Trendsin numbersof petrelsattractedto artificiallightssuggest populationdeclinesin Tenerife,CanaryIslands AIRAMRODR´IGUEZ,1,*BENEHARO

RODR´IGUEZ2MATTHEWP.LUCAS3

1DepartmentofEvolutionaryEcology,Estacio´n

Biolo´gicade Don˜ana(CSIC),AvenidaAme´rico

VespucioS⁄N,Seville,41092,Spain

2CalleLa MalecitaS⁄N,BuenavistadelNorte,

38480,S⁄Cde Tenerife,Canarias,Spain

3KauaiEndangeredSeabirdRecoveryProject,Pacific

CooperativeStudiesUnit, UniversityofHawaii,POBox

458, Waimea,HI96796,USA

Thesecretivebreeding behaviour ofpetrels makesmoni- toringtheirbreedingpopulations challenging.Toassess population trendsofCory’sShearwaterCalonectrisdio- medea, Bulwer’sPetrel BulweriabulweriiandMacarone- sianShearwaterPuffinusbaroliinTenerife from 1990 to

2010, weused data from rescue campaigns that aim to reduce themortalityoffledglingpetrels attractedtoarti- ficiallightsasproxies fortrends inbreeding population size. Despite increases in human population size and light pollution, the number of rescued fledglings of Cory’sShearwater andBulwer’sPetrelincreasedand remained stable, respectively, whereas numbersof res- cued Macaronesian Shearwaterssharply declined. Inthe absence of more accurate population estimates, these resultssuggesta worryingdeclineintheMacaronesian Shearwater’s breeding population.

Keywords: Canary Islands,lightpollution,popula- tionsize,rescuecampaigns, seabirds.

TheProcellariiformes(including petrels andshearwaters, hereafterpetrels) comprise one ofthe most endangered birdtaxa,as manyspecieshaveundergonesubstantial declinesinrecent times(Butchartetal.2004). Themain causesoftheir populationdeclinesareintroducedpreda- tors, interactionswith fisheries and nesting habitat loss

*Correspondingauthor.

Email:

(Carlile etal.2003, Bourgeois Vidal 2007, LeCorre

2008). Some conservationactions are being developed forcharismatic andemblematicspecies(e.g.Carlileetal.

2003). However,forsmallerspecies,populationdeclines often gounnoticedandonlyafewrecovery programmes havestarted sofar(e.g.BourgeoisVidal2007). Moni- toring and estimating the breeding population size of petrels ischallenging because oftheir secretive breeding behaviour; the majority are nocturnal, nest in under- ground burrows that are difficult to count and often occupy inaccessible, remote nesting sites (Bretagnolle etal. 2000, Gregory etal. 2004). Therefore, in the absence ofaccurate censusdata, other indicesorproxies for populationsizemight be used to assesspopulation trends (Gregory etal.2004).

On archipelagos worldwide, thousands of fledglings ofdifferent petrel speciesareattractedtoartificiallights during their firstflightsfrom nest-burrowsto the sea,a phenomenon called ‘fallout’ (Reed etal. 1985, Telfer etal.1987, LeCorre etal.2002, Rodríguez Rodríguez

2009, Miles etal. 2010, Rodrigues etal. 2011).

Grounded birdsarevulnerable tostarvation, predation, dehydration and collision with vehicles. Rescue cam- paigns have been carried out inmany places (LeCorre etal.2002), andmost ofthe rescued birds (>90%)are later released into the wild (Telfer etal.1987, Ainley etal.2001, LeCorre etal.2002, Rodríguez Rodríguez

2009, Milesetal.2010, Fontaine etal.2011).

The Canary Islands are an importantbreeding area for petrels in the northeastern subtropical Atlantic (Arcosetal.2009). Atleastsevenpetrel species(includ- ingshearwatersandstorm-petrels)breed regularlyinthe archipelago (Lorenzo 2007). Tenerife Island isthe larg- est, highest and the second most inhabited(2034 km2, up to 3718 m asl, over 900000 inhabitants) of the Canary Islands(ISTAC 2011) andishome tosixbreed- ing petrel species (Lorenzo 2007). There are no long- term monitoringprogrammeson the Canary Islands to detect population trends of petrels (Lorenzo 2007).

Following the work of Ainley etal. (2001) on the endangeredNewell’s Shearwater Puffinus newellion Kauai(Hawaiian archipelago),weuseddatafromrescue campaigns toevaluate thepopulationtrends ofthethree mostcommonpetrel speciesinvolvedinfalloutonTene- rife: Cory’s Shearwater Calonectrisdiomedea, Bulwer’s Petrel Bulweria bulweriiandMacaronesian Shearwater Puffinus baroli(formerly Little ShearwaterPuffinus assimilis)(Rodríguez Rodríguez 2009). At present, these three speciesareprotectedunder theNational and Regional Catalogue ofThreatenedSpecies (Spanish and Canarian Governments)and they are listed as nation- allyvulnerable (Cory’sShearwater) andendangered (Bulwer’s Petrel and Macaronesian Shearwater)because their populationsarethoughttobeindecline (Madroño etal. 2004). Cory’s Shearwater is the most abundant Canarian seabirdandiswidespreadontheislands,breed-

ingincliffs, slopesandravines,oftenseveralkilometres inland. Bulwer’s Petrel and Macaronesian Shearwater breed inmarine rocksandseacliffs,although thefinding of someBulwer’sPetrelsathighaltitude andseveral kilometres inland suggeststhat this species could breed inhighinterior mountains.Cory’sShearwater andBul- wer’sPetrel breed onallthe Canarian islands.No accu- rate estimates of the breeding populationfor Tenerife are available, but in the period 1997–2003,ca. 2000–

3000, 400and70breeding pairshavebeenestimatedfor Cory’sShearwater, Bulwer’sPetrelandMacaronesian Shearwater, respectively(Lorenzo 2007,Rodríguez& Rodríguez 2009andreferences therein).

Here, wereport the numberofrescued fledglingsof these three petrel species during 21years on Tenerife, Canary Islands,usethese toassesspopulationtrends and propose appropriateconservationmeasures.

MATERIAL ANDMETHODS

We collected data on the numbersof fledglingsrecov- ered onTenerife Islandbythe rescue campaigns during

1990–2010forthe three petrel species,followingRodrí- guezandRodríguez (2009).Therescueprogrammeisan island-wide public conservation programmethat has recovered disorientated petrels attracted to artificial lights on their firstflightsfrom the nest-burrowout to the seasince1990 (Anonymous1995). Thisprogramme hasbeenimplemented bytheLaTahonillaWildlife RehabilitationCentre and funded by the local govern- ment (Cabildo Insular deTenerife).The numberofres- cued fledglingsofeach species wasrecorded every year foreachmunicipalityseparately.

Thenumberofrescued birdsmight depend onseveral factors (Ainley etal. 2001). First, the annual rescue effort probably varied because it is a volunteer programmeanddepends ontheefficacyoftheawareness campaign conductedeach yearbefore the fledglingsea- son,aswellasthenumberofprevious campaigns carried out (LeCorre etal.2002, Salamolard etal.2007). Asit was not possible to know how many people were involvedintherescuecampaignandtheir levelofaware- ness each year, we used human populationsize, taken from ISTAC (2011),asan informal assessment of the size of the rescue campaign. Secondly, the amount of lightpollution couldvarythrough thestudyperiod. To evaluate the evolution of light pollution in Tenerife Island, three stable average artificial light layers from

1992, 2000 and 2008 (911.25·911.25 m resolution; pixel values ranking from 0 to 63 relative units) were obtained from the National Geophysical Data Center (USA)( loadV4composites.html),following amethodology simi- larto that ofRodrigues etal.(2011).The annual mean valueofstableartificiallightswasextractedusingARCGIS (version 9.2; ESRI, Redlands, CA, USA) for the three

layersandthe observed increase wascalculated.Inaddi- tion to measuredlight, the grosselectricity production was also evaluated (ISTAC 2011). Thirdly, the lunar phase has animpact on the numberofretrieved fledg- lings(Ainley etal.2001), sowe identified years where the peak fledgling period coincided with a full moon. Finally, the numberof retrieved fledglings is likely to reflecttheoverallsizeof thebreedingpopulationand breeding success. Thus the number of fledglings pro- duced cannot exceed the numberofbreeding pairs(Day etal.2003) (petrels layasingleeggperbreeding season; Warham 1990).

RESULTS

Thethree petrel speciesshowed different trends in recoveries throughout the study period (Fig.1). The number of rescued Cory’s Shearwater increased, Bul- wer’s Petrelrecoveriesremained stable,butrecoveries declined in Macaronesian Shearwater. The Canarian human populationincreased steadily from 623823 in

1991to906854in2010.Similarly,thegross electricity productionincreased from1431 Gigawatt hours in1991 toa peakof3550Gigawatt hoursin2008andslightly decreased to 3357 Gigawatt hours in the last 2years (2009 and2010). Humanpopulationsizeandgrosselec- tricity production were highly correlated (r=0.98, P0.001, n=20). Asaconsequenceofthe growth in human populationand electricity production,light pol- lution intensity (and consequently area affected) increased by 22.6% from 1992 to 2000 and by 13.1% from2000to2008(Fig.2).

DISCUSSION

Althoughchanges inthe numberofrescued birds over thelasttwodecadescouldprovideaproxyforpopulation trends, theycanbebiasedbyseveralfactors(i.e.variation inannualeffortof therescuecampaign,lightpollution intensity andmoon phase). Asaconsequenceofincreas- ingawareness, the numberofrescued birds islikelyto haveincreased during the firstfewrescuecampaigns (Le Correetal.2002). Indeed,onTenerifeinthethirdrescue campaign (1992) more than twice asmany birds were rescued than inthe firsttwo seasons(1990–91;Fig.1). Thus, wecanassume that atleastinitiallythe awareness levelhasincreasedduringourstudyperiod.In addition, human population, grosselectricity consumptionand urbanizedareahavenotably increased inTenerife, which could haveincreased both rescue effort andlight pollu- tion. From 1992 onwards, the use of shielded lights increasedandmercuryandhalogenlights begantobe substitutedwithlowsodium vapour lightsasimposed by the Sky Law (Ley del Cielo 31⁄1988, Real Decreto

243⁄1992), whichaimstoprotecttheresearchactivityof the Astrophysical Institute of the Canary Islands by

Figure1.Number ofrescued fledglingsofthreepetrelspecies collected bytherescuecampaigns onTenerife(Canary Islands). Opencirclesindicateyearswherethepeakfledging periodcoincidedwithafullmoon(±4days)(seeRodr´ıguez& Rodr´ıguez2009).

reducing lightpollution.However,accordingtoouranal- ysis this law was not sufficient to reduce the overall extent of the area affected by light pollution and its intensity.Breedingsuccesscanvary fromyeartoyear, although these annual variations should not skew long- term fledglingfallout trends, unlesslowbreeding success hasbeensustained overalongperiod oftime. Thisthen wouldresultinasmallernumber ofyoungbirdsbeing recruitedasbreeders in the following years and in the longterm wouldresultinapopulation decline.Lastly, numbersof rescued fledglings can be lower when the peakfledglingperiod coincideswith afullmoon (Ainley etal.2001), but inour dataitdoesnot appear toaffect long-termtrendsofthefallout(seeFig.1).

Ifbreedingpopulationsizeis stable,onewouldexpect thatthenumbers offledglingsrescuedwouldcorrelate positively with theannualeffortoftherescuecampaigns andtheincrement oflightpollution.Undercurrent conditions onTenerife, with increasingrescue effort and lightpollution,evenadecliningpopulationcouldshowa positive trend in the numberof rescued fledglings. A declineinthenumber ofrescuedfledglings,however, suggests a worrying situation because this indicates a decline in the breeding populationand⁄or adecline in breedingsuccesssustained overseveralconsecutive years.

The three petrel species showed different trends in the numbersofrescued fledglingsduring the study per- iod. Inthe caseofCory’sShearwater,wecannot confi- dently predict areal populationtrend. Population size might be increasing given that at least aproportion of thethousandsofrescued fledglingsduringthelastrescue campaigns would have been recruitedasbreeders, but noaccurate informationonrecruitmentofreleasedbirds isavailable. Anecdotal data, based on unoccupiedbur- rowsduringthelastfewyearsandhighnumbersofadult birds predatedbyferal cats (A. Rodríguez B.Rodrí- guez pers. obs.), as well as the assumptions of other authors (Madroño etal.2004, Lorenzo 2007), contradict the apparentpopulationincreasefromthisstudy. Thisis alsosupportedbytheviabilitymodels recently published forthisspeciesintheAzoresarchipelago, which requires unrealistic demographic parameters to permit popula- tion stability (Fontaine etal.2011). Furthermore, lights fromnewurbanizedareasthat appeared onTenerife dur- ingthe study period may attract individuals from colo- niesnot previously affected (Figs2and3), which could contributetotheobserved increaseofrescued birds.This patternoflightattractionspreading tonewly developed urban areaswasalsoseenonKauai(Ainley etal.2001). Wealsobelievethat the levelofpublic awareness isnot completelysaturated.Therefore,ifthe breeding popula- tionis actuallydecliningwhilecitizenship awarenessis increasing, we expect to find the number of rescued Cory’sShearwaterfledglingstopeakwithin thenextfew years, asreportedfor the declining population of Newell’s Shearwater (Ainley etal. 2001, Day etal.

2003, Duffy2010).

ForBulwer’sPetrelandMacaronesian Shearwater,the situation appears lessfavourable. ForBulwer’sPetrel, the patternofrecoveries wasstable during the years1992–

2009. The most important breeding colonies for this species are located on shoreline marine rocks, places where fledglingseasilyreach the seaand light pollution hasincreasedonlymildlyincomparisonwith other areas (e.g. touristic cities; Figs2and 3). However,the stable number ofrescuedbirdscouldstillindicatethatthe populationisdeclining, ashasbeen suggested (Madroño etal.2004, Lorenzo 2007).

ForMacaronesian Shearwaters, thesituation appears critical, as the pattern of recoveries suggests a sharp

Figure2.Levels ofartificiallightintensity duringthenightonTenerife Islandin1992,2000and2008.Darkershadingindicatesmore illuminatedareas.Linesindicate municipallimitsandcoastline.Artificial lightintensitywastakenfromsatelliteimagesbytheNational GeophysicalDataCentre(see text).

Figure3.GeographicaldistributionofbreedingcoloniesandthenumberoffledglingpetrelsattractedtoartificiallightonTenerife Islandduring 1998–2010. Certaintyofbreedingatpossiblecolonylocationsineach5·5-kmsquareisgivenasconfirmed,probable andpossible(modifiedfromLorenzo2007).

populationdecline, likemany petrel species inthis and other archipelagos (Day etal. 2003, Le Corre etal.

2003, Bourgeois Vidal 2007, Rodríguez etal.2008). The change to more efficient lighting systems sincethe SkyLawcameintoforcecouldbeanon-mutuallyexclu- siveexplanationforthe negative trend inthe numberof rescued birds, asMacaronesian Shearwatermight not be attractedto the modified light sources (see Reed etal.

1985, Salamolard etal.2007). Passerines, for example,

are affected differently by different qualities of lights (Poot etal.2008). However,this seems unlikely to be the completeexplanationand apopulationdecline has also been suggested by limited field observations (Madroñoetal.2004, Lorenzo 2007, pers.obs.).Forthe Newell’s Shearwater population on Kauai, additional datafromradarcountsandpopulation viabilitymodels confirmed the population decline initially suspected from the decreasing numbersof rescued birds (Ainley etal.2001, Dayetal.2003).

Trends in numbersof rescued birds are affected by severalsourcesoferror(seeabove), but giventhelackof accurate censusdataonthese petrel populations,aswell asthe secretive breeding behaviour andthe inaccessibil- ityoftheir breeding sites,thisiscurrently theonlyavail- ableinformation, atleastforBulwer’sPetreland Macaronesian Shearwater. However,tocarryout more formalanalysesin futurestudies,wewouldneedto(1) implement a massive ringing effort of nestlings every year to assess accurately the percentage of fledglings affected bylight pollution;(2) determineburrow occu- pancyrateandcorrelate thiswith thenumberofrescued fledglings;(3) survey citizens’ awareness; and (4) begin other programmestoestimate populationtrends suchas radar surveys (Day etal.2003). Because preventionis ofteneasierthancure,forMacaronesian Shearwater we recommendthe immediatedesign, publicationandexe- cution ofaconservationplanasrequired bySpanishand Canarian law.Thisconservationplanshould bebasedon three keypoints: (1)control ofcatsandratsatcolonies; (2) enhancement of the rescue campaign during the fledgingmonths toimprove citizens’awareness and improve rehabilitationprocedurestoreduce the mortal- ity of rescued birds by providing food and liquid, and testwaterproofpropertiesofplumage; and(3)execution ofadetailed surveyprogrammeonthedistribution,pop- ulationsizeandbreedingparameters toevaluateits demographicdynamics.

Specialthanks toallthe anonymouspeople who kindlyhelped rescuethebirds,andtothestaffon‘La Tahonilla’Wildlife RehabilitationCentre(Cabildo Insular de Tenerife),who pro- videduswithdataof dazzledbirds.ThankstoSolnyA.Adal- steinsson, LuisCadahía, David G. Ainley, Morten Frederiksen, RuediNagerandtwoanonymous reviewersforcomments and suggestions onthe earlydrafts ofthis manuscript.A.R. wishes tothank NickD.Holmes andAndrea Erichsenfortheir support

during my research stay on Kauai. A.R. was supportedby an

I3Ppre-doctoralfellowship fromtheCSIC.

REFERENCES

Ainley, D.G.,Podolsky, R.,Nur,N.,Deforest, L.Spencer, G.A.2001.Statusandpopulation trendsoftheNewell’s Shearwater onKauai:amodelforthreatenedpetrelson urbanizedtropicaloceanicislands.Stud.AvianBiol.22:

108–123.

Anonymous.1995.Rescatedepardelasen Canarias.La Gar- cilla92:6.

Arcos, J.M., Be´cares, J., Rodr´ıguez, B. Ruiz, A. 2009.

A´reasImportantesparalaConservacio´n delasAvesMari- nasen Espan˜a.LIFE04NAT⁄ES⁄000049.Madrid:Sociedad

Espan˜oladeOrnitolog´ıa(SEO⁄BirdLife).

Bourgeois,K.Vidal,E.2007.TheendemicMediterranean Yelkouan Shearwater Puffinus yelkouan: distribution, threatsanda pleaformoredata.Oryx42: 187–194.

Bretagnolle, V., Attie´, C. Mougeot, F. 2000. Audubon

ShearwaterPuffinuslherminieriinReunio´nIsland:behav- iour,census,distribution,biometricsandbreeding biology. Ibis 142:399–412.

Butchart,S.H.M.,Statterfield,A.J., Bennun, L.A., Shutes,

S.M.,Akc¸akaya,H.R.,Baillie,J.E.M.,Stuart,S.N.,Hilton- Taylor,C.Mace,G.M.2004.Measuringglobaltrendsin thestatusofbiodiversity:redlistindicesforbirds.PLoS Biol. 2:e383.

Carlile, N.,Priddel, D.,Zino, F.,Natividad, C.Wingate, D.B.2003.Areviewoffoursuccessful recovery pro- grammesforthreatened,sub-tropical petrels.Mar.Ornithol.

31: 185–192.

Day,R.H.,Cooper,B.A.Telfer,T.C.2003.Decline of Townsend’s(Newell’s) Shearwaters(Puffinusauricularis newelli)onKauai,Hawaii.Auk 120:669–679.

Duffy, D.C.2010.ChangingseabirdmanagementinHawai’i: fromexploitationthroughmanagement torestoration. Waterbirds33: 193–207.

Fontaine, R.,Gimenez, O.Bried, J.2011.Theimpactof introducedpredators,light-induced mortalityoffledglings andpoaching onthedynamicsoftheCory’sShearwater (Calonectrisdiomedea) population fromtheAzores,north- easternsubtropicalAtlantic.Biol.Conserv.144:1998–2011.

Gregory, R.D.,Gibbons, D.W.Donald, P.F. 2004.Bird censusandsurveytechniques.InSutherland,W.J.,New- ton,I.Green,R.E.(eds)BirdEcologyandConservation: AHandbook ofTechniques:17–56.Oxford:OxfordUniver- sityPress.

ISTAC.2011.InstitutoCanario deEstad´ıstica.Gobiernode Canarias.Availableat: istac (accessed1 March2011).

LeCorre,M.2008.Cats,ratsandseabirds.Nature451:134–135.

LeCorre, M.,Ollivier, A.,Ribes, S.Jouventin,P.2002.

Light-induced mortality of petrels: a 4-year study from

Re´unionIsland(IndianOcean).Biol. Conserv.105:93–102.

LeCorre,M.,Ghestemme,T.,Salamolard, M.Couzi,F.-X.

2003.RescueoftheMascarenePetrel,acriticallyendan- geredseabirdof Reu´nionIsland,IndianOcean.Condor

105:387–391.

Lorenzo, J.A.(ed.)2007.AtlasdelasAvesNidificantesenel Archipie´lagoCanario(1997–2003).Madrid:Direccio´nGen- eraldeConservacio´n dela Naturaleza-SEO⁄BirdLife.

Madron˜o,A.,Gonza´lez,C.Atienza,J.C.(eds)2004.Libro RojodelasAvesdeEspan˜a.Madrid:Direccio´nGeneral parala Conservacio´n dela Biodiversidad-SEO⁄BirdLife.

Miles,W.,Money,S.,Luxmoore,R.Furness, R.W.2010.

Effectsof artificiallightsandmoonlightonpetrelsat St

Kilda.Bird Study57: 244–251.

Poot, H., Ens, B.J., de Vries, H., Donners, M.A.H., Wernand,M.R.Marquenie,J.M.2008.Green lightfor nocturnallymigratingbirds.Ecol.Soc.13: 47.

Reed,J.R.,Sincock, J.L.Hailman,J.P.1985.Lightattrac- tioninendangered Procellariiformbirds:reductionby shieldingupwardradiation.Auk102:377–383.

Rodrigues, P., Aubrecht, C., Gil, A., Longcore, T. Elvidge, C.2011.Remote sensing tomapinfluenceoflight pollutiononCory’sShearwaterin Sa˜oMiguelIsland,Azores Archipelago.Eur.J.Wildl.Res.doi:10.1007/s10344-011-

0555-5.

Rodr´ıguez, A.Rodr´ıguez, B.2009.Attractionofpetrelsto artificiallightsintheCanaryIslands: effectofthemoon phaseandageclass.Ibis 151:299–310.

Rodr´ıguez, A., Rodr´ıguez, B., Barone, R., Pe´rez, B. Herna´ndez,A.2008.Statusandconservationrequirements ofManxShearwatersPuffinuspuffinus onTenerife(Canary Islands).Alauda76:72–74.

Salamolard, M.,Ghestemme, T.,Couzi,F.-X.,Minatchy, N.

LeCorre, M.2007.Impactsdese´clairagesurbainssur les petrels de Barau, Pterodromabarauisur I’lle de la Re´unionetmeasurespourre´duirecesimpacts.Ostrich78:

449–452.

Telfer, T.C.,Sincock, J.L., Byrd, G.V.Reed,J.R.1987.

Attraction of Hawaiian seabirds to lights: conservation effortsandeffectsofmoonphase.Wildl.Soc.Bull.15:

406–413.

Warham, J..1990.ThePetrels:TheirEcologyandBreeding

Systems.London:AcademicPress.

.

ª2011TheAuthors

Ibisª2011British Ornithologists’Union