carrasco et al., 2003). Two organisms were used in this study to assess the sedimentquality in the Bay of Cádiz: the benthic diatom  Cylindrothecaclosterium , and the harpacticoid copepod  Tisbe battagliai . Microalga C  .  closterium was chosen as a representative of microphytobenthos forvarious reasons: it presents a high ecological relevance as primaryproducer, it plays a signi 󿬁 cant role for matter and energy transferthrough the benthic food webs, for oxygen production, and for Environment International 35 (2009) 831 – 841 ⁎  Corresponding author. Tel.: +34 956 832 612; fax: +34 956 834 701. E-mail address:  cristiano.araujo@icman.csic.es (C.V.M. Araújo).0160-4120/$  –  see front matter © 2009 Elsevier Ltd. All rights reserved.doi:10.1016/j.envint.2009.02.003 Contents lists available at ScienceDirect Environment International  journal homepage: www.elsevier.com/locate/envint  determining the structure and function of benthic intertidal zones(Miller et al., 1996; Blanchard et al., 2000), and it is a promising test organismforecotoxicologicalstudies(Moreno-Garrido et al., 2003a,b,2007; Araújo et al., 2008). Copepoda is the principaland essentiallink betweenthephytoplanktonandhighertrophiclevels.Asacomponentof marine meiobenthos fauna they are exposed to sediment particlesnot only through direct contact, but also through ingestion. Thisincreases the ecological relevance of this group in sediment bioassays(Green et al.,1993). The species  T  .  battagliai  has beenwidely used as atestorganism(ISO,1999),buttherearenoreferencesofitsapplicationfor whole-sediment assays. This species has recently been included instudies of sediment quality dealing with pore water tests (Mackenet al., 2008), despite the fact that aqueous experiments conducted onobligate benthos are not environmentally realistic (Hagopian-Schle-kat et al., 2001). Thus, the use of these two species representing twodifferent trophic levels is extremely important in ecotoxicologicalstudies. Sediment quality assessment through toxicity assays tointegrate the complex effects of compounds has been widelyemployed (Volpi Ghirardini et al., 2005), since data on compoundconcentration separately do not provide an effective basis forestimating the toxic potential for living resources (Long et al., 1995).When both ecotoxicological and physical – chemical data are inte-grated in multivariate statistical techniques, such as cluster analysisand principal component analysis (PCA), a better predictive powerand understanding can be reached (Shin and Fong, 1999; VolpiGhirardini et al., 2005; McPherson et al., 2008).Aseriousproblemencounteredwhencarryingoutsedimenttoxicityassays is the presence of natural predators in the sediment. Thisinterference may be solved due to the predation on the test organismthat will lead to a false positive result; moreover, the presence of morphologically similar species and competitor organisms is another “ confounding factor ”  (Day et al.,1995; DeFoe and Ankley,1998). These two factors can either underestimate or overestimate the results. Dayetal.(1995)showedthatthepresenceofindigenousspeciesinsedimentcancompromisetheinterpretationofresults.Awidelyusedprocedureissievingthesedimentthrough1to2mmmesh;howeveritmightnotbeef  󿬁 cient in removing cocoons, eggs, and young instars of manyinvertebrates,suchasoligochaetewormsandchironomids(Reynoldsonet al., 1994). Other potential methods for removing the indigenousmacrofauna include removal by hand, freezing, heating or drying,chemical sterilants and irradiation (SETAC,1993). However, sometimesthosemethodsdonotseemveryef  󿬁 cientforremovingthemicroorgan-isms and sometimes they modify characteristics of the samples due totheir aggressiveness. In this study, immersion of sediment samples inliquid nitrogen ( − 196 °C) was employed in order to remove theindigenousbiota.Itshouldbepointedoutthatthismethodwasnotusedto store the samples, because sediment freezing as a storage method isnot recommended (Malueg et al.,1986).Thisstudyaimsto(i)identifythesedimenttoxicitylevelsintheBayofCádizforbenthicmicroalgae C  . closterium andharpacticoidcopepod T  . battagliai ;(ii)obtainageneralspatialvisionthroughclusteranalysison similarity between sampled stations; and (iii) discriminate veryimportant variables to sediment toxicity using principal componentsanalysis, in order to point out possible contamination sources. In thepresent study a new methodology for short-term assays focusing onsublethal effects on both organisms  C  .  closterium  and  T. battagliai  towhole-sediment assayswas established.Additionally, the ef  󿬁 ciencyof using liquid nitrogen as a treatment to eliminate autochthonous biotapresent in the sediment was investigated. 2. Materials and methods  2.1. Study site and sampling  The Bay of Cádiz is located at the southwest margin of the IberianPeninsula, between geographic coordinates 36°23' – 36°27' N and6°08' – 6°15' W, and it is surrounded by the cities of Cádiz, SanFernando,andPuertoReal(Carrasco etal., 2003;Ligero et al.,2005).It can be divided into two principal maritime regions: the Outer Bay(open to the Atlantic Ocean), and the Inner Bay (protected fromwaveaction and connected to the Atlantic Ocean through intertidalchannels and salt marshes) (Ligero et al., 2004). The Inner Bay,where this study was centred, is a zone protected from the surge and,partially, the wind, so that water dynamics are tidally controlled(Ligero et al., 2002). Sediment samples were collected from ten stations in the Inner Bay as indicated in Fig.1.The surface layer ( 󿬁 rst 0.5 cm) was sampledduring low tide, usingbox drags (Fig. 2) designed for this work, blended and placed intoseveral plastic  󿬂 asks (100 mL capacity) which were totally  󿬁 lled andhermetically closed to avoid headspace. After sampling, sedimentswere maintained at 4 °C until they were processed (no more than4 days). A sample of acid-washed sand obtained from a non-pollutedarea in southwest Spain was ground and used as negative controlsediment due to its non-toxic characteristics (Moreno-Garrido et al.,2003b). Sediment samples were not sieved.  2.2. Sample characterization Values of pH and redox potential (pH Meter 330 WTW) wereregistered in pore water and directly in sediment for each sample.Organic matter content in the sediment was obtained throughcalcinations (loss of ignition) in a muf  󿬂 e furnace at 450 °C. Sedimentgrain-size distribution was determined by a successive wet sieving.Organic carbon content was determined using the techniquedescribed by Gaudette et al. (1974), modi 󿬁 ed by El-Rayis (1985).Sedimentsampleswerefreeze-driedanddigestedasadaptedfromtheprocedure described by Loring and Rantala (1992). Brie 󿬂 y, sedimentsamplesweredigestedwith10mLMilli-Qwater,5mLHNO 3 ,1mLHCland4mLHFwiththefollowingprogram:ramptime30min,holdtime20 min at 175 psi, later 30 mL boric acid was added and the programwas run: ramp time10 min, hold time 5 min at 100 psi. The digestionwas carried out in a microwave oven (CEM Corporation, MARS 5).Analyses of metal concentrations (Al, Cd, Co, Cr, Cu, Fe, Hg, Mn, Ni, Pb,Sn, V, and Zn) were carried out by inductive coupled plasma opticalemission spectrometry (ICP – OES, Perkin Elmer Optima 2000 DV). Alltrace metal determinations were veri 󿬁 ed by analyses of certi 󿬁 edreference materials (BCR, 277-Estuarine sediment) whose measure-ment accuracy with regards to the certi 󿬁 ed value ranged between 82and 97%. The results were expressed as total concentration in µg g − 1 dry weight. Natural variability in trace metal concentrations and grainsize can cause confusion in data interpretation regarding anthropo-genic loadings; a widely used procedure to compensate for thisdifference and to obtain comparable results consists in normalizingthe concentration of metal based on levels of elements that are notcommonlyassociatedtoanthropogenicinputssuchaslithium(Loring,1990) and aluminium (Windom et al., 1989). Both elements showed similar correlation with analysed trace metals (Table 1); however,lithium was used as a normalizer because it showed to be the mostadequate element in a nearby zone with similar sediment character-istics (Campana, 2006). All measurements were carried out in alaboratory at room temperature. Although there is an importance of polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocar-bons (PAHs) as potentially toxic compounds, there is no evidence of contamination by these compounds in the Bay of Cádiz (Riba et al.,2004; Martín-Díaz et al., 2008), thus they were not measured.  2.3. Treatment with liquid nitrogen In the laboratory, the collected samples were placed in a plastictray, forming a  󿬁 ne layer of sediment, on which liquid nitrogen( − 196 °C) was poured. Once the liquid nitrogen evaporated, thesamples were left to thaw at room temperature (approx. 20 °C), 832  C.V.M. Araújo et al. / Environment International 35 (2009) 831 – 841  during 3 to 4 h. After thawing the samples were immediately used inthe ecotoxicological assays. Pore water samples (obtained bycentrifugation at 4 °C, and 2719 ×  g   during 5 min) were taken beforeand after immersion in liquid nitrogen to assess possible changes inthe toxicological characteristics due to ultra-freezing. Sedimentcharacterization was also carried out before and after immersion inliquidnitrogen.Ultra-freezingwasnotusedasastoragemethodbutasa procedure to eliminate the autochthonous biota.  2.4. Toxicity assays with C. closterium 2.4.1. Sediment assays A strain of   C. closterium  (Ehrenberg) Lewin & Reimann (formerly Nitzschia closterium  (Ehrenberg) W. Smith) was isolated in May 2000fromasaltpondsituatedneartheAndalusiaInstituteofMarineScience(ICMAN – CSIC) in Puerto Real (SW of Spain). This strain is currentlyincluded in the ICMAN's Culture Collection of Marine Microalgae(CCMM – ICMAN, BIOCISE). For routine cultures, seawater  󿬁 lteredthrough a GF/C, Whatman  󿬁 lter, autoclaved, and later enriched withGuillard's f/2 medium (Guillard and Ryther,1962) and 50 mg L  − 1 SiO 2 was used. Arti 󿬁 cial substitute ocean water (ASTM, 1975) with SiO 2 (50mgL  − 1 ),NO 3 − (150mgL  − 1 ),andPO 43 − (10mgL  − 1 )supplementedas nutrients was used in all toxicity assays. This arti 󿬁 cial seawater wasemployed to avoid the effect of the chelator EDTA present in theGuillard'sf/2formulation.Routinecultureandassayswereperformedat20±1 °Cundercontinuous white light (35.2±1.1 µmol m − 2 s − 1 ) in acontrolled culture chamber (Ibercex). Three-day algal cultures Fig. 2.  Box drag used for sampling.  Table 1 Correlation matrix of aluminum and lithium concentrations with trace metals.Al Cd Co Cr Cu Fe Hg Mn Ni Pb Sn V ZnAl  −  0.96 0.89 0.61 0.43 0.97 0.63 0.82 0.82 0.30 0.15 0.97 0.84Li 0.99 0.90 0.88 0.59 0.40 0.95 0.60 0.79 0.79 0.25 0.11 0.96 0.82 Fig. 1.  Study site with sampled stations named as S01 to S10.833 C.V.M. Araújo et al. / Environment International 35 (2009) 831 – 841  (exponential growth phase) were used for the inoculation of sedimentsamples.Algalinoculumwasobtainedbycentrifugationduring4minat4123×  g  andre-suspendedinacleanmedium(arti 󿬁 cialsubstituteoceanwater)beforethestartoftheassays.Thisprocedurewasrepeatedthreetimes. Initial cellular concentration of 10 4 cells mL  − 1 was adopted forgrowth-inhibitionassay(OECD,1998).Assayprocedurewasfollowedasdescribed by Moreno-Garrido et al. (2003b) with some changes. Erlenmeyer  󿬂 ask (125 mL capacity) assays were used, topped withsynthetic cotton (Perlon). Each  󿬂 ask, containing 2 g (dry weight) of sediment, was  󿬁 lled up to 50 mL with the above mentioned arti 󿬁 cialseawatersupplementedwithnutrients.Thevolume(inmL)of2gofdryweight was previously calculated for each sediment sample. Thisprocedure was established to obtain a standard  󿬁 nal volume in each 󿬂 ask because each sediment sample presented a different density.Sedimentandoverlyingwater(arti 󿬁 cialseawater)wereplacedin 󿬂 asks24 h prior to inoculation of microalgae in order to allow the samples toequilibrate. Acid-washed sand obtained from a non-polluted area wasground and used as negative control sediment. Arti 󿬁 cial seawatersupplementedwithnutrientswasusedascontrol.Flaskswereincubatedunderthesameconditionsastheculturementionedabove.Attheendof the assay (72 h) cells were sampled after vigorous hand agitation toprevent settling, and counted in a Neubauer chamber, using a lightmicroscope (Olympus BH-2). Due to the dif  󿬁 culty in discriminatingbetweencellsandsedimentparticles, 󿬂 uorescenceemissionwasusedinmicroscopycountingthroughablueexcitationlightandabarrier 󿬁 lterof 530 nm resulting in an bright-red  󿬂 uorescence emitted by cellchloroplasts. Percentage of response (after 72 h) was calculated foreach sediment sample based on the following formula:Inhibition  k ð Þ  =  Dc  − DsDc  − Di    × 100  ð 1 Þ where  Dc   is the  󿬁 nal cell number in the control sediment (arti 󿬁 cialsediment  –  negative control),  Ds  is the  󿬁 nal cell number in thesediment samples, and  Di  is the initial cell number. Negative andpositive values indicate respectively stimulation and inhibitionregarding the negative control. During all assays the initial and  󿬁 nalpH values were registered in all the samples (Crison, MicropH 2001,electrode Hamilton, Slimtrode). Forall bioassays, the pH drift was lessthan 0.8 pH unit over 72 h.  2.4.2. Pore water assays Pore water samples were obtained foreach sedimentbycentrifuga-tion at 4 °C, and 2719 ×  g   during 5 min. This procedure was carried outwith natural samples (without treatment), and after ultra-freezing( − 196°C).Forstation7,collectedvolumeofporewaterwasnotenoughto carry out the assay due to the sandy particle-size distribution of thesedimentinthatlocation.Allsampleswere 󿬁 lteredbeforetheassaysby0.45 µm. Assays with pore water were carried out in a microplate(Greiner96-well 󿬂 atbottomwhite,12×8,400µLcapacity)accordingtoAraújo et al. (2008). Each well was  󿬁 lled with 300 µL of the sample.Sampleswerenotdiluted.Control(arti 󿬁 cialseawater)usedinthisassaypresented similar nutrient levels to the mean nutrient concentrationfound in the sample pore water (2 mg SiO 2  L  − 1 , 1 mg NO 3 − L  − 1 , and0.2 mg PO 43 − L  − 1 ). This procedure was adopted because a highernutrientconcentrationincontrolwouldhaveresultedinhighergrowth,and sample toxicity could have been overestimated. During incubationmicroplates were not shaken and they were closed non-hermeticallywith their transparent covers. Five replicates for each sample, controland blank (only culture medium) were used. To avoid possible edgeeffects of circumferential wells only centrally-located wells were used.To reduce evaporation, wells along the margin of the microplate were 󿬁 lled with Milli-Q water (Lukavský, 1992). Samples were randomlydistributedinthemicroplateinordertoavoidpseudoreplicationeffectsdue to possible spatial differences in illumination and temperature.Fluorescence values were compared with the mean value of   󿬂 uores-cence activity of algae incubated in the culture medium, considered as100%activity.Measurementswereperformedat0,24,48,and72hbyamicroplatereaderwith 󿬂 uorescencedetector(TECANMultiplateReader,GENios). Fluorometric data were obtained with the following settings:excitation wavelength: 485 nm and emission wavelength: 680 nm.Valuesof chlorophyll 󿬂 uorescencewere usedtocalculatetheinhibitionpercentages,whichwerecalculatedbycomparisonoftheareaunderthecontrolpopulation(100%)andthesample 󿬂 uorescencecurves(Hampelet al., 2001) obtained every 24 h throughout a 72 h period. Microalgastrain used in these assays was the same used during the sedimentassays. Both assays with sediment and pore water were carried outsimultaneously. During the assays no tests using reference substanceswerecarried outconcomitantly; nonetheless,beforeeachexperimentabatteryofassaysinmicroplateandErlenmeyer 󿬂 askswasperformedtoattestthehighreproducibilityandreplicabilityofthemethod.Moreover,thecellgrowthincontrolwasmonitoredinordertoverifythatthe 󿬁 nalcell number was at least 16 times higher than the initial cellconcentration. Finally, the variability of chlorophyll  󿬂 uorescenceobserved in the control treatment was always lower than 10%.  2.5. Toxicity assays with T. battagliai 2.5.1. Sediment assaysT. battagliai  Volkmman – Rocco had been cultured at the AndalusiaInstitute of Marine Science (ICMAN – CSIC) over several months. Thecopepods were cultured in 0.7 µm  󿬁 ltered natural sea water collectedfromtheRíoSanPedro,inPuertoReal(Spain).5Lvesselswereused,andanimals were maintained at 20±1 °C in 16/8 h light – darkness cycle,with weekly water renewal (ISO,1999). A living diet based on cells of  Isochrysis  aff.  galbana  (Primnesiophyceae) was provided twice perweek in all culture vessels (Williams and Jones,1994). Bioassays wereconducted on adult females carrying egg-sacs, based on Diz et al.(in press), although adapted to allow the development of the whole-sediment assays. Thus, toxicity assays were conducted in 12-wellsuspension culture-plates (17.8/16 MM and 127.8/85/19 MMCELLSTAR, Greiner Bio-One). Sediment samples (0.4 g of dry weight)were placed in each well (after ultra-freezing with liquid nitrogen),which were then  󿬁 lled with 4 mL of arti 󿬁 cial seawater preparedaccording to ASTM (1975). Microplates were prepared 1 day beforestarting the assay, and were maintained overnight in a culture room at20±1°C,withoutaeration,andwithaphotoperiodof16/8hlight/dark.Thesameconditionswerekeptthroughoutthedurationoftheassay.Tenreplicates (ten wells) were used for each sediment sample; they wererandomly distributed in microplates containing one adult female each.Additionally, ten replicates were prepared with arti 󿬁 cial sediment(control sediment). Wells were checked after 72 h using a dark 󿬁 eldbinocularmicroscope(LeicaGZ610×/22).Mortalityandthenumberof spawningfemalesweredetermined.Establishedmortalitycriterionwasabsence of movement after slight disturbing with a pipette. Spawningfemales were observed in order to con 󿬁 rm that over 80% of females incontrol sedimenthad spawned atthis timeof the assay. Oncemortalitywasveri 󿬁 ed,sampleswerestainedand 󿬁 xedwith10%Rose-Bengal(thisconcentration was previously assessed for sediment assays in ourlaboratory) in 10% buffered formaldehyde. Nauplii were counted usingthe same dark 󿬁 eld binocular microscope in order to determine thefecundity. Fecundity inhibition was the endpoint adopted, and itsinhibition percentage was calculated based on the number of naupliiproduced by females exposed to control sediment (considered as 100%fecundity).  2.5.2. Pore water assays Porewater was obtained, as mentionedabove, bycentrifugation of thesediment(beforebeingultra-frozen),anditwas 󿬁 lteredthrougha0.42 µm  󿬁 lter before being used in the assays in order to eliminatesuspended particles. The methodology followed was as described by 834  C.V.M. Araújo et al. / Environment International 35 (2009) 831 – 841  Diz et al. (in press), with the only difference that females with egg-sacs were selected, and not only those that carried the  󿬁 rst egg-sac.Thismodi 󿬁 cationwasadoptedbecausetheclutchsizesinharpacticoidcopepods are usually similar (Hart,1990). Ten females were exposedto each pore water sample (in ten wells); a control (ten replicates)with arti 󿬁 cial seawater (ASTM,1975) was also used. At the end of theassay (72 h) samples were  󿬁 xed in 10% buffered formalin and stainedwith 10% Bengal-Rose to count nauplii. Fecundity inhibition percen-tage was also calculated in all the pore water assays.  2.6. Multivariate analyses Two multivariate techniques, principal component analysis (PCA)and cluster analysis, were simultaneously employed. PCA was used tosimplify the data interpretation, since it reduces the entire data set toprincipal components (factors) which represent axes of majorvariation (McPherson et al., 2008; Morales-Caselles et al., 2007). Principal component is an axis that maximises the variances betweendata set, and so ordinates the variables based on the natural distancebetween two points in space (Euclidean distance). The parametersused to calculate the similarity matrix were: sediment pH and redoxpotential after ultra-freezing, sediment toxicity data for  C  .  closterium and  T  .  battagliai , metal concentration in ultra-frozen sediment,granulometry (% of silt and sand), organic matter and total organiccarboncontentin theultra-frozen sediment.Alldatawerenormalisedso that they had comparable scales. Three principal components wereestablished, and varimax rotation was used to maximize the sum of the variance. Analyses were performed with the statistical computersoftware package SPSS version 15.0.Cluster analysis (or classi 󿬁 cation) was employed to group stationsthat presented similar characteristics. For the clustering technique, thesimilarity between samples was represented by a dendrogram (hier-archical agglomerative method) based on the Euclidean distance.Parameters used to calculate the similarity matrix were the same onesusedforthePCAordination.Alldataweretransformedusinglog(1+  y ).AnalyseswereperformedwiththecomputersoftwarepackagePRIMER (Plymouth Routines in Multivariate Ecological Research) (Clarke andWarwick,1994).Another similarity matrix was generated from the ecotoxicogicaldatabasedontheEuclideandistancethatallowedanewordinationof the stations. Similar ordination was generated only with environ-mental variables. Afterwards, both matrices were superposed usingBIOENVAnalysis(basedonSpearmanrankcorrelation)toselectthoseenvironmental variables that best explain the stations pattern with  Table 2 Values of pH and redox potential (Eh) in pore water and sediment, and total organic carbon (TOC), organic matter (OM), and sand and silt percentage in sediment before and afterultra-freezing.Stations Pore water SedimentpH Eh (mV) pH Eh (mV) TOC (%) OM (%) Sand (%) Silt (%)N UF N UF N UF N UF N UF UF UF UFS01 7.77 7.79  − 62  − 63 7.58 7.33  − 50  − 37 4.04 4.22 10.60 4.4 95.6S02 7.96 7.80  − 73  − 63 6.95 6.89  − 15  − 11 2.99 2.90 11.59 41.4 58.6S03 7.92 7.74  − 69  − 60 7.19 7.10  − 19  − 24 0.87 0.70 1.96 75.4 24.6S04 7.60 7.68  − 52  − 57 7.89 7.48  − 68  − 45 1.02 1.22 2.56 76.4 23.6S05 7.78 7.85  − 62  − 66 7.23 7.35  − 30  − 38 3.53 3.12 8.89 92.6 7.4S06 7.91 7.84  − 70  − 66 7.23 7.26  − 31  − 33 0.84 0.95 1.59 82.9 17.1S07 7.66  − − 56  −  7.12 7.29  − 25  − 35  −  0.42 0.62 100.0 0.0S08 7.88 7.68  − 69  − 57 7.00 7.12  − 17  − 25 1.86 1.75 4.00 32.1 67.9S09 7.92 8.01  − 70  − 75  −  6.92  − − 14 1.90  −  5.28 10.7 89.3S10 7.93 7.73  − 71  − 60 7.18 7.20  − 28  − 30 1.62 1.69 4.39 55.0 44.0Control  − − − − − − − − −  0.27 0.39 93.1 6.9N: natural sediment (before ultra-freezing); UF: after ultra-freezing.  Table 3 Total metal concentration in sediment samples before (N) and after (UF) ultra-freezing (values in µg g − 1 ).Stations Li Cu Zn Fe Co Ni Mn V Al Cr Hg Cd Pb SnS01 N 88 47 150 34000 15 38 300 130 67000 98 0.58 0.56 49 5.4UF 92 49 150 35000 16 40 330 140 71000 100 0.58 0.60 50 5.6S02 N 45 130 170 29000 16 46 320 100 40000 180 0.97 0.47 230 68UF 45 140 170 28000 16 47 310 100 40000 190 1.10 0.40 220 73S03 N 24 20 56 9000 5.6 16 130 27 24000 26 0.20 0.18 24 2.5UF 23 18 50 8300 5.2 16 120 25 23000 22 0.21 0.18 30 2.7S04 N 21 9.5 41 7900 5.7 18 190 29 21000 25 0.11 0.14 16 1.4UF 19 8.5 35 7700 6.0 16 200 27 20000 24 0.10 0.17 16 1.4S05 N 72 38 120 31000 13 34 260 110 58000 83 0.44 0.47 47 4.6UF 74 38 120 31000 13 34 240 110 61000 84 0.42 0.48 46 4.8S06 N 17 8.2 33 6400 5.5 16 110 25 15000 23 0.09 0.23 12 1.0UF 15 7.7 33 6100 5.2 17 91 23 14000 24 0.07 0.16 11 1.1S07 N  − − − − − − − − − − − − − − UF 10 2.3 12 2000 3.1 13 64 6.3 7900 8.0 0.11 0.07 5.5 0.40S08 N 51 27 87 22000 11 31 300 79 44000 62 0.37 0.45 30 3.2UF 52 28 89 23000 11 30 320 82 45000 65 0.41 0.41 33 3.3S09 N 68 33 95 28000 12 30 450 99 56000 72 0.48 0.43 36 4.0UF  − − − − − − − − − − − − − − S10 N 38 22 70 17000 8.3 23 190 56 34000 43 0.37 0.32 32 3.0UF 39 22 71 17000 8.6 26 200 58 33000 45 0.39 0.31 32 2.9Control UF 8.2 1.6 12 2600 2.8 12 86 8.5 51000 8.0 0.03 0.07 4.2 0.2835 C.V.M. Araújo et al. / Environment International 35 (2009) 831 – 841
Related Search
Similar documents
View more...
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