iosg newsletter - 8

IOSG Newsletter - 8 -

7

4°)AdronesurveytoinventoryandcharacterizetheRed-footedBoobycolonies(Sulasula)ofFarquhar

atoll

MatthieuLeCorre,SabineOrlowski,IsabelleUrbina-Barreto,MatthewMorgan,Francois

Baguette,Pierre-AndréAdam,AnabelleCupidon&GerardRocamora

Monitoring seabird colonies canbe challenging iftimeinthefieldislimited,ifbreedinghabitatsaredifficult to reach or if the number of breedingbirds is extremely high. Furthermore, walkingamongstseabirdcoloniesforsurveyproposescangenerate significant disturbance and reducebreedingsuccess.Dronesare increasinglyused insuch situations as they can cover large areasconsiderablyfasterthantraditionaltechniques.Atthe correct altitude, they can fly over entirecolonies without disturbing birds, whilstsimultaneously collecting high resolutionphotographsthatcanbeusedforanin-labcensusafterthefieldwork.

“Barachois3”,themainnestingcolony.Farquhar Atoll (Seychelles) holds one of thelargestRed-footedBoobycoloniesintheWesternIndian Ocean. During November 2019, Island

Conservation Society (ICS) and UMR ENTROPIEconducted a pilot experiment to assess thefeasibility of using a drone to census the seabirdpopulationontheatoll.Thisresearchwaspartofa project (“Abundance, habitat selection andmovements at sea of the Red-footed Booby asinformative tools for conservation managementwithin the Seychelles Marine Spatial Plan”)conducted by ICS, UMR ENTROPIE (University ofRéunion Island) and the Zoological Society ofLondon. The project is funded by the SeychellesConservation and Climate Adaptation Trust(SeyCCAT) and supported by the Government ofSeychelles, Islands Development Company andFarquharFoundation.Additionaltechnicalsupportand supervision was provided by the IslandBiodiversity and Conservation Centre of theUniversityofSeychelles.

MatthewandFrançois,thedronepilotandco-pilot!Theexperimentwentextremelywell.Preliminaryflight tests at various altitudes showed that thedronedidnotfrightenboobies.Weconducted21flights in 5 days and covered the two maincolonies of the atoll (63 ha). These flightsproduced3168photos thatwereused tobuild amosaic orthophotograph with the dedicatedphotogrammetrysoftwareAgisoftMetashape.Wethen used the orthophoto in QGIS for imageinterpretation and nest counting. Nests werecounted manually and classified in terms ofbreeding status: adult on a nest (protecting anempty nest, incubating or brooding); chick on anest; fledging.We also pointed all resting adults(notonanest).Intotalwecounted9112breedingpairs, which is similar to previous land-basedsurveys.Wearenowcollaboratingwithanexpertin automated image analysis to automaticallyclassify and countboobynests andnon-breedingbirds.


8

5°)TrackingjuvenileSootyTernsfromBirdIsland,Seychelles

ChrisFeare,RachelBristol&Christine

LaroseAs a continuation of our investigations of the at-sea movements of Bird Island’s Sooty Terns, inAugust 2019 we deployed satellite tags on 15juvenilesinthefewdaysbeforefledging.Ouraimwastoinvestigatewherethejuvenileswerelikelyto attain independence from their parent(s), anarea that we postulated would likely be apredictably profitable feeding area where adultscouldcontinuetoprovisiontheiryoungandwherejuveniles could gain their required prey-captureskills.

Lotek6gArgosPoint

MicrowaveTelemetryInc5gsolarPTTWe used Teflon tube thigh-loop harnesses toattachfiveMicrowaveTelemetryInc5gPTTsandten Lotek 6 g ArgosPoint tags to juvenilesestimatedtobewithintwoweeksoffledging,andofminimum bodymass of 180 g. The tags weredeployed 27-30 August 2019. One of these birds

failedtofledge.Theremainingbirdslefttheislandbetween 6 and 17 September but one birdremained in the colonyuntil 26 September. Theyweretrackedforupto11months.

TracksofjuvenileSootyTernsfromBirdIsland(whitedot)upto7October2019

TracksofthethreesurvivingjuvenilesinJuly2020.BirdIslandisthewhitedot,coloureddotsrepresentthelastfixesfromthesebirdsOnleavingBird Islandtwelveofthebirdsheadednorth.Theothertwobirdsinitiallyheadedwestorsouthwestforc.150kmbeforetheytooheadednorth.Themostconspicuousbathymetric featureat the centre of their distribution shortly afterdeparture is theCocodeMer ridge,a submarinemountainrangethesouthernendofwhichliesc.300kmnorthofBirdIsland,andtheridgeextendsc. 350 km NNW. The tracks have yet to be fullyanalysed but after their initial heading the birdsthen dispersed widely, with some individualsvisiting offshore of the East African coast fromSomalia south to the Comores in the northernMozambiqueChannel,northintotheArabianSea,others ranging in the vicinity of the Chagos-Laccadive and 90°East Ridges, and one birdfeeding around the Andaman Islands and in theBayofBengal.Fromthesedataitlooksasthough


9

juveniles in their first year establish distributionpatterns similar to those discovered of adults inourearliergeolocatorstudies.Remarkably, one of our tagged birds was foundalive inlandnorth ofMumbai on 27 July 2020. Itwas emaciated and died two days later. TheMTItagwasrecoveredbutwehavestillnotbeenabletoretrieveit!Our study was funded by the SeychellesConservation and Climate Adaptation Trust(SeyCCAT)andsupportedbyBirdIslandLodge,forwhichweareextremelygrateful.6°)PetrelWatch–MonitoringRoundIsland

petrelswithcitizenscience

KirstyFranklin,JohannesChambon,NikCole,VikashTatayah,KevinRuhomaun&Malcolm

NicollEffective monitoring of seabird colonies isessential to understanding long-term populationtrends.However, for tropical seabirdswhichmaybreed all year round and often in remotelocations, visiting breeding colonies regularly islogistically difficult, can cause high disturbance,and is often expensive in terms of time andmoney. One method which has the potential toovercomesomeofthesechallenges isautomatedtime-lapsephotography.

In November 2019, we installed 10 time-lapsecameras on Round Island, Mauritius in order totrial this methodology to monitor the hybridpopulation of Pterodroma petrels that breedthere. Preliminary analysis of long-term surveydata has revealed seasonal variation in breedingsuccess, but further detailed data and analysesare required to understand why this pattern

occurs.We have therefore programmed camerasto take photos at hourly intervals in order tocapture this seasonal variation, as well as theidentificationof keyphenologicaldates (e.g. egg-laying,hatching,chickfledging)andnestsuccess.Since their deployment, these cameras havegenerated an enormous number of images, andwenowneedtoturnthiswealthofphotographicinformation into a dataset that allows us toanswer our research questions. One methodwhichhasprovedsuccessfulinspeedingupimageprocessing, particularly before AI (ArtificialIntelligence) can be trained on a new species, iscitizen science. We have therefore joined forceswiththeteamatSeabirdWatch,acitizenscienceproject previously aimed at measuringpopulationsofseabirdsintheNorthAtlantic.Images from Round Island are hosted on theSeabird Watch website (www.seabirdwatch.org)and volunteers are asked to ‘tag’ petrels byclicking on them. Volunteers are also asked toclassify each record as an ‘adult’, ‘chick’, ‘egg’ or‘other’. This latter group is used to tag theotherinhabitantsofRound Island,suchas theendemicRound Island boa, Telfair’s skinks, giant tortoisesortropicbirds.

So, if you’re intriguedby this unusual populationof petrels and fancy getting involved, or knowofsomeoneelsewhowouldbe,thenwewouldloveyourhelp!Itisincrediblyeasy,andthereisatwo-minutetutorialonthewebsite.Everyclickcounts!Acknowledgements: Thanks to the British Ornithologist’sUnion (BOU) for funding this studentship, and to theUniversityofEastAngliaandZoologicalSocietyofLondonfortheir support. Also, thanks to the Mauritian WildlifeFoundation and Mauritian Government National Parks andConservation Service for their long-term support for theRoundIslandpetrelproject.

iosg newsletter - 8

Documents