S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 8 ) 2 4 2 – 2 4 9

ava i l ab l e a t www.sc i enced i rec t . com

www.e l sev i e r. com/ loca te / sc i to tenv

Assessing the effects of the Prestige oil spill on the Europeanshag (Phalacrocorax aristotelis): Trace elements andstable isotopes

Carola Sanperaa,⁎, Sonia Valladaresa, Rocío Morenoa, Xavier Ruiza,1, Lluis Joverb

aDept. Biologia Animal, Facultat de Biologia, Universitat de Barcelona, Diagonal 645, 08028 Barcelona, SpainbDept. Salut Pública, Facultat de Medicina, Universitat de Barcelona, Casanova 143, 08036 Barcelona, Spain

A R T I C L E D A T A

⁎ Corresponding author. Departament de BioBarcelona, Spain. Tel.: +34 934021452; fax: +3

E-mail address: csanpera@ub.edu (C. Sanpe1 Xavier Ruiz passed away on 27th April 20

0048-9697/$ – see front matter © 2008 Elsevidoi:10.1016/j.scitotenv.2008.07.052

A B S T R A C T

Article history:Received 26 February 2008Received in revised form 24 July 2008Accepted 28 July 2008Available online 18 September 2008

ThePrestigeoil spill resulted in themortalityof several seabird species on theAtlanticNWcoastof Spain. Shag casualties were particularly relevant, since populations are resident in the areathe whole year round and because of several features which make them highly vulnerable toenvironmental hazards. Ecological catastrophes give us the opportunity of collecting sampleswhich, otherwise, would be difficult to obtain. We examine the potential of shag corpses asbioindicators of inorganic pollution and the possible factors of variability, such as biologicaltraits (sex, age) or nutrition status. We determined trace elements (Hg, Se, Cr, Pb, Zn and Cu)and isotopic signatures (15N, 13C) in soft tissues (muscle, liver) and in primary feathers formedatdifferent times (before andafter thePrestige) in individuals of knownsexandage, collected atthe time of the Prestige disaster. These were compared with data from another group of shagstrapped accidentally in fishing gear in 2005.Our results did not seem to be affected by sex or age on any of the analysed variables. Thehighernitrogen isotopic signatures in the soft tissues of the Prestige shagsmay be related to thenutritionstresscausedbyapoorerbody condition,which isalso reflected in increasing levels ofsomemetals in the liver.This isotopic enrichmentwasalsoobserved innewly forming featherswhen compared to the old ones. On the other hand, the lower δ15N and Hg values in shagfeathers from2005point toa shift in feeding resources topreyof lower trophic levels.We foundthat feather features (being an inert tissue and having a conservative composition), ifcombined with careful dating and chemical analysis, offer a very useful tool to evaluatetemporal and spatial changes in seabird ecology in relation to pollution events.

© 2008 Elsevier B.V. All rights reserved.

Keywords:ShagOil spillTrace elementStable isotopesPollution

1. Introduction

In November 2002, the tanker Prestige was wrecked off theAtlantic north-west coast of Spain (Galicia) and spilledapproximately 60,000 tm of oil. There were two major oilspills and the western Galician coastline was severely affectedby fuel contamination. As a consequence of the catastrophe,

logia Animal, Facultat d4 934034476.ra).08.

er B.V. All rights reserved

thousands of seabirds of different species died and many ofthem were washed ashore. The protected area of the ‘ParqueNacional Marítimo-Terrestre de las Illas Atlánticas’, made upof several coastal archipelagos of small islands located at themouth of the rias, was also affected (Fig. 1). This area holds themost important Iberian breeding colonies of the Europeanshag (Phalacrocorax aristotelis).

e Biologia, Universitat de Barcelona, Avgda, Diagonal 645, 08028

.

Fig. 1 –Sampling area. Cíes and Ons Islands located at the mouth of the ‘Ría de Vigo’ and ‘Ría de Pontevedra’, respectively.

243S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 8 ) 2 4 2 – 2 4 9

Most of thedead seabirdswereAlcidaewintering in thearea(mainly auks, guillemots…) which originate from a number oflarge breeding colonies to the north. Therefore oil-relatedmortalitymay be thinly spread among these colonies reducingthe impact of mass-mortality (but see Votier et al., 2005).However European shags have a high oil vulnerability index(Williams et al., 1995) because of several characteristics ofpopulations in the area affected by the Prestige. Firstly, theNWAtlantic coast of Spain constitutes the southern limit for thisspecies. Galicia holds approximately 86% of the Iberian shagpopulation (ca. 2000 pairs), which is resident in the area thewhole year round. Secondly, most of them (85%) are concen-trated in just two breeding colonies, in the Illas Cíes and IllasOns (Parque Nacional de Illas Atlánticas, see Fig. 1); thisdistribution makes these populations highly sensitive to anyenvironmental hazard, such as oil spills (Velando and Alvarez,2004). Finally, the Galician shag populations were reported tobe slightly declining prior to the oil spill (Velando and Freire,2002) and a 50% decrease in breeding success was observed forthe oiled colonies of Illas Cíes andOnsduring the 2003breedingseason (Velando et al., 2005).

The Prestige transported heavy fuel oil (type 6, UK classifi-cation) which as well as containing several hydrocarbons,included high amounts of trace elements, mainly Fe, Ni, V, As,Co, Cr, Cu, Se, Zn (CSIC, 2003a). An increase of some heavymetals in the water column was detected following thedisaster (CSIC, 2003b) and also in surface waters around theIllas Cíes and Ons (Santos-Echeandía et al., 2005).

Seabirdmassmortalities resulting from ecological disasters,such as oil spills, represent an excellent opportunity to collecthuge number of species' samples that are not commonlyavailable, although the use of beached bird surveys as amonitoring instrument has been hotly debated (Furness and

Camphuysen, 1997). The carcasses of oiled birds may be usefulfor carrying out pollutant studies, as long as pathologicalfeatures of dead birds are carefully evaluated (Wenzel andAdelung, 1996; Debacker et al., 2000, 2001a; Mochizuki et al.,2002). Some tissues, such as feathers, allow retrospectivestudies formonitoring heavymetal pollution by lead, cadmiumand many other elements. Metals can be deposited onto thesurface of feathers from the atmosphere, or incorporated intogrowing feathers from blood. For the Prestige, Balseiro et al.(2005) reported that although some birds died when treated atrescue centres,mostdid so because of the ingestionofoil,whichresults in severe impairment of a bird's health. Previousresearch analysing trace metal levels in tissues of seabirdsfrom the Prestige period did not observe any increase in thelevels of inorganic elements as a consequence of the spilled oilwhen comparing beached birds from different periods ordifferent areas (Pérez López et al., 2005, 2006; Carbonell et al.,2006; Gallego Rodriguez et al., 2007).

In this study, we have combined trace element analysiswith that of stable isotopes in several tissues (feather, muscle,liver) of shag carcasses collected at the time of the Prestigeaccident and we have compared them with data from freshsamples from a group of healthy shags, which died in 2005after being trapped accidentally in fishing gear in the samearea. We wanted to determine the impact of Prestige oil ontrace element concentrations of soft tissues (muscle, liver) andhow these were related to biological variables (gender, age),trophic status or nutritional stress. Nitrogen and carbonisotopic signatures could provide information not only aboutthe trophic ecology of sample birds, but also on the nutritionalstress they might have suffered previously, since the ratio15N/14N is known to increase in relation to severe starvationperiods (Hobson et al., 1993; Cherel et al., 2005). Because

244 S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 8 ) 2 4 2 – 2 4 9

feathers remain unchanged once formed, we collected feath-ers from the same individuals, but which had grown atdifferent times (before, during and after the spill) in 2002–03,and compared to those of 2005, to track the changes ininorganic pollutants and stable isotope values to get cluesabout shag trophic ecology during this period and to infer thesignificance of the changes observed in soft tissues.

2. Materials and methods

The Autonomous Government (Conselleria de MedioAmbiente, Xunta de Galicia) organised the collection ofseabird carcasses beached along the coast and took them tothe Wildlife Recovery Centre at Cotorredondo (Pontevedra),where the material was kept frozen until tissue analysis.

We sampled 58 carcasses of European shags collected deadbetween November 2002 and April 2003 in the Rias Baixas offGalicia (NW Spain) (see Fig. 1). They were aged by plumagemorphology and included 32 juveniles and 26 adults. Duringthe necropsy, pectoral muscle and liver samples weredissected and animals were sexed by gonadal inspection,resulting in 31 females and 22males; in 5 individuals sex couldnot be determined because of the degree of decomposition(Martinez-Abrain, personal communication). In those animalswithout severe oiling, the whole wing and some scapularmantle feathers were sampled. Following the moultingpattern described for this species (Potts, 1971; Cramp andSimmons, 1977) we assigned moulting scores to each primaryfeather according to Gargallo (2000). Because the 7th primaryfeather (P7) had already moulted in all the birds, weconsidered them to be ‘control’ feathers, which were formedbefore the Prestige oil spill. Eighteen individuals had primaryfeathers which were growing at the time of the Prestigedisaster (Moulting primaries, ‘MP’) and these were consideredto be affected by changes which occurred during the spill.

The same samples were collected in 8 juvenile shags fromthe same area (Illa de Ons, Parque Nacional Illas Atlánticas)found dead in June 2005 after being trapped in fishing gear. Allsamples were kept frozen at −20 °C until analysis.

3. SIA analysis

Muscle and liver samples were lyophilized and lipid extractedfollowing the Folch method (Folch et al., 1957) in order tominimize the differences in δ13C caused by the variable contentof tissue lipids (Hobson andClark, 1992). Featherswere rinsed inNaOH (0.25 M) and oven dried (T=60 °C). All samples werehomogenized to an extremely fine powder using an impactormill (Freezer Mill 6850, Spex CertiPrepH Inc., Metuchen, NewJersey) operating at liquid nitrogen temperature.

Weighed subsamples (0.7 mg feathers; 0.2 mg muscle/liver)were placed into tin buckets and crimped for combustion. N andC isotopic analyses were carried out by EA-IRMS (elementalanalysis–isotope ratio mass spectrometry) by means of aThermo-Finnigan Flash 1112 elemental analyzer coupled to aDelta isotope ratiomass spectrometer via a CONFLOIII interface(Serveis Científico-Tècnics, University of Barcelona).

Stable isotope ratios were expressed in conventional deltanotation as parts per thousand (‰) according to the followingequation: δX=[(Rsample/Rstandard)−1]×1000 where X is 15N or13C and R is the corresponding ratio 15N/14N or 13C/12C.

The international standards for N and C are atmosphericnitrogen (AIR) and Peedee Belemnite (PDB) respectively.Accuracy was ≤0.1‰ for δ13C and ≤0.3‰ for δ15N.

4. Trace element analysis

Chemical determination of trace elements (Se, Hg, Pb, Cr, Cu,Zn) was carried out by means of ICP-OES, Perkin-Elmer Elan6000 (Serveis Científico-Tècnics, University of Barcelona).Feather and muscle samples (ca. 100 mg) were digested inTeflonTM containers using HNO3 (1–2 ml) and H2O2 (0.5–1 ml)for 14 h at 90 °C. Liver samples (ca. 200 mg) were digested in amicrowave oven (Milestone Ethos Plus) using HNO3 (4 ml) andH2O2 (4 ml) for 90 min at 210 °C. All concentrations areexpressed in μg g−1 on a dry weight basis.

Accuracy of analysis was checked by measuring certifiedreference tissue: Human Hair (BCR 397) for feathers,and Dogfish Liver (DOLT-3) and Lobster Hepatopancreas(TORT-2) for muscle and liver. Mean recoveries ranged from92% to 105% for Se, 80–92% for Hg, 96–119% for Pb, 82–115% forCu and 91–114% for Zn, and no corrections were made.

5. Statistical methods

Values of metal concentrations and stable isotope ratios wereroutinely checked for normal distributions using Q–Q plots.Only trace metal concentrations showed skewed distributionswhich were normalized by applying a logarithmic transforma-tion. Comparisons between sexes, adult vs. juveniles orbetween two sampled periods were made using the t-teststatistic for independent samples on the original units forstable isotopes or using the log-transformed values for metalconcentrations. Paired differences between tissues (liver vs.muscle) or between feathers (7th primary vs. moultingprimary) were normally distributed and a paired t-test wasusedon the original values. Pearson correlation coefficientwasused to assess the relationship between tissues or betweenfeathers using log transformed values of metal concentrationsand original ‰ isotope ratios. Unadjusted p-values werereported in tables and text, but sequential Bonferroni correc-tion, using Holm's procedure, was applied to each group oftests (comprising between 8 and 18 individual tests) tomaintain the table-wise type I error rates at 5%.

SPSS v15.0 statistical package was used to carry out dataanalysis.

6. Results

6.1. Muscle and liver tissues

Descriptive statistics for trace elements and stable isotopesignatures analyzed in both tissues are shown in Table 1. Lead

Table 1 – Descriptive statistics for metal concentrations (μg g−1 dw) and stable isotope signatures (‰) of liver andmuscle arepresented according to sampling period and age class

Se Hg Cu Zn Pb Cr δ15N δ13C

PrestigeAdults Liver Mean 11.54 28.83 31.92 249.26 15.13 −15.50

Median 10.72 22.88 29.53 220.38 15.15 −15.48IQR (S.D.) 5.53 17.84 20.40 242.92 (0.57) (0.72)N (detected) (25) (26) (26) (26) (0) (0) 26 26

Muscle Mean 2.64 4.19 24.25 78.55 0.25 1.71 13.29 −15.76Median 2.63 4.24 23.04 81.42 0.12 1.68 13.22 −15.72IQR (S.D.) 1.25 2.52 6.42 34.27 0.40 0.16 (0.60) (0.50)N (detected) (26) (26) (26) (26) (3) (25) 26 26

Paired t-test 13.34 7.83 3.24 6.95 19.03 2.80p-value b0.001⁎ b0.001⁎ 0.003⁎ b0.001⁎ b0.001⁎ 0.009⁎

Pearson r 0.41 0.49 0.34 0.84 0.64 0.77p-value 0.04 0.01 0.09 b0.001⁎ b0.001⁎ b0.001⁎

PrestigeJuveniles Liver Mean 8.43 19.90 40.96 221.12 14.99 −15.26

Median 7.66 18.60 32.86 210.54 14.93 −15.28IQR (S.D.) 4.82 12.84 31.72 205.35 (0.67) (0.61)N (detected) (31) (32) (32) (32) (0) (0) 32 32

Muscle Mean 2.41 4.63 23.73 76.31 0.16 1.76 13.36 −15.63Median 2.05 4.65 23.15 67.38 0.15 1.72 13.25 −15.57IQR (S.D.) 0.91 2.78 8.17 46.70 0.04 0.20 (0.68) (0.54)N (detected) (32) (32) (32) (32) (5) (32) 32 32

Paired t-test 12.09 9.51 4.49 6.66 17.913 4.323p-value b0.001⁎ b0.001⁎ b0.001⁎ b0.001⁎ b0.001⁎ b0.001⁎

Pearson r 0.64 0.83 0.54 0.85 0.71 0.64p-value b0.001⁎ b0.001⁎ 0.001⁎ b0.001⁎ b0.001⁎ b0.001⁎

2005Juveniles Liver Mean 5.27 5.36 23.17 125.17 14.03 −16.34

Median 5.08 3.76 23.09 102.45 13.92 −16.33IQR (S.D.) 0.55 3.79 7.10 13.01 (0.38) (0.29)N (detected) (7) (8) (8) (8) (0) (0) 8 8

Muscle Mean 1.71 2.04 18.73 62.59 0.14 1.65 12.58 −16.68Median 1.47 1.94 18.36 54.31 0.14 1.62 12.49 −16.65IQR (S.D.) 0.29 3.13 2.60 13.06 0.03 0.17 (0.33) (0.22)N (detected) (8) (4) (8) (8) (2) (8) 8 8

Paired t-test 27.87 3.06 2.82 2.68 11.777 3.641p-value b0.001⁎ 0.055 0.026 0.031 b0.001⁎ 0.008⁎

Pearson r 0.89 0.93 0.55 0.87 0.52 0.18p-value 0.007 0.072 0.16 0.005⁎ 0.18 0.19

As a spread measure, interquartile range (IQR) was used for skewed distributions and standard deviation (S.D.) for stable isotopes. Paired t-testand Pearson correlation coefficient between values of tissues are given. Test p-values with ⁎ mark denote that they are significant at table-wiseerror type I of 5%.

245S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 8 ) 2 4 2 – 2 4 9

in both tissues as well as chromium in the liver were belowthe limit of detection in most samples (LOD: Pbb10 ppb,Crb5 ppb).

No significant differences betweenmales and femaleswerefound in the adult and juvenile shags of 2002–03, either intrace metal levels or in isotopic signatures.

When comparing age groups, differences were significantonly for hepatic selenium of adults, which was shown to behigher (Seliver: t-test=3.67, p=0.0006). Moreover, no differenceswere found in stable isotope signatures between juveniles andadults.

With few exceptions, when comparing hepatic and muscletrace element concentrations, levels in the liverwere shown tobe significantly higher. In most cases, concentrations of trace

elements in both tissues were positively correlated (seeTable 1). Both in adults and juveniles of 2002–03, δ15N andδ13C values in the liver were higher than in the muscle, andsignatures between both tissues were positively correlated. Inthe 2005 group, liver isotopic signatures were higher, but noliver/muscle correlation was observed (Table 1).

When comparing juveniles from both time periods, Prestigeshags revealed higher values of selenium (t-test=4.63; pb0.001),mercury (t-test=5.52; pb0.001) and copper (t-test=3.36; p=0.002)in the liver, but not in themuscle (see Fig. 2). On the other hand,stable isotopes of nitrogen and carbon in both tissues weresignificantly higher in the Prestige shags (liver: δ15N t-test=3.9,δ13C t-test=4.89; muscle: δ15N t-test=4.65, δ13C t-test=8.47, allresults with pb0.001).

Fig. 2 –Graphic showing the values distribution of trace elements (in log scale) and stable isotopes in different tissues of juvenileshags from the two periods: Prestige (dark boxes) and year 2005 (light boxes). The length of the box is the interquartile range(IQR), and the line inside marks the medians. Values more than 1.5 IQR's from the ends of the box are labeled as outliers (o).

246 S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 8 ) 2 4 2 – 2 4 9

Table 2 – Descriptive statistics for metal concentrations (μg g−1 dw) and stable isotope signatures (‰) of 7th primary and amoulting primary are presented according to sampling period and age class

Se Hg Cu Zn Pb Cr δ15N δ13C

PrestigeAdults 7th primary Mean 1.84 6.41 10.30 108.19 0.61 3.35 14.78 −14.89

Median 1.63 6.31 10.45 107.19 0.35 2.38 14.86 −14.92IQR (S.D.) 1.19 1.81 2.20 14.15 0.20 1.68 (0.67) (0.61)N (detected) (6) (9) (9) (9) (9) (9) 9 9

Moulting Primary Mean 1.71 5.57 11.45 109.69 0.34 2.46 15.16 −15.08Median 1.36 5.73 10.34 105.88 0.16 2.39 15.20 −14.81IQR (S.D.) 1.55 2.32 1.63 9.27 0.37 0.46 (0.39) (0.68)N (detected) (4) (5) (5) (5) (5) (5) 5 5

PrestigeJuveniles 7th Primary Mean 2.27 10.73 10.04 115.86 0.60 2.92 14.51 −15.05

Median 2.23 12.84 10.35 113.04 0.31 2.51 14.48 −14.69IQR (S.D.) 1.18 9.25 3.43 25.80 0.69 0.61 (0.45) (0.87)N (detected) (17) (17) (19) (19) (18) (19) 19 19

Moulting primary Mean 1.85 6.97 11.03 111.68 1.16 2.87 15.94 −14.64Median 1.60 3.19 10.48 107.72 1.03 2.33 16.10 −14.45IQR (S.D.) 0.54 3.46 1.76 22.97 1.15 0.23 (0.58) (0.51)N (detected) (10) (13) (13) (13) (12) (13) 13 13

Paired t-test 1.17 1.28 −1.17 0.76 −2.38 0.38 −8.87 −0.34p-value 0.27 0.23 0.27 0.46 0.039 0.71 b0.001⁎ 0.74Pearson r −0.24 −0.69 −0.20 −0.45 −0.84 −0.20 0.53 0.24p-value 0.50 0.018 0.52 0.12 0.001⁎ 0.52 0.07 0.44

2005Juveniles 7th primary Mean 1.42 1.41 11.47 102.01 0.73 2.25 13.85 −15.62

Median 1.28 0.99 11.06 99.76 0.30 2.05 13.65 −15.73IQR (S.D.) 0.93 0.78 2.39 25.82 0.33 0.31 (0.59) (0.28)N (detected) (3) (8) (8) (8) (8) (8) 8 8

As a spreadmeasure, interquartile range (IQR) was used for skewed distributions and standard deviation (S.D.) for stable isotopes. Paired t-test andPearson correlation coefficient betweenvaluesof tissues are given.Testp-valueswith ⁎markdenote that theyare significant at tablewiseerror type Iof 5%. Differences were not tested in adults because of the small sample size. Juveniles from 2005 did not have moulting primary feathers.

247S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 8 ) 2 4 2 – 2 4 9

6.2. Feathers

Descriptive statistics for trace elements and stables isotopesin different kinds of feathers (7th primary ‘P7’, primarymoulting feather ‘MP’) are shown in Table 2.

Neither adult nor juvenile shags from 2002–2003 showedsignificant differences between sexes in trace element levelsor isotopic signatures of different feathers. There were nodifferences in relation to age, either.

Paired comparisons among ‘P7’ and ‘MP’ were carried outfor juveniles given that the small number of MP of adultsprecludes statistical testing. Excepting lead, which showed anegative correlation between the two types of feathers, nodifferences and no correlations were found for trace elements.A significant increase in nitrogen isotopic signatures ofmoulting primary feathers was observed.

When comparing juvenile ‘P7’ between periods we foundsignificant differences in Hg levels (t-test=5.01; pb0.001) andnitrogen signatures (t-test=3.18; p=0.004) both being higher inthe 2002–03 period (see Fig. 2).

7. Discussion

Oiled seabirds offer a unique opportunity to obtain a largenumber of samples from specieswhich are often not available.

This is the case of the European shag in Galicianwaters, whichpresents a clustered distribution, nesting on rather inacces-sible rocky shores and cliffs, andwhose populations are highlysensitive to human disturbance.

However, seabird mortalities may not be a representativesample of the whole population; some age classes or gendercould be biased because of certain life-history traits. Forexample, female and immature seabirds comprise a signifi-cantly greater proportion of casualties compared with maleand adult seabirds in the Prestige accident (Sociedade Galegade Ornitoloxia, 2005). In this study we found no significantdifferences in trace elements or isotope ratios of Shagcarcasses as a function of age and sex, indicating that suchdifferences are not a consistent feature of oil-relatedmortalityin seabirds (Wenzel and Adelung, 1996). Thus we suggest thatthe differential mortality of oiled seabirds associated with thePrestige spill is related to other factors, such as behaviouralsegregation or differences in habitat selection by different ageor sex groups during that time of the year. Martinez-Abrainet al. (2006) attributed the skew to the despotic model ofbreeding habitat selection by male shags and becauseimmature birds were not defending territories at colonies.

When comparing juvenile shags from the two time periods(2002–03 vs. 2005), differences were very obvious. Birds thatdied just after the Prestige accident presented higher hepaticlevels of most trace elements (Se, Hg and Cu); isotopic

248 S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 8 ) 2 4 2 – 2 4 9

signatures in both tissues were also higher. Our results onhepatic mercury are similar to those published by Carbonellet al. (2006) on shags collected after the Prestige accident.However, these authors did not find a significant decreasingtrend of liver Hg concentrations with time (2003 vs. 2005),because their sampling scheme considered only beached birdsprobably with a deficient nutritional status (see later). Forshags from the two periods, the levels of lead and chromiumwere low, in many cases under the limit of detection. The factthat not only those metals present in the oil spill (such as Cu),were found in higher concentrations in the affected animalsbut also other metals that were not reported in the oil (i.e. Hg),and conversely, that some of themetals present in the oil wereat low levels in the shag tissues (Cr), suggests that otherphysiological effects must play a part in explaining the highermetal load in oiled shags. Several studies have reported theinfluence of body condition on trace element levels in tissues.Marine birds affected by the Prestige's oil spill were reported tobe severely dehydrated and emaciated, indicating that theydied after suffering previous weakness (Balseiro et al., 2005).

Previous work has shown a tendency for a negativerelationship between trace elements and nutritional status.Debacker et al. (2001a,b) reported that an increase in heavymetal levels (Cu and Zn) in internal tissues of birds, both in thewild and in the laboratory, parallel to increasingly severecachexia, pointed to a general redistribution of heavy metalswithin the organs as a result of starvation and proteincatabolism (cachectic status). The increased levels of thesemetals, which are particularly toxic, could well represent anadditional pressure to birds already facing stressful conditions(Debacker et al., 2000).

Wenzel and Adelung (1996) suggested that oiled guillemots(Uria aalge) can be suitable monitoring organisms for traceelement contamination provided that birds of the same age ornutritional condition are compared. However, due to the factthat the oiled bird (sub)population is a non-random hetero-geneous sample, this requirement will be difficult to meet. Inthis situation, determining nitrogen isotopic signatures oftissues could give us information about nutritional stress,even if these signatures could be masked by ecologicalprocesses such as diet selection (Kempster et al., 2007;Williams et al., 2007). We found higher levels of δ15N andδ13C in both muscle and liver of juvenile shags of 2002–03,which is in agreement with the observed enrichment in tissue15N in experimental and field studies on starving birds(Hobson et al., 1993; Cherel et al., 2005). Although nutritionstress is thought to affect mainly nitrogen signatures, somestudies have shown a parallel increase in 13C (Hatch et al.,1995).

Analysis of feather composition is widely used to monitorcontaminants and stable isotope ratios (Furness and Green-wood, 1993; Thompson et al., 1998; Becker, 2003). Because thecomposition of feather keratin is fixed after it has grown,stable isotopes in feathers reflect the diet and the environ-mental conditions at that time. Thus, if the moult pattern andtime of feather formation are known, the information is time-specific. The 7th primary feathers (P7) of 2002–03 carcasseswere therefore formed before the accident, being indicative ofthe shag's foraging ecology in the preceding period. On theother hand, those primary feathers found in active moult,

were forming at the time of the animal's death, andpresumably under conditions of nutritional stress. Thisexplains the fact that in these last formed feathers (MP)nitrogen signature was enriched. As Cherel et al. (2005)reported for feathers of fasting King Penguins (Aptenodytespatagonicus), the increase of δ15N would point to differentpathways for resource allocation (endogenous vs. dietary) inorder to grow feathers.

When comparing P7 between juveniles from the two timeperiods, isotopic differences appear, values being higher in2002–03. This would suggest either that the shags have shiftedtheir diet in 2005 to preys of lower trophic level, or, that therehas been a general decrease in the signatures of theecosystem. Anyway, considering that Hg is the only metalwhich is biomagnified along the food web, its observeddecrease in P7 concentrations in the year 2005 would pointto the change in the trophic level of resources consumed.Nonetheless, we hope that further workwe are carrying out onshag chicks sampled on an annual basis will clarify the sourceof these variations.

The analysis of these tracers in feathers, if coupled to theknowledgeofmoultingpatterns, canprovidea reliabledatabaseto carry out spatial or temporal comparisons. For example,Burger et al (2007) compared feathers of Pigeon Guillemots(Cepphus columba) between pristine and oiled areas from theExxonValdez accident and found only slight differences for leadand cadmium. Our results reveal that sampling soft tissues forpollutants or stable isotopes, even if carcasses are fresh, canresult in severe bias because the individuals can be subject todifferent degrees of nutritional stress.

Acknowledgements

Thanks are given to the Conselleria de Medio Ambiente(Autonomous Government, ‘Xunta de Galicia’) and to theParque Nacional de las Illas Atlánticas de Galicia, for facilitat-ing the shag corpses. Cristobal Pérez (University of Vigo, Spain)helped us in the collection of the shags at Cotorredondo.Gabriel Gargallo (Institut Català d'Ornitologia, Barcelona)assessed us on moulting scores. Mike O'Neill made thelanguage revision of the manuscript. Rocío Moreno wassupported by a FPU grant (MEC, Spain). Funding for thiswork, as well as the work done by Sonia Valladares, wasprovided by project VEM2004-08524 from the Ministerio deEducación y Ciencia (Spain).

R E F E R E N C E S

Balseiro A, Espí A, Márquez I, Pérez V, Ferreras MC, Martín JF, et al.Pathological features in marine birds affected by the Prestige oilspill in the north of Spain. J Wildl Dis 2005;41(2):371–8.

Becker PH. Biomonitoring with birds. In: Markert, BA, Breure, AM,Zechmeister, HG, editors. Bioindicators & Biomonitors: Principle,Concepts and Applications. Elsevier; 2003. p. 677–736. ciutat.

Burger J, Gochfeld M, Sullivan K, Irons D. Mercury, arsenic,cadmium, chromium, lead and selenium in feathers of PigeonGuillemots (Cephus Columba) from Prince William Sound andthe Aleutian Islands of Alaska. Sci Total Environ2007;387:175–84.

249S C I E N C E O F T H E T O T A L E N V I R O N M E N T 4 0 7 ( 2 0 0 8 ) 2 4 2 – 2 4 9

Carbonell MG, Fernández Torija C, López Beceiro A, FidalgoÁlvarez LE, Hernández Moreno D, Soler Rodriguez F, et al.Contenido hepático demercurio y plomo en cormoránmoñudo(Phalacrocórax aristotelis) y alcatraz atlántico (Morusbassanus) procedentes de las costas de Galicia (España). RevToxicol 2006;23:95–100.

Cherel Y, Hobson KA, Bailleul F, Groscolas R. Nutrition, physiology,and stable isotopes: new information from fasting andmoltingpenguins. Ecology 2005;86:2881–8.

Cramp S, Simmons KEL. Handbook of the birds of Europe, theMiddle East and North Africa: the birds of the WesternPalearctic, vol.1. Oxford University Press; 1977. 722 pp.

CSIC. Caracterización del vertido y evolución preliminar en elmedio. CSIC Technical Report, Vol. 1; 2003a. http://csicprestige.iim.csic.es/Informes.htm.

CSIC. Presencia demetales pesados en la zona de hundimiento delpetrolero Prestige y composición de metales y complejantesdel fuel emulsionado de la costa., Vol. 2; 2003b. http://csicprestige.iim.csic.es/Informes.htm.

Debacker V, Jauniaux T, Coignoul F, Bouquegneau JM. Heavymetalcontamination and body condition of wintering guillemots(Uria aalge) at the Belgian coast from 1993 to 1998. Environ ResA 2000;84:310–7.

Debacker V, Schiettecatte LS, Jauniaux T, Bouquegneau JM.Influence of age, sex and body condition on zinc, copper,cadmium and metallothioneins in common guillemots (Uriaaalge) stranded at the Belgian coast. Mar Environ Res2001a;52:427–44.

Debacker V, Rutten A, Jauniaux T, Daemers C, Bouquegneau JM.Combined effects of experimental heavy-metal contamination(Cu, Zn, and CH3Hg) and starvation on quail's body condition.Biol Trace Elem Res 2001b;82:87–107.

Folch J, LeesM, Stanley GHS. A simplemethod for the isolation andpurification of total lipids from animal. J Biol Chem1957;226:497–509.

Furness RW, Greenwood JJD. Birds as monitors of environmentalchange. London: Chapman & Hall; 1993. 343 pp.

Furness RW, Camphuysen CJ. Seabirds as monitors of the marineenvironment. ICES J Mar Sci 1997;54:726–37.

Gallego Rodríguez ME, Hernandez Moreno D, Fidalgo Álvarez LE,Rigueiro L, López Becerro A, Soler Rodríguez F, et al. Niveles deplomo y cadmio en tejido hepático de aves marinas afectadaspor el accidente del Prestige en Galicia. Ardeola 2007;54:41–51.

Gargallo G. La nueva ficha de muda. In: Pinilla J, editor. Manualpara el anillamiento científico de aves. Madrid: SEO/Birdlife;2000. p. 99–113.

Hatch KA, Sacksteder KA, Treichel IW, Cook ME, Porter WP. Earlydetection of catabolic state via change in 13C/12C ratios ofblood proteins. Biochem Biophys Res Commun1995;212:719–26.

Hobson KA, Clark RG. Assessing avian diets using stable isotopesII: Factors influencing diet-tissue fractionation. Condor1992;94:189–97.

Hobson KA, Alisauskas RT, Clark RG. Stable-nitrogen isotopeenrichment in avian tissues due to fasting and nutritionalstress: implications for isotopic analyses of diet. Condor1993;95:388–94.

Kempster B, Zanette L, Longstaffe FJ, MacDougall-Shackleton SA,Wingfield JC, Clinchy M. Do stable isotopes reflect nutritionalstress? Results from a laboratory experiment onsong sparrows. Oecologia 2007;151:365–71.

Martinez-Abraín A, Velando A, Oro D, Genovart M, Gerique C,Bartolomé MA, et al. Sex-specific mortality of European shagsafter the Prestige oil spill: demographic implications for therecovery of colonies. Mar Ecol Prog Ser 2006;318:271–6.

Mochizuki M, Hondo R, Kumon K, Sasaki R, Matsuba H, Ueda F.Cadmium contamination in wild birds as an indicator ofenvironmental pollution. Environ Monit Assess2002;73:229–35.

Pérez López M, Cid Galán F, Hernández Moreno D, OropesaJiménez AL, López Becerro A, Fidalgo Álvarez LE. Contenido demetales pesados en hígado y plumas de aves marinasafectadas por el accidente del Prestige en la costa de Galicia.Rev Toxicol 2005;22:191–9.

Pérez-López M, Cid F, Oropesa AL, Fidalgo LE, López-Beceiro A,Soler F. Heavy metal and arsenic content in seabirds affectedby the Prestige oil spill on the Galician coast (NW Spain). SciTotal Environ 2006;359:209–20.

Potts GR. Moult in the shag Phalacrocorax aristotelis, and theontogeny of the ‘staffelmauser’. Ibis 1971;113:298–305.

Santos-Echeandía J, Prego R, Cobelo-García A. Copper, nickel andvanadium in the Western Galician Shelf in early spring afterthe Prestige catastrophe: is there seawater contamination?Anal Bioanal Chem 2005;382:360–5.

Sociedade Galega de Ornitoloxia. El impacto del Prestige sobre lasaves marinas de Galicia. http://www.sgosgo.org/archivos/informe3anosprestige.pdf2005.

ThompsonDR, Bearhop S, Speakman JR, Furness RW. Feathers as ameans of monitoring mercury in seabirds: insights fromstables isotope analysis. Environ Pollut 1998;101:193–200.

Velando A, Freire J. Population modelling of European shags(Phalacrocorax aristotelis) at the southern limit: conservationimplications. Biol Conserv 2002;107:59–69.

Velando A, Alvarez D. Cormoranmoñudo Phalacrocorax aristotelisaristotelis. In: Madroño A, González C, Atienza JC, editors. LibroRojo de las Aves de España. Madrid: Ministerio de MedioAmbiente-SEO/Birdlife; 2004. p. 60–2.

Velando A, Álvarez D, Mouriño J, Arcos F, Barros A. Populationtrends and reproductive success of the European shagPhalacrocorax aristotelis on the Iberian Peninsula following thePrestige oil spill. J Ornithol 2005;146:116–20.

Votier SC, Hatchwell BJ, Beckerman A, McCleery RH, Hunter FM,Pellat J, et al. Oil pollution and climate havewide-scale impactson seabird demographics. Ecol Lett 2005;8:1157–64.

Wenzel C, Adelung D. The suitability of oiled guillemots (Uriaaalge) as monitoring organisms for geographical comparisonsof trace element contaminants. Arch Environ Contam Toxicol1996;31:368–77.

Williams JM, Tasker ML, Carter IC, Webb A. A method of assessingseabird vulnerability to surface pollutants. Ibis 1995;137:S147–52.

Williams CT, Buck CL, Sears J, Kitaysky AS. Effects of nutritionalrestriction on nitrogen and carbon stable isotopes in growingseabirds. Oecologia 2007;153:11–8.