Ornis Fennica 72:159-164 1995

Incidence of a Haemoproteuslari parasitemia in a threatened Goll: Larus audouinii

Xavier Ruiz, Daniel Oro & JacobGonzález-Solís

Ruiz, X., Oro, D., González-Solís, J., Dept. de BiologiaAnimal (Vertebrats), Universitat : de Barcelona, Av. Diagonal 645. 08028-Barcelona, Spain

~ Received 15 August 1995, accepted 11 October 1995

Data on haematozoan parasiternias in are scarce and, since blood parasites may influence fitness, it is important to report their incidence in threatenedspecies such as Audouin' s . Blood smearsof 90 Audouin' s caught during incubation at their two main breeding sites, fue Ebro Delta and fue Chafarinas Islands, were exarnined for Haemoproteus lari parasiternia. Overall prevalence of fue parasite was 92.1 % for fue Chafarinas Is. and 28.9% for fue Ebro Delta. To OUTknowledge, prevalences reported in ¡~' this study are fue highest ever recorded for Laridae. Differences in prevalence between localities were significant, but not differences between fue sexes within each locality. Males were more intenselyparasitized than femalesin Chafarinas,and gulls from Chafarinas were more intensely parasitized than those from fue Ebro Delta. Differences in intensity were also checked using data from sympatric Yellow-legged Gulls, L. cachinnans, and only fue factor locality was significant. There was no significant relationship betweenbody condition and intensity of parasiternia in males or females in either locality. These data indicate higher susceptibility to H. lari for gulls breeding at fue Chafarinas Islands.

~_. Introduction infections in wild birds are neededto evaluate their potential effects (Bennett et al. 1992, May 1995), The interaction ofhaematozoa and their avian hosts particularly in fue case of rafe or endangered spe- has received increasedattention since parasiteshave cies, as such Audouin' s Gull, one of fue few SPEC : been recognized as agents influencing many host species (Species ofEuropean Conservation Con- fitness components (reviews by M011eret al. 1990 cero) belonging to category 1 (species of global and Clayton 1991). This influence may occur conservation concero) (Tucker & Heath 1995).The l through sexual selection (Harnilton & Zuk 1982, main potential threat for this species comes from Clayton 1991), through differential survival rates fue concentration of 84% of fue world population in of parasitized vs. non-parasitized birds (Atkinson only two breeding colonies, those ofthe Ebro Delta & van Riper 1991, Davidar & Morton 1993) or andChafarinasIslands (Pedrocchi & Ruiz 1995). through lowered reproductive success(Johnson & This situation is particularly sensitive to fue action Boyce 1991, Korpimaki et al. 1993, Riitti et al. of epizooties, thus deserving particular interest to 1993). fue study of prevalence (the proportion of individu- Therearefewdataonhaemotazoanparasitemias als infected in a population) and intensity (the in Laridae, andstudies reporting fue incidence of number of infected red blood cells in a standard

1. "' t Pc 160 ORNIS FENNICA Vol. 72, 1995 sample from an infected individual) of any kind of (Witt et al. 1982). Smears were air dried and flXed parasitemias,especially those affecting fue popula- in methanol within a few minutes of sampling. A tion during fue breeding season. numeric code identified each slide. In fue labora~ In this contribution we report and analyze data tory fue slides were stained with GIEMSA and on prevalence and intensity of fue haematozoan blind scores were obtained by exarnining smears Haemoproteus lari in Audouin's Gulls Larus underoilat 1000X. audouinii while breeding at its two main world Haemoproteus lari was identified following colonies and compare data of this species with Bennett et al. (1992). No other blood parasiteswere those obtained for a sympatric gull species, fue found (G.F. Bennett, pers. comm.). Smears were Yellow-legged Gull L cachinnans. scoredby one observer.Prevalence was established through inspection of 100 fields containing about 100 erythrocytes each, thus on fue basis of 10.000 2. Material and Metbods erythrocytes observed per sample. Intensity was established by counting fue number of infected 2.1. Study areas erythrocytes in 40 fields, thus on fue basis of 4.00~ red blood cells observed per sample. Additionally, ~ The Chafarinas Islands (Melilla, Spain: 35 ° 11'N, some blood smearsof Yellow-legged Gulls breed- 3° 46'35"E) are three small volcanic islands situated ing sympatricallywith fue Audouin' s Gulls were 4.5 Km offshore from fue Moroccan Mediterra- alsocollected both in fue Ebro delta (N = 26) and nean coast. The only inhabited island is fue central fue Chafarinas (N =6), and processed in fue same one (Isabel 11)and gulls nest on fue other two: Rey way. Francisco, placed very near Isabel 11,which con- tains most of fue Audouin' s Gull nests (ca. 3.700 in 1994) and about 600 nests ofYellow-legged Gull 2.3. Statistics (La rus cachinnans); and Congreso Is., where most of the nests correspond to Yellow-legged Gull Two collections of 15 random1y choosen smears (1.000), but algo holding about 800 nests of fromeachlocalitywerescoredtotestintraobserver Audouin's Gull in 1994. On fue Rey Is. fue domi- variations in prevalence values using fue kappa nant vegetation is formed by Suaeda vera, Lycium index of concordance (Fleiss 1981), while varia- intricatum and Atriplex halimus, whereas on tions in fue degreeof intensity recorded were tested Congreso Is. Salsola oppositifolia and Pancratium using fue intraclass correlation index, approached foetidum are almost exclusive. In both cases fue by means of fue Model 11of single factor ANOV A vegetation covers almost completely fue islands (Zar 1984). Intra-observer repeatibilities were high surface (Blanco 1988). both for prevalence (Kappa Index of Concordanc~ The Ebro Delta (40037'N, 0°21'E), is an alluvial k = 0.88) and for intensity of parasitemias(intraclasc ---"" plain of about 250 Km2 presenting a sandy flat correlation ri =0.93). peninsula of 2500 Ha, where Audouin' s Gull breeds Independence among prevalence, sex, and 10- (more than 10.000 pairs in 1994) together with cality was tested using chi-square statistics on a other species.The landscapeis formed by a three-dimensional contingency table (Zar 1984). mosaic of saltmarshesand scatteredhalophytic veg- Independenceof intensity and sex was testedusing etation (Oro & Martinez 1994). fue Mann- Whitney U test, and associationbetween body condition (body mass/tarsus length) and in- fection intensity, in each locality and sex, was tested 2.2. Sampling parasites using fue Pearson corre1ation. Since, in fue caseof Yellow-legged Gull, some expected frequencies Blood smears of 90 incubating Audouin' s Gulls, were lower than 1%, we usedfue Fisher' s exact test caught under license using nest-trapson Apri11994, to check for differences in prevalence between 10- were obtained by venipuncture both in fue Ebro calities in this species.A two-factor ANOV A was Delta (N = 44) and fue Chafarinas Islands (N = 46). used on log-transformed data (K-S Lilliefors =0.07, were sexed using a bill biometry index d.f. = 74, n.s.) to check for significant differences in

L ~A Ruiz et al.: Blood parasites in Audouin's Gu/l 161

intensity betweenlocalities using datafrom both 4. Discussion species. All tests were two-tailed. The scant data available about haematozoans in Laridae show that fue prevalences are usually very 3. Results low (ca.5% or less,see e.g. Peirce 1981, Bennet et

Table 1 shows fue number of Audouin's Gulls Table 1. Frequencies of parasitism by Haemoproteus infected vs. non-infected for each locality and sexo lari on Audouin's Gul/ by sex and locality. . Globalprevalencewas92.1%fortheChafannasIs. EBRO DELTA CHAFARINAS (94.1% formales and90.5% forfemales) and28.9% d' Q d' Q for fue Ebro Delta (17.6% for males and 35.7% for females).Overallindependenceamongprevalence, PARASITIZED 3 10 20 23 sex andlocalitywasrejected(M2=62 01 P

r-- and tests of partial rndependence for each varIable failed to reject fue null hypothesis only for sex (X42 =2.48, n.s.). Thus, a 2 x 2 contingencytable wasused to testfue independencebetween preva- Table 2. Intensityof Haemoproteus lari parasitism lence and locality. Results show that gulls from among infected individuals by sex and locality. Mean Chafarinas Is. presented significantly higher levels number of infected erythrocytes (standard deviation) of prevalence (X21= 39.11, P < 0.0001) thanthose andsample size are given. ofthe Ebro Delta. EBRO DELTA CHAFARINAS The results obtained checking for differences by locality using fue Yellow-legged Gull sam- MALES 30.0 (40.9) N = 3 27.35 (16.6) N = 20 pIes (Table3) alsoindicated significantly higher FEMALES 2.8 (3.36) N = 10 12.91 (8.1)N = 23 prevalencesfor fue Chafarinascolony (Fisher's exact test P = 0.002). Parasitism intensity values are given in Table Table 3. Prevalence of parasitism by Haemoproteus . lari on Yel/ow-legged Gul/s by locality. Differences 2. Srnceprevalence fo.r males at th~ Ebro Delta are statistical/y significant (Fisher's exact test P=0.002) was much lower than rn fue ChafarInasIs., thus indicating higher prevalences in the Chafarinas colony. providing a largely unbalanced sample to test intensities(Table 2), fue associationof intensity EBRO DELTA CHAFARINAS with sex was teste.d~nly at fue.Chaf.annas. ~s., PARASITIZED 7 6 /"""" wheremales had slgmficantly hlgher rntenslties NON PARASITIZED 19 O than females(U = 96.0,P < 0.001).The correla- TOTAL 26 6 tion analysesof body conditionwith intensity showedthat fue correlation was not significant l' mal l' al t .th 1 alit (Eb D lta Table 4. Intensity of Haemoproteus lari parasitism J.or esor J.em es a el er oc y ro e . f d . d. 'd I b . d I I. among In ecte In IVI ua s y specles an oca Ity . males, rl = 0.56, P = 0.62; Ebro Delta females, Mean number of infected erythrocytes (standard de- rs = -0.08, p:;; 0.83; Chafarinas males, rlO= 0.57, viation) and sample size are given. A two factor P = 0.052; Chafarinas females,' rlS= -0.07, ANOVA on long trans formed data has shown that P = 0.75). only locality is responsible ,for si~nificant ~ifferences S'.fi d'f~ ... f (F1 = 29.3, P = 0.0001), whlle nelther specles nor the 19m lcant 1 erences rn rntenslty o para- interaction between species and locality were signifi- sitemiabetween localities were also checkedus- canto ing datafrom both species,and only locali,tywas significant (F¡ = 29.3, P = 0.0001)(Táble 4); EBRODELTA CHAFARINAS while neither species(FI = 2.9, P = 0.09),nor . . . b . dI l. (F 11 Lcachlnnans13.5(14.03)N=722.7(14.33)N=6 rnteraction etweenspeclesan ocalty 1= ., Laudouinii 8.7 (19.97) N = 13 19.6(14.26)N=43 P = 0.3) was significant.

..

"""" 162 ORNIS FENNICA Vol. 72, 1995 al. 1992). To OUTknowledge fue present results poorer feeding conditions than that of fue Ebro show much larger prevalences than ever recorded Delta. However, since, in contrast, Yellow-legged for Laridae. Since large seasonal variation in preva- Gulls showed larger proportions of two-egg clutches lence occurs in other species (Weatherhead & at fue Ebro Delta (30%) than in Chafarinas (20%) Bennett 1991, 1992; Allander & Bennett 1994, (authors, unpublished), buthigherincidence ofpara- Norris et al. 1994), fuese high values might be sites at Chafarinas, other variables non mutually attributable to fue fact that all OUTsamples were exclusive with breeding effort, such as fue degree taken during fue breeding season, when chronic of exposure to parasite vectors, Ceratopogonid flies infections may relapse owing to either the ofthegenusCulicoides(Allander&Bennet1994), irnmunosuppressive action of sexual hormones (Zuk or fue incidence of other irnmunosuppressive agents, 1991, Atkinson & van Riper 1991, Folstad & Karter e.g. organochlorine pollutants such as PCBs (Peakall 1992), increased exposure to parasite vectors (M~ller 1986), need to be considered. 1994) or breeding effort (N orris et al. 1994). According to Allander & Bennett (1994), and At present three striking patterns in prevalence G. F. Bennett (pers. comm.), prevalence values in ofblood parasites amongst birds have been recog- haematozoan parasitemias are mainly dependent "' nized (Norris et al. 1994), one of which is that on fue abundance and activity ofparasite vectors, amongst nonraptorial altricialland birds prevalences i.e. exposure, so we can expect higher densities or tend to be larger at higher latitudes (Ricklefs 1992). activity of Culicoides at Chafarinas than at fue Ebro According to OuTdata northern populations of both Delta. We have no data on Culicoides density at fue gull species showed lower prevalence than more two colony sites, but, since temperature and wind southern ones. Clearly further research is needed to may affect fue feeding activity of dipteran parasite ascertain whether this is an exceptional case or vectors (McCreadie et al. 1986, quoted in Allander whether exhibit different trends to land & Bennett 1994), we can report that weather pa- birds (see Peirce 1981 for values in European coun- rameters are consistent with this: during fue incuba- tries). Ricklefs (1992) explains this trend at tion stage (April month) fue Chafarinas Is. were interspecific level through a possible link between hotter (mean temperature 18°C) than fue Ebro Delta fue length afilie embryonic period and fue matura- (mean temperature 13°C), and, furthermore, fue tion of an effective immune system, since preva- Ebro Delta is a windy afea, and fue vegetation lence is inversely related to fue length afilie incu- cover of fue colony site is scarcer than in fue batían period for a given egg-size. However, this Chafarinas (see methods). should not be fue case in an intraspecific compari- The fact that fueTe were no significant differ- son in which only slight differences in fue incuba- ences in prevalence between fue sexes indicates a tion period are found (authors, unpubl.). similar degree of exposure to parasite vectors dur- an fue other hand, N orris et al. (1994) hypoth- ing incubation. This is consistent with OuTobserva- haematozoanesized that allprevalence three patternsin birds observed(Read 1991,for tionssiteatthisbreeding on fue time spentstage by (authors each sex unpubl.). at fue colonyThisis ~ ~

Ricklefs 1992) can be explained through breeding algo fue most usual trend found in fue literature on effort. Thus, fue fact that gulls from fue Chafarinas blood parasites in birds (Stacey et al. 1990, Allander Islands showed significantly higherprevalences than & Bennett 1994, O'Dell & Robbins 1994), although those of fue Ebro Delta, would mean that gulls fue possibility of finding differences may rely on breeding in Chafarinas invest greater parental ef - sampling date (Weatherhead & Bennett 1992). forts than gulls in fue Ebro Delta. This is consistent Because Haemoproteus infections have pre- with data on reproductive parameters such as clutch patent periods of about 14 days (G.F. Bennett, pers. gire, since fueTe was a larger proportion of two-egg comm.), in Audouin' s Gull fue significantly higher clutches in Chafarinas (25%) than in fue Ebro Delta intensity of parasitemia shown by males at (12%) (authors unpubl.). Since in gulls, clutch-size Chafarinas colony, should reflect greater suscepti- variability reflects mainly nutritional constraints dur- bility to H. lari during fue pre-laying periodo This ing fue egg-synthesis period (Reid 1987, Salzer & coincides with fue courtship feeding and egg-syn- Larkin 1990, Bolton et al. 1992), fuese datacould thesis period, which lasts about 20 days in gulls indicate that fue Chafarinas population was under (Salzer & Larkin 1990), and also with other costly

/A~/ A

~" ~, ~ Ruiz et al.: Blood parasites in Audouin's Gull 163

male activities such as spenn competition (Birkhead Ebro joen suistossaEspanjan it'árannikolla. Linnut & M~ller 1992, Norris et al. 1994). Moreover, pyydystettiinhaudonta-aikaanpesimasaariltaanveri- during fue pre-laying period drastic changesoccur naytteiden ottoa varten. Loisittuja Villimerenlokki- in hormone balance of fue birds, which mar inter- yksiloita esiintyi runsaarnmin Chafarinas saarilla act in a complex way with fue immunocompetence (92% yksiloistií loisittu) kuin Ebron suistossa(29% of fue sexes(Korpimaki et al. 1993, Saino & M~ller yksiloistií loisittu). Nain korkeita loisintafrekvensseja 1994). In any case,differences in intensity between ei lokkilinnuilla ole aikaisemmin havaittu. N aaraita localities are also indicative of gulls being more ja koiraita oli loisittu yhta suurella frekvenssilla, susceptible to parasites when reproducing in fue mutta koirailla loisia esiintyi lukuma'árilisesti Chafarinas Islands. enemman loisittuja yksiloa kohti. Tulokset Kelta- The absenceof relationship between body con- ja1ka1okiltaolivat samansuuntaisia.Linnun kunnolla dition and parasitemia intensity does not suggest ja loistartunnan voimakkuudella ei nayttanyt olevan . any effect of fueseparasites on gulls' health. Moreo- yhteyttii. Kirjoittajat pohtivat loisien esiintymiseen ver, fue only relation approaching significance is vaikuttavia tekijoitiíjaloisten vaikutuksiaisantiinsa. ¡-- that of males from Chafarinas, but it is in fue re- verse direction, i.e. males in best condition were those more intensely parasitized. This could be References because Haemoproteus are relatively benign . 993) dth mal . b d" Allander,K.&G.F.Bennett.1994:Prevalenceandrnten- (B ennett 1 , ~ e.n es m e~er con lbon, sity of haematozoaninfection in a population of Great could be older blrds, l.e. those WhlCh have had TitsParusrnajorfromGotland,Sweden.-J.ofAvian longer exposure to fue parasites. Values of relation- Biology 25 :69-74. ship for females are too low to suggest any trend. Atkinson, C. T. & C. van Riper ill. 1991: Patbogenicity However, fue effects of parasites on other fit- and epizootiology of avian haematozoa:Plasmodium, ness components, such as breeding success (includ- Leucocytozoon, an~ Haemo~ro~us. -. In: J. E. Loye . . " ) d . & M. Zuk (eds.), Blrd-paraslte rnteractIons:Ecology, mg extra-patr paternlbes or en urance to Inlgra- evolution, and behaviour: 19-48. Oxford University tory effort, need to be thoroughly investigated be- Press, Oxford:19-48. fore concluding that such parasitemias, at their Bennett,G. F. 1993:Phylogenetic distribution and possible present level, are irrelevant to gulls' fitness. evolution of tbe avian speciesof tbe Haemoproteidae. - SysternaticParasitology 26:39-44. Acknowledgments.The autbors are grateful to Gordon Bennett,G. F., R. A. Earlé,H. du Toit & F. W. Huchzermeyer. F. Bennett for help in parasite determination, cornments, 1992: A host-parasitecatalogue ofthe Haematozoaof and critical reading of tbe manuscript. To Lluis Jover for tbe Sub-Saharanbirds. - OnderstepoortJ. vetoRes. statistical advice, and to two anonymous reviewers for 59:1-73. improvementson tbe submittedmanuscript. José Luis Tella Birkhead, T. R. & A. P. M~ller. 1992: Sperm competition ",--. suppliedus inforrnation and metbodologyabout blood para- in birds. Evolutionary causes and consequences.- sites. Xell Genovart and Joan CarIes Abella helped us in Academic Press.London. tbe field work. Robin Rycroft (EIM, Universitat de Barce- Blanco, E. 1988: Plantas de las Islas Chafarinas y lona) improved tbe English texto Autborities of tbe Ebro descripción de su paisaje vegetal. Actes del Simposi Delta Natural Parc give us logistic support. Funds were Internacional de BotAnica Pius Font i Quer. vol 11. provided by ICONA and DGICYT grant PB91-027I of tbe Fanerogamia:333-343. . SpanishGovernment. The Audouin's gull project is spon- Bolton, M., D. Houston & P. Monaghan. 1992:Nutritional sored by IBERlA., Líneas Aéreas Españolas. constraintson egg forrnation in tbe LesserBlack-backed Gull: an experimental study. - J. of Ecology 61:521-532 . .. Clayton, D. H. 1991: The influence of parasites on host Selostus: Haemoproteus lan- verllolSen sexual selection. - Parasitol. Today 7:329-334. esiintyminen ViilimerenlokiUa Davidar, P. & E. S. Morton. 1993: Living witb parasites: prevalence of a blood parasite and its effect on Kirjoittajat tutkivat Haemoproteus lari -veriloisen . survivorship in tbe. ~Ie Martin. - 110: 109-116.

.. t . +,.. Villi' . 1 kill (La ud .. .) . Flelss, J. L. 1981: StatIstIcal metbods for rates and propor- eSlm YInlSLa meTen O a rus a oumu Ja tI. J h W ' '.. . ons. o n 1I ey & Sons. N ew Y or.k 321 pp. keltaJa1ka1okilla (L cachmnans) kahdella tutkimus- Folstad, l. & A. J. Karter. 1992: Parasites,bright males. alueella, Chafarinas saarilla,jotka sijaitsevat lahella and tbe irnmunocompetencehandicap. - Am. Nat. Afrikan pohjoisrannikkoa Marokon edustalla, ja 139:603-622. 164 ORNIS FENNICA Vol. 72, 1995

Harnilton,W. D. & M. Zuk. 1982:Heritable true fitness Peirce,M. A. 1981:Distribution and host-parasite check- and bright birds: A role for parasites?- Science list of fuehaernatozoa of birs in WestemEurope. - J. 218:384--386. of NaturalHistory 15:419-458. JohnsonLL. & M. S. Boyce. 1991:Fernale choice of Ratti,O., R. Dufva& R. V. Alatalo.1993: Blood parasites maleswith low parasiteloads in SageGrouse. -In: J. andmale fitness in fuePied Flycatcher. - Oecologia E. Loye & M. Zuk (eds.),-parasite interactions: 96:410-414. Ecology,evolution, and behaviour. Oxford University Read,A. F. 1991:Passerine polygyny: a role for para- Press,Oxford:377-388. sites?- Am. Nat. 138:434-459. Korpimiiki, E., H. Hakkarainen& G. F. Bennett:1993: Reid,W.V.1987:ConstraintsonclutchsizeintheGlaucous- Bloodparasites and reproductive success ofTengrnalm' s wingedGull. - Studiesin Avian Biology 10:8-25. owls:detrimental effects on fernalesbut noton males? RickIefs,R. E. 1992:Embryonic development period and - FunctionalEcology 7:420-426. fueprevalence of avianblood parasites. - Proc.Nat. May,R. M. 1995:Disease and fue abundance and distribu- Acad.Sci. USA 89:4722-4725. tion ofbird populations:a sumrnary.-Ibis 137Suppl. Saino,N. & A. P. M~ller. 1994:Secondary sexual charac- 1:85-86. ters,parasites and testosteronein fue Barn swallow, M~ller, A. P. 1994:Parasites as an environmental compo- Hirundorustica. - Anim. Behav.48:1325-1333. nentswallow of reproductionHirundo rustica. in birds- Ardea as exemplified82: 161-172. by fue Salzer,feedingD. W.on & clutchG. J. Larkin. and third-egg 1990:Impact size inof Glaucous- courtship ~ .; - M~ller,A. P.,K. Allander&R.Dufva. 1990: Fitness effects wingedGulls. - Anim. Behav.39:1149-1162. - of parasiteson passerinebirds: A review. - In: J. Stacey,L. M., C. E. Couvillion,C. Siefker& G. A. Hurst. Blondel,A. Gosler,J.-D. Lebreton & R. H. McCleery 1990:Occun-ence and seasonal transmission ofhernatozoa (eds.),Population Biology of passerinebirds: An inte- in wildturkeys. - J.ofWild1ife Diseases 26:442-446. gratedapproach.SpringerVerlag,Berlin:269-280. Tucker,G. M. & M. F. Heath. 1995:Birds in Europe: Norris, K., M. Anwar & A. F. Read.1994: Reproductive Their ConservationStatus. Birdlife ConservationSe- effort influencesfue prevalence ofhaematozoan para- ries 3. Birdlife Intemational.Cambridge sitesin GreatTits. - J.Anim. Ecol.63:601-610. Weatherhead,P. J. & G. F. Bennett.1991: Ecology of O'Dell, J. P. & L. W. Robbins.1994: Hematozoa ofwood Red-wingedBlackbird parasitism by haematozoa.- ducks (Aix sponsa)in Missouri. - J. of Wildlife Can.J. ZooI69:2352-2359. Diseases30:36-39. Weatherhead,P. J. & G. F. Bennett.1992: Ecology of Oro, D. & A. Martinez.1994: Factors affecting Klepto- parasitismof Brown-headedCowbirds by haematozoa. parasitismandpredationrates upon acolony of Audouin's - CanJ. Zoo170:1-7. Gull (Larusaudouinii) by Yellow-leggedGulls (Larus Witt, H.-H.,N. Stempel,E. deJuana & J. M. Varela.1982: cachinnans).-Colon.Waterbirds 17:35-41. Geschlechtsunterschiedebei Korallenmowe(Larus Peakall,D. B. 1986:Accumulation and effects on birds.- audouinii)nach meBbaren Merkrnalen. - Die Vogel- In: J. S.Waid (ed.), PCBs and fue Environment, vol. II warte31:457-460. CRCPress. Boca Raton, Florida:31-47. Zar, J. H. 1984:Biostatistical Analysis. 2nd. Ed. Prentice Pedrocchi,V. & X. Ruiz. 1995:On fue currentstatus of Hall Intemational.NewJersey. Audouin'sgull Larusaudouinii in fueMediterranean. Zuk, M. 1991:Parasites and bright birds:New dataand a -In: Tasker,M. L. (ed.),Threats to seabirds:Pro- newprediction.-In:J.E.Loye&M.Zuk(eds.),Bird- ceedingsof fue 5th IntemationalSeabird Group con- parasiteinteractions: Ecology, evolution, and behav- ~ ference.Seabird Group, Sandy. p. 40. iour.Oxford University Press, Oxford:317-327. J

~

--'