Unraveling theimportanceofricefieldsforwaterbird populations inEurope

Gregorio M.Toral • Jordi Figuerola

Abstract Rice fieldsareanalternativehabitatforwaterbirdsandprovidefoodand shelterformany avianspecies,butthereisalackofinformationabouthowtheuseofrice fields translatesintopopulationleveleffects.Theaimofthisstudywastotesttherela- tionshipbetweentheuseofricefields byEuropeanwaterbirdsandtrendsintheirpopu- lations.Wetestedthisrelationshipduringtheautumnmigrationseasonandduringthe breeding season.Basedoncountsconducted overthelast23yearsinnaturalmarshesand areasof ricefieldsinDon˜ana(SWSpain),an indexofricefieldusewasconstructedfor 76 birdspecies,whichwasthencomparedtothesespecies’Europeanpopulation trends obtainedfromtheliterature.Apositiverelationship wasfoundbetweenwaterbirdpopu- lationtrendsand theuse ofricefieldsduringautumnmigrationseason.Ourstudysuggests thatchangesintheCommonAgriculturePolicyinEuropeleadingtoreductionsinareasof ricecultivationmay haveimportanteffectson waterbirds.Therestorationofformermarsh areasandthemaintenanceofricecultivationwould seemtobemore environmentally friendlyapproachesthantheuseoftheseareastogrowalternativecropsorsolarfarms.

Keywords Agriculture Don˜ana Migration Shorebird Wetland

Introduction

Intensehabitattransformationlinkedtohumanactivityhasmoldedthecompositionand abundanceofaviancommunitiesthroughouttheworld(Murphy2003;Lemoineetal.

2007)andagriculturalintensificationhasbeenassociatedwithseriousdeclinesinthe populationsofmanyfarmlandbirdsoverrecentdecadesinEurope(Donaldetal.2001,

2006;Wretenbergetal.2007).Therearealsomany studiesabouttheeffectsof agriculture

G.M.Toral() J.Figuerola

DepartmentofWetlandEcology,Don˜anaBiologicalStation,C.S.I.C.,Avda. Ame´ricoVespucios/n41092,P.O.Box1056,41080Seville,Spain

e-mail:

J.Figuerola

e-mail:

onwaterbirdpopulations.Duncanetal.(1999)showedanegativeeffectofagriculture intensification onwinteringduckspopulations.Longetal.(2007)concludedthatan increaseintheareaofagriculturallandcouldbeassociatedwithdecreasingpopulationsin Anseriformes.However,somecropscouldact asanalternativeor‘buffer’habitat forsome waterbirdspeciesthatfeedorbreedinagriculturalareas(Sa´nchez-Guzma´netal.2007)and therearesomestudiesreportingpositive effectsofagricultureoncertainspeciesof waterbirds(e.g.,Gauthieretal.2005;Fasolaetal.2010).

Wetlandsareoneofthemostproductiveecosystems onEarth.Nevertheless,theyhave beensubmittedtointensivetransformationanddrainageandtheyaretodayindecline throughout theworld(CzechandParsons2002).Itisestimatedthatover50%ofthe world’swetlands havebeenlostsince1900.Overthelastfewcenturies thenaturalwet- landsoftheMediterranean regionhavebeenreducedinareaby80–90%duetopressure fromhumanpopulation growthandtheconversionofwetlandsintoagriculturaland urbanizedareas(Finlaysonetal.1992).InsouthernEuropeandNorthAfricalargeareasof pristinewetlandshavebeentransformedintoricefields, acropthatcovers581,978hain Europe(FerreroandNguyen2004).

Previousworksuggeststhatricefields maybeanimportanthabitatforwaterbirds throughout theworldandinsomeareasmayinfactbetheprimaryforaginghabitat availableto them(CzechandParsons2002).Severalstudieshaveshowntheimportance of ricefieldsasawinteringsiteforwaterbirdsindifferentlocationsaroundtheworld,suchas California(ElphickandOring1998;ElphickandOring2003)or Cuba(Acostaetal.1996) inNorthAmerica;Portugal(Lourenc¸oand Piersma2008)orSpain(Rendo´netal.2008)in Europe;orJapan(Maeda2005)inAsia.Furthermore,ricefields areusedbyavarietyof waterbirdsasbreedingsites(FasolaandRuiz1996),althoughtoalesserextentthanas foraging sites(CzechandParsons 2002). Despite alltheinformationabouttheimportance ofthiscropforwaterbirds,thepossibilitythattheuseofricefieldsactuallytranslates into populationleveleffectshasreceivedlittleattention.Sincericefields differinhabitat structure,watertabledepth,anddisturbance levelsfromnaturalmarshes,andgiventhat thesefactorsareknowntoaffecthabitatselectionbywaterbirds (BolducandAfton2004; Canepucciaetal.2007),itislikelythatnotallspecieswill use this alternativehabitatwith thesameintensity.

HundredsofthousandsofbirdsusetheareaofDon˜ana(SWSpain),whichincludes marshesandricefields,inthecourseofayear.TheDon˜anamarshesarethemostimportant winteringsiteformigratingwaterbirds intheMediterranean(Rendo´netal.2008)andone ofthemain migratorystopoversforwaterbirdsusingtheEastAtlantic flyway.Inthispaper wetestthehypothesisthatthespeciesthatusericefieldsmostintensivelywillbenefitat thepopulationlevelfromthepresenceofthisalternativehabitat.Wetestthishypothesisby analyzingtherelationshipbetweenricefield useandpopulationtrendsinEuropefor76 speciesofwaterbirdspresentinnaturalmarshesand/orricecropsinsouthernSpain.

Methods

InSpain,ricefields cover118,000(FAOSTAT2004)andthelargestsuchareainSpain (36,000ha)issituatednearthemarshesoftheDon˜anaNationalPark, a55,000hawildlife reservenorthoftheGuadalquivir estuaryinAndalucia(SWSpain).Ofthe180,000haof freshandbrackishmarshespresentin1900,36,000 haweretransformedintoricefields between1926and1997(Garc´ıa-NovoandMart´ın-Cabrera2005;Rendo´netal.2008). Additionally,othertypesoftransformationhavereducedtheareaofnaturalmarshesto

30,000haatpresent(Enggass1968;Garc´ıa-NovoandMart´ın-Cabrera 2005)andchanges inthemarshes’hydrologicalstructurehavedramaticallyalteredtheflood patternsofthe area.DuetosummerdroughtsandtheunpredictabilityofrainfallinMediterraneanclimate zones,theperiodsofmaximumfloodlevelsinricefieldsandnaturalmarshesdonot overlapintime,therebyensuringthatricefields canactasanalternativehabitatfor waterbirdsinthearea.

Bird’scountsandpopulationindices

Weusedbirdcountsconductedoverthelast23years(1980–2003)intheDon˜anaNational Parkandsurroundingareathatarepartofthelong-termmonitoringprogramruninthearea ( analyzed terrestrial sur- veysfromthreesectorsofricefields(2,756ha)andsevensectorsofmarsh(3,192ha) sincethehighestnumberofsurveysexistforthesetensectors.Anindexofricefieldusein relationtonaturalmarshusewascalculatedforeachspeciesbasedontheproportion of positivecounts (countsinwhichatleastonebirdofaparticularspecies wasdetected)in rice fieldsin relationtothe numberofpositivecountsin bothhabitats(naturalmarshesand ricefields).Wedidn’tbasetheindexofricefielduseonabundance datatoavoidthebias producedbybigflocksofwaterbirds.WeusedEq.1tocalculatetherice-fielduseindex foreachspeciesandmonthbasedonitspresence/absence.

Rice-fielduseindex¼MPR=ðMPRþMPMÞ:ð1Þ

MPR(MeanPresenceinRicefields)=numberofpositivecountsinricefields/total numberofcountsinricefields.

MPM (Mean Presence in Marshes)=number of positive counts in marshes/total numberofcountsinmarshes.

Indiceswereonlycalculatedforspecieswithatleastfivepositivecountsinthearea (76species).Wecalculatedtwoseparateseasonal-use indices(breedingandautumn migration)adjustedtocropandbirdphenology.Wecalculatedseasonalindicesasthemean valueofthemonthsincludedintheseason.Thebreedingseasonindexwascalculatedusing datafromApriltoJuly,aperiodthatlargelycoincides withriceseeding andgrowth. The autumnmigrationseasonindexwascalculatedwithdatafromOctober toJanuary,months includingautumnmigrationandthewinteringperiodthatcoincidewith cropripening,soil tilling, andthe shallowfloodingoffields,whichdryoutbythe endofDecember–January.

Wedefinepopulationasa‘distinctassemblageofindividualswhichdoesnotexperience significant emigrationorimmigration’(DelanyandScott2006).Europeanpopulation trendsforwaterbirdswereobtainedfrom‘WaterbirdPopulation Estimates’(Delanyand Scott2006),wheretrendsareexpressedasoneofthefollowingcategories:increasing, stable,decreasing,fluctuatingandextinct. Thosecategorieswerescoredforthe analysesas positive(2),notrend(1),ornegative(0),andtheresultingdependentvariablewastreatedas anordinalvariableintheanalyses. Weassigned notrendvalue(1)toMarmaronetta angustirostris,theonlyspecieswithafluctuatingpopulationtrendforEurope.Inadditionto ricefieldusewe alsoestimatedmeanmigrationdistanceasdegreesoflatitudebetweenthe meanbreedingandmeanwinteringranges(seedistribution mapsinCramp1977–1985).

Statisticalmethods

TherelationshipbetweenpopulationtrendsinEurope(asadependentvariable)andthe rice-fielduseindex(asanindependentvariable)wastestedusingGeneralizedLinear

MixedModels(GLMMs).Weconsideredthepossiblelackofindependence ofoursam- ples,sincespeciesfromthesamefamilyshareanumber ofmorphologicalandlifehistory featuresandthereforecouldhavesimilarpreferences intheirhabitatselectionforfeeding and/orbreeding.Toavoidthispossiblebiaswemadeanexploratory analysis,usinga nestedGLMtoexaminehowtrendsinEuropeanbirdpopulations varybetweendifferent taxonomiclevels(order,familyandgenera).Mostofthevarianceinpopulationtrendswas unrelatedto taxonomy(82%).Familyandgeneraexplained12and6%,respectively,ofthe variance;novariancewasexplainedbyorder.Furthermore,weusedanestedGLMto examinethevariationintheuseofricefields betweendifferenttaxonomiclevels(order, family,andgenera).

TheGLMMswerefitusingtheGLIMMIXprocedureinSAS9.2.Thepopulationtrend inEuropewasmodelledbyspecifyingamultinomialerrordistribution andacumulative logitlinkfunction,withtheexplanatoryvariablescodedasfixed effectsandfamilyand generanestedwithinfamilyasrandomeffects(orderwas excludedasrandomeffectbased ontheresultsoftheexploratoryanalysis).Becausethetaxonomywasincorporated inthe modelwith ahierarchicalstructure(i.e., generanestedwithinfamilies),themodelhelps to correctforthephylogeneticeffectsdescribedbetweenthesetaxonomiccategories.

Twodifferentsetsofmodelswerefitted tothedata,oneforeachofthetimeperiods usedtocalculatetherice-field useindex.Theexplanatoryvariablesintheinitialmodels includedrice-fielduseindex,migrationdistance(anditsquadraticterm),aswellasthe two-wayinteractionwiththerice-field useindex.Weincludedmigrationdistanceasan explanatoryvariableandalsoincludeditsquadratictermtoexplorethepossiblediffer- encesin populationtrendsbetweenshort,intermediate, andlongdistancemigrantsbecause previousstudieshavefoundAfro-Palearctic migrantbirdstohavenegativepopulation trends overthelastfewdecades incomparisontoshort-distancemigrantsorresident birds (Sandersonet al.2006).Wealsoaddedtheinteraction ofrice-fielduseindexandmigration distance sincethishabitatisveryimportantasafeedingsiteandcouldhaveadifferent valueforwaterbirdspeciesdependingontheirrequirementsoffood,whichcouldbe relatedtothedistancetheyhavetocoverduringmigration.Wefollowedabackwards regression-model selectionprocedure,andthefactorcontributingleasttothemodelwas removedateachstepbeforethemodelwasrefitted.Variableswereretainedinthemodel untilalltheinteractionsincludingthevariablehadbeenexcludedfromthemodel;alower probabilitylimitof0.10wasusedtoretainafactorinthemodel.Onlyfactorswith P\0.05were interpreted as statistically significant.We usedthis methodof model selectioninsteadofAkaikeInformationCriterion(AIC)becausewehaverandomeffectsin ourmodels.Although therangeofapplicationofAICcanbeexpandedtorandomeffect, thisextensionisinanearlystageofdevelopment(BurnhamandAnderson2002).We intendedtoincludealsobreedinglatitudeasanexplanatory variableinouranalyses. However it was strongly correlated with migration distance (r2 =0.41; F=48.2; DF=1,70;P\0.001),andtheinclusionofbothvariablesintheanalysesmayproduce spuriouscorrelations. Forthesereasonsweconfirmedinseparateanalyses(notpresented) thattherelationships foundwereinfactwithmigrationdistanceandwerenotsignificant whenusingbreedinglatitudeinsteadofmigrationdistance.

Results

Although theautumnmigrationperiodandbreeding seasonindiceswerepositively cor- related,therelationshipwasnotstrong(r2 =0.29;F=27.4;DF=1,69;P\0.001),

Fig.1 Variationin thenumberofwaterbirdsthat usetherice fieldsandmarshesinDon˜ana duringtheyear. Notethatthenumbers referonlytotheareasincluded inourstudyandnottothecompletesurfaceareaof marshesandricefieldspresentinDon˜ana

suggestingthat somedifferencesoccurinspecies’habitatpreferencesbetweenautumnand spring.Birdabundanceinricefields incomparisontonaturalmarsheswashigherduring theautumnmigrationseasonthaninspring(Fig.1).

Therewasimportantinterspecificvariabilityintherice-fielduseindex(seeTable1) thatwasnotexplainedbytaxonomy, withvaluesrangingbetween0(e.g.,Aythyanyroca) and1(e.g.,Sternanilotica).Varianceexplainedattheorderlevelvariedbetween20and

24%dependingonthetimeperiodanalyzed.Atthefamilylevel, thevaluesvariedbetween

0and21%.Novariancewasexplainedatthegeneralevel.

Thefinalmodelfortheautumnmigrationseasonretainedthreevariables:useindex, migrationdistanceanditsquadraticterm(Table2).According tothismodel,population trendsinEuropewerepositivelyrelatedtotheautumnrice-fielduseindex,andhada quadraticrelationship withmigrationdistance,withmorenegativepopulationtrendsfor speciesmigratingovermediumdistancesthanforlongdistanceandresidentspecies.Inthe caseofthebreedingseason,thefinal modelwasverysimilarandretainedfourvariables (Table2).Therelationshipofpopulationtrendswithmigrationdistancewassimilarto whatwefoundduringtheautumn.Rice-fielduseindexduringthebreedingseasontended tobepositivelyrelatedtopopulationtrends(P=0.08)andmorestronglyinresidentthan forspeciesmigratingoverlongerdistances(P=0.06)(Table2).

Discussion

Wefoundimportantinterspecificdifferencesinthefrequencyofuseofricefieldsas opposed to natural marshes in Don˜ana, which were related to population trends in waterbirdpopulationsinEurope.Thissuggeststhatthealternativehabitatsprovidedby ricefieldsduringtheautumnmigrationperiodfavorthepopulationsof certainspecies.We alsofoundthisrelationduringthebreeding seasonalthough itdidn’treachsignificance, suggestingthatwaterbirdsbenefitfromusingricefieldsmainlyasafeeding habitatduring theautumnmigrationperiod.

Table1 Valuesfortherice-fielduseindexforthespeciesanalyzedinthetwodifferenttimeperiodstaken intoaccount

Specie / Migrationseason index / Breedingseason index / Order / Family
Aythyanyroca / 0.00 / 0.00 / Anseriforms / Anatidae
Oxyuraleucocephala / 0.00 / 0.00 / Anseriforms / Anatidae
Aythyaferina / 0.01 / 0.00 / Anseriforms / Anatidae
Nettarufina / 0.00 / 0.00 / Anseriforms / Anatidae
Aythyafuligula / 0.03 / 0.00 / Anseriforms / Anatidae
Tadornatadorna / 0.03 / 0.00 / Anseriforms / Anatidae
Anascrecca / 0.10 / 0.00 / Anseriforms / Anatidae
Anasstrepera / 0.10 / 0.00 / Anseriforms / Anatidae
Anaspenelope / 0.19 / 0.00 / Anseriforms / Anatidae
Anasclypeata / 0.14 / 0.03 / Anseriforms / Anatidae
Anasplatyrhynchos / 0.14 / 0.04 / Anseriforms / Anatidae
Anasacuta / 0.33 / 0.00 / Anseriforms / Anatidae
Marmaronetta angustirostris / 0.34 / 0.03 / Anseriforms / Anatidae
Anasquerquedula / 0.19 / 0.00 / Anseriforms / Anatidae
Anseranser / 0.50 / 0.00 / Anseriforms / Anatidae
Egrettaalba / 0.00 / 0.00 / Ciconiiforms / Ardeidae
Nycticoraxnycticorax / 0.00 / 0.00 / Ciconiiforms / Ardeidae
Ixobrychusminutus / – / 0.18 / Ciconiiforms / Ardeidae
Ardeacinerea / 0.60 / 0.14 / Ciconiiforms / Ardeidae
Ardeolaralloides / 0.93 / 0.13 / Ciconiiforms / Ardeidae
Ardeapurpurea / 0.99 / 0.14 / Ciconiiforms / Ardeidae
Bubulcusibis / 0.94 / 0.08 / Ciconiiforms / Ardeidae
Egrettagarzetta / 0.78 / 0.21 / Ciconiiforms / Ardeidae
Burhinusoedicnemus / 0.50 / 0.00 / Ciconiiforms / Burhinidae
Recurvirostraavosetta / 0.17 / 0.11 / Ciconiiforms / Charadriidae
Charadrius dubius / 0.20 / 0.00 / Ciconiiforms / Charadriidae
Himantopushimantopus / 0.47 / 0.14 / Ciconiiforms / Charadriidae
Pluvialissquatarola / 0.48 / 0.21 / Ciconiiforms / Charadriidae
Vanellusvanellus / 0.59 / 0.08 / Ciconiiforms / Charadriidae
Charadrius alexandrinus / 0.44 / 0.33 / Ciconiiforms / Charadriidae
Charadrius hiaticula / 0.67 / 0.43 / Ciconiiforms / Charadriidae
Pluvialisapricaria / 0.64 / 0.00 / Ciconiiforms / Charadriidae
Ciconiaciconia / 0.81 / 0.53 / Ciconiiforms / Ciconiidae
Ciconianigra / 0.97 / 0.00 / Ciconiiforms / Ciconiidae
Glareolaprantincola / – / 0.42 / Ciconiiforms / Glareolidae
Larusgenei / – / 0.00 / Ciconiiforms / Laridae
Larusminutus / 0.12 / 0.00 / Ciconiiforms / Laridae
Sternaalbifrons / 0.00 / 0.13 / Ciconiiforms / Laridae
Larusfuscus / 0.65 / 0.00 / Ciconiiforms / Laridae
Larusridibundus / 0.56 / 0.13 / Ciconiiforms / Laridae
Chlidoniasniger / 0.51 / 0.21 / Ciconiiforms / Laridae
Table1continued
Specie / Migrationseason index / Breedingseason index / Order / Family
Laruscachinnans / 0.79 / 0.08 / Ciconiiforms / Laridae
Sternanilotica / 1.00 / 0.39 / Ciconiiforms / Laridae
Chlidoniashybridus / 0.94 / 0.37 / Ciconiiforms / Laridae
Phalacrocorax carbo / 0.24 / 0.00 / Ciconiiforms / Phalacrocoracidae
Phoenicopterusruber / 0.17 / 0.00 / Ciconiiforms / Phoenicopteridae
Platalea leucorodia / 0.40 / 0.09 / Ciconiiforms / Phoenicopteridae
Podicepscristatus / 0.00 / 0.00 / Ciconiiforms / Podicipedidae
Podicepsnigricollis / 0.00 / 0.00 / Ciconiiforms / Podicipedidae
Tachybaptusruficollis / 0.02 / 0.00 / Ciconiiforms / Podicipedidae
Arenariainterpres / 0.00 / 0.00 / Ciconiiforms / Scolopacidae
Calidrisalba / 0.00 / 0.00 / Ciconiiforms / Scolopacidae
Numeniusphaeopus / – / 0.00 / Ciconiiforms / Scolopacidae
Limosalapponica / 0.33 / 0.00 / Ciconiiforms / Scolopacidae
Tringahypoleucos / 0.47 / 0.00 / Ciconiiforms / Scolopacidae
Tringaerythropus / 0.52 / 0.00 / Ciconiiforms / Scolopacidae
Numeniusarquata / 0.35 / 0.00 / Ciconiiforms / Scolopacidae
Calidrisferruginea / 1.00 / 0.12 / Ciconiiforms / Scolopacidae
Calidriscanutus / 0.33 / 0.42 / Ciconiiforms / Scolopacidae
Limosalimosa / 0.44 / 0.16 / Ciconiiforms / Scolopacidae
Tringatotanus / 0.36 / 0.25 / Ciconiiforms / Scolopacidae
Calidrisalpina / 0.47 / 0.12 / Ciconiiforms / Scolopacidae
Calidrisminuta / 0.61 / 0.00 / Ciconiiforms / Scolopacidae
Philomachuspugnax / 0.64 / 0.38 / Ciconiiforms / Scolopacidae
Tringanebularia / 0.69 / 0.17 / Ciconiiforms / Scolopacidae
Gallinagogallinago / 0.74 / 0.00 / Ciconiiforms / Scolopacidae
Tringaochropus / 0.88 / 0.12 / Ciconiiforms / Scolopacidae
Tringaglareola / 0.97 / 0.41 / Ciconiiforms / Scolopacidae
Lymnocryptesminimus / 1.00 / – / Ciconiiforms / Scolopacidae
Plegadisfalcinellus / 0.98 / 0.09 / Ciconiiforms / Threskiornithidae
Grusgrus / 0.45 / 0.00 / Gruiforms / Gruidae
Rallusaquaticus / 0.00 / – / Gruiforms / Rallidae
Fulicaatra / 0.11 / 0.03 / Gruiforms / Rallidae
Fulicacristata / 0.16 / 0.00 / Gruiforms / Rallidae
Gallinulachloropus / 0.27 / 0.00 / Gruiforms / Rallidae
Porphyrioporphyrio / 0.43 / 0.10 / Gruiforms / Rallidae

OurresultssupportthehypothesisthatthecreationofricefieldsaftertheSecondWorld WarintheMediterraneanRegionhelpedincreasethepopulationsofbreedingandmigrating waterbirds(CzechandParsons2002).Previousworksuggeststhatincreasesinthenumber oflittleegretsinItalyweremainlydrivenbyanincreasingavailabilityofricefields(Fasola etal.2010).Sa´nchez-Guzma´n etal.(2007)recordedpopulationsizesofinternational importance ([1% ofthe biogeographicalpopulationusingthe EasternAtlantic Flyway)for

Table2 RelationshipbetweenEuropeanpopulationtrendsandthevariablesconsideredintheanalysis

FactorMigrationperiodBreedingperiod

Estimate / F1,27 / P / Estimate / F1,28 / P
Useindex / 2.0306 / 5.85 / 0.023 / 7.1779 / 3.2 / 0.085
Migrationdistance / -0.1423 / 7.52 / 0.011 / -0.1153 / 5.93 / 0.022
Migrationdistance9migrationdistance / 0.0018 / 5.06 / 0.033 / 0.001894 / 5.01 / 0.033
Useindex9migrationdistance / -0.2516 / 3.74 / 0.063

Significantvaluesareinbold.Estimates correspondtotheregression parametersofeachfactorincludedin thefinalmodelobtainedbybackwardsregression

severalwaterbirdspeciesinthericefieldsofExtremaduraincontinentalsouthwest Spain, andsuggestthatthecreationofthesericefields couldbemodifyingthewinteringand/or feedingsitesofsomeofthewaterbirdsusingthisflyway.Ackermanetal.(2006)foundthat theincreaseinricecultivatedareainCalifornia’sCentralValleyallowedthewintering PacificGreaterWhite-frontedGeese(Anseralbifronsfrontalis)populationtoconcentrate theirhabitatuse,thusreducingthepopulationrangeandroost-to-feed distancesbetween decades.

Ourresultssuggest thatthepositiveeffectofricefieldsonwaterbird populationscould beoperatingatlargescalesformanyspeciesacrossEurope,andmaybeaglobalpattern.

Weanalyzedtheautumnmigrationperiodseparately toavoidtakingintoaccountthe monthsinwhichricefields arenotavailableashabitatforwaterbirds.Aswell,previous studieshavefoundthatricefields areparticularlyattractivetowaterbirdsduringautumn andearlywinterintheMediterranean region,wheninvertebratebiomasslevelspeakand temporarywetlandsareusuallydry(Gonza´lez-Sol´ıs etal.1996).Duringthisperiod,this habitatisofinternationalimportanceasastopover siteinthewesternMediterranean (Finlaysonetal.1992).IntheDon˜anaricefieldsthepeakinthenumberofbirdsusing thesefieldsoccursinautumn(Fig.1).

PreviousworkfoundAfro-Palearctic migrantbirdstohavenegativepopulationtrends overthelastfewdecadesincomparisontoshort-distance migrantsorresidentbirds (Sandersonetal.2006).Ourresultsshowthattherelationship betweenpopulationtrends andmigrationdistance isquadratic,withsomespecieswithextrememigrationdistances (morethanfifty degreesoflatitudebetweentheirmeanbreedinglatitudeandmeanwin- teringlatitude)havingmorepositivepopulationtrendsthanthosewithmediumvalues.The questionofwhyextremelong-distant migratingspecieshavemorepositivetrendsmerits furtherstudy,butitispossiblethattheybenefitfromhavingaverylargewinterrange (althoughtherelationship betweenextremelong-distant migrationandlargewinterrange shouldalsobeexplored). Thiswouldallowthemtocopewiththealterationsoccurring in somewintering areas,includingchangesinrainfallpatternsorvariations inprimary productivity intheSahelregion,thelatterisafactorwhichhasbeenlinkedwithdeclines andpopulationfluctuationsinanumberofEuropeanmigrantsthatwinteredthereduring thedroughtofthe1980s(DenHeld1981;Kanyamibwaetal.1990).

Shultzetal.(2005)suggestthattheBritishfarmlandbirdswhosepopulations have sufferedmostunderagriculturalintensificationarethosewiththemostspecializedresource andhabitatuseandleastcognitiveabilities.Inthissense,benefitsfromtheexploitationof ricefieldscouldbeduetoacoincidencewiththecharacteristicsofthenaturalhabitats exploitedbythesespecies,butcouldalsobeduetoahighplasticityinhabitatselection.

Inthiscase,analternativeexplanationforthepositiverelationshipbetweenricefielduse andpopulationtrendswouldbethatthemostplasticspeciesarebetteratdealingwith human-transformedenvironments.

Althoughwecan’tdemonstrate thatthereisacausalrelationbetweentheuseofrice fieldsandEuropeanpopulationtrends,ourapproachidentifiespotentialimpactsofchanges onricefieldmanagementonwaterbirdsthat shouldbetaken intoaccountbypolicymakers andfutureresearchprojects.Therelevanceofricefields asanalternativehabitatfor waterbirdsinEuropeshouldnotbeunderestimatedsincethelossofnaturalwetlands meansthattheseartificial habitatsnowrepresentanimportantproportionofthetotal habitatareasuitableforwaterbirdsintheregion.Thesurfaceareaofricecultivatedin southernEurope(Italy,Spain,Greece,andPortugal)represents 30.7%ofsuitablehabitat forwaterbirds.InSpain,thispercentagehasnowrisento49.5%(Table3).

Understandinghow ricefieldmanagementaffectsbirdsisimportanttothedevelopment ofsustainable agriculturalsystemsthatcombineeconomicallyviablefarmingandbird conservation. Infact,previousstudieshaveshownthatthereareagriculturalbenefits derivedfromhavingwaterbirdsinricefields,sincetheyimprovestrawdecomposition (Birdetal.2000)orweedcontrol(VanGroenigen etal.2003).However,changesinthe CommonAgriculturePolicy mayreduce thesurface ofricefieldsinsouthernSpainandin therestofEurope.The2003reformoftheCommonAgriculture Policyintroduced a numberofmodificationssuch asthewithdrawalofsupportforfarm production,whichhas meantareduction inthegrantsreceivedbyricefarmers, whoarenowchanging to alternativecropswiththeconcomitantlossofhabitatforwaterbirds. Inlightofourstudy, theawardingofgrantsbytheEUtoensurethemaintenanceofthesurfaceareaofrice cultivated,togetherwiththeimplementationofecologicalagricultureorthetransformation ofricefields intonaturalmarshlands,wouldseemabetteroption.Furthermore,well- designedenvironmentalagriculturalschemesinsouthernEuropemayalsoincreasethe wintersurvivalofbirdpopulationsbreedingfurthernorth(Wretenbergetal.2007).

It is important to note that we are not advocating the transformationof natural marshlandintoricefields:however, managers andpolicymakersshouldbeawarethatthe transformationofricefieldsintoothertypesofcroplands(forexample,cotton)orintosolar farmswillhaveanimportantnegativeimpactonwildlifeconservationinEurope.

Climatechangeinvolvespotentially negativeeffectsonbiodiversity.Itis important that the developmentofalternative sustainableenergyproductionbefavoured,sincethe impact ofclimatechangeneedtobemitigatedthroughlowcarbonenergyproduction. However, theindiscriminatesitingofwindorsolarfarmswithoutanydetailedassessment oftheir environmentalimpactoralternativesmayhaveimportantnegativeimpactsonwildlife. Solarfarmsareusuallysituatedonformeragriculturelandbecauseaflat surfaceisnec- essary.Nowadays,theenvironmentaldamagecausedbytheimplementationofsolarfarms isnotconsideredtoberelevantinAndalucia(PlanAndaluzdeSostenibilidadEnerge´tica

Table3 Areaofnaturalmarshes andricefields insouthernEuro- peancountries,takenfromthe

Pan-EuropeanWetlandInventory

CountryMarshes(ha)Rice

fields(ha)

%Areaof ricefields

(2004)andFAOSTAT(2004) ( WKBASE/)

Spain120,537118,00049.5

Portugal130,94325,19816.1

Italy450,563218,67632.7

Greece166,79422,41311.8

Total868,837384,28730.7

PASENER2007-2013).However, bearinginmindtheresultspresentedhere,thenegative effect that the placing ofsolarfarmsonformerrice fieldswouldhaveonDon˜ana waterbird communitiesmustbetakenintoaccount.Spainnowhasthethird-largestsolarcapacityin theworld,behindonlytheUnitedStatesandGermany (Abend 2008)andcurrentsolar- energyprojectsintheDon˜anaareawilltransformabout1,000haofricefields:this transformationofricefields intosolarfarmsmayrepresentanimportantandsilentsec- ondarylossofwetlandsinsouthernEurope.

Acknowledgments TheJunta deAndaluc´ıafundedthisstudyviatheprojectcontract‘‘Lasaves acua´ticas deDon˜anayelcultivodelarroz:lainteraccio´nentrelaagriculturaylaconservacio´ndelaszonashu´medas’’. GregorioM.ToralwasfundedbyanI3P-CSICgrantfortheformationofResearchers.The‘‘Equipode SeguimientodeProcesosNaturales’’belongingtotheDon˜anaBiologicalStation(C.S.I.C.)hasbeencol- lectingthecensusdatausedinthisstudyformanyyears.

References

AbendL(2008)Spanishsolarfirmsaccusedoffraud,NatureNews.doi:10.1038/news.2008.1326

AckermanJT, TakekawaJY,OrthmeyerDL,FleskesJP,YeeJL, KruseKL (2006)Spatialusebywintering greaterwhite-frontedgeeserelativetoadecadeofhabitatchangeinCalifornia’sCentralValley.

JWildlManag70:965–976

Acosta M, MugicaL,MancinaC,Ruiz X(1996)ResourcepartitioningbetweenGlossyandWhiteIbisesin aricefieldsysteminsouthcentralCuba.Waterbirds19:65–72

BirdJA,Pettygrove GS,EadieJM(2000)Theimpactofwaterfowlforagingonthedecomposition ofrice straw:mutualbenefitsforricegrowersandwaterfowl.JApplEcol37:728–741

BolducF,AftonAD(2004)Relationshipsbetweenwinteringwaterbirdsandinvertebrates, sedimentsand hydrologyofcoastalmarshponds.Waterbirds27:333–341

BurnhamKP,AndersonDR(2002)Modelselectionandmultimodelinference:apracticalinformation-

theoreticapproach,2ndednedn.Springer-VerlagInc,NewYork

CanepucciaAD,IsacchJP,Gagliardini DA,EscalanteAH,IribarneOO(2007)Waterbirdresponseto changesinhabitatareaanddiversitygenerated by rainfallinaSWAtlanticcoastallagoon.Waterbirds

30:541–553

CrampS,SimmonsKEL(eds)(1977–1985)BirdsoftheWesternPaleartic:HandbookoftheBirdsof

Europe,theMiddleEastandNorthAfrica.volI–IV.OxfordUniversityPress.Oxford

CzechHA,ParsonsKC(2002)Agriculturalwetlandsandwaterbirds:areview.Waterbirds25(Special

Publication2):56–65

DelanyS,ScottD(2006)WaterbirdPopulationEstimates,4thedn.WetlandsInternational,Wageningen

DenHeldJJ(1981)PopulationchangesinthePurpleHeroninrelationtodroughtinthewinteringarea.

Ardea69:185–191

DonaldP,GreenR,HeathM(2001)AgriculturalintensificationandthecollapseofEurope’sfarmlandbird populations.ProcRSoc268:25–29

DonaldP,SandersonF,Burfield I,VanBommelF(2006)Furtherevidenceofcontinent-wideimpactsof agricultural intensification on European farmland birds, 1990–2000. Agric Ecosyst Environ

116:189–196

DuncanP,HewisonAJM, HouteS,RosouxR,TournebizeT,Dubs F,BurelF,BretagnolleV(1999)Long- termchangesinagriculturalpracticesandwildfowlinginaninternationallyimportantwetland,and theireffectsontheguildofwinteringducks.JApplEcol36:11–23

ElphickC,OringL(1998)WintermanagementofCalifornianricefieldsforwaterbirds.JApplEcol

35:95–108

ElphickC,OringL(2003)Conservationimplicationsoffloodingricefieldsonwinterwaterbirdcommu- nities.AgricEcosystEnviron94:17–29

EnggassP(1968)LandreclamationandresettlementintheGuadalquivirdelta-LasMarismas.EconGeogr

44:125–143

FAOSTAT(2004)Agriculture.FoodandAgricultureOrganizationoftheUnitedNations.Rome,Italy.

FasolaM,RuizX(1996)Thevalueofricefieldsassubstitutesfornaturalwetlandsforwaterbirdsinthe

MediterraneanRegion.Waterbirds19:122–128

FasolaM,RuboliniD,MerliE,Boncompagni E,BressanU(2010)Long-termtrendsofheronandegret populations inItaly,andtheeffectsofclimate,human-induced mortality,andhabitatonpopulation dynamics.PopulEcol52:59–72

FerreroA,NguyenN(2004)Constraintsandopportunitiesforsustainabledevelopment ofrice-based productionsystemsinEurope.In:FAORiceConference,Rome,Italy:12–13

FinlaysonC,HollisGE,DavisTJ(1992)ManagingMediterraneanwetlandsandtheirbirds.International

WaterfowlandWetlandsResearchBureauandIstitutoNazionaledi Biologiadella Selvaggina.Special

PublicationNo.20,Slimbridge

Garc´ıa-Novo F, Mart´ın-Cabrera C (2005) Don˜ana: Agua y biosfera. UNESCO/Ministeriode Medio

Ambiente,Madrid

GauthierG,GirouxJF,ReedA,Be´chetA,Be´langerL(2005)Interactionsbetweenlanduse,habitatuse,and populationincreaseingreatersnowgeese:whataretheconsequencesfornaturalwetlands?Glob

ChangBiol11:856–868

Gonza´lez-Sol´ısJ,BernadiX,RuizX(1996)SeasonalvariationofwaterbirdpreyintheEbrodeltarice fields.Waterbirds19:135–142

Kanyamibwa S,Schierer A,PradelR,LebretonJD(1990)Changesinadultannualsurvivalratesina westernEuropeanpopulationofthewhitestorkCiconiaciconia.Ibis132:27–35

LemoineN,BauerHG,PeintingerM,Bohning-Gaese K(2007)Effectsofclimateandland-usechangeon speciesabundanceinacentralEuropeanbirdcommunity.ConservBiol21:495–503

LongPR,SzekelyT,KershawM,O’ConnellM(2007)Ecologicalfactorsandhumanthreatsbothdrive wildfowlpopulationdeclines.AnimConserv10:183–191

Lourenc¸oPM,PiersmaT (2008)Stopoverecologyofblack-tailed godwitsLimosalimosainPortugueserice

fields: aguideonwheretofeedinwinter.BirdStudy55:194–202

MaedaT(2005)BirduseofricefieldstripsofvaryingwidthintheKantoplainofcentralJapan.Agric

EcosystEnviron105:347–351

MurphyM(2003)Avianpopulationtrendswithintheevolvingagriculturallandscapeofeasternandcentral

UnitedStates.Auk120:20–34

NivetC,FrazierS(2004)AreviewofEuropeanwetlandinventoryinformation.Reportpreparedinthe frameworkof‘ApilotstudytowardsaPan-EuropeanWetlandInventory’.WetlandsInternational

Rendo´nMA,GreenAJ,AguileraE,AlmarazP(2008)Status,distributionandlong-termchangesinthe waterbirdcommunitywinteringinDon˜ana,south-westSpain.BiolConserv141:1371–1388

Sa´nchez-Guzma´nJM,MoranR,MaseroJA,CorbachoC,CostilloE,VillegasA,Santiago-QuesadaF(2007) IdentifyingnewbufferareasforconservingwaterbirdsintheMediterraneanbasin:theimportanceof

thericefieldsinExtremadura,Spain.BiodiversConserv16:3333–3344

SandersonF,DonaldP,PainD,BurfieldI,VanBommelF(2006)Long-termpopulationdeclinesinAfro- palearcticmigrantbirds.BiolConserv131:93–105

ShultzS,BradburyRB,EvansKL,GregoryRD,BlackburnTM(2005)Brainsizeandresourcespecial-

izationpredictlong-termpopulationtrendsinBritishbirds.ProcRSoc272:2305–2311

VanGroenigenJW,BurnsEG,Eadie JM,HorwathWR,VanKesselC(2003)Effectsof foragingwaterfowl in winterfloodedrice fieldson weed stressand residuedecomposition.AgricEcosystEnviron

95:289–296

WretenbergJ,Lindstro¨mA,SvenssonS,Pa¨rtT(2007)Linkingagriculturalpoliciestopopulationtrendsof

Swedishfarmlandbirdsindifferentagriculturalregions.JApplEcol44:933–941