Notesandrecords - PDF document

1 Notesandrecords Evidenceforlocaldeclines
Notesandrecords EvidenceforlocaldeclinesinTanzania’spuku antelope( Kobusvardoni Livingstone,1857) populationbetween1999and2003 ColinBonnington 1 *,MarkD.Steer 2 ,JulienLamontagne- Godwin 2 ,NishaOwen 2 andMatthewGrainger 1 1 Frontier-TanzaniaSavannaResearchProgramme,POBOX 9743,DaresSalaam,Tanzania, 2 SocietyforEnvironmental Exploration ‡ Frontier,50-52RivingtonStreet,EC2A3QP, U.K. Introduction TheKilomberoValleyinTanzaniaholds c. 75%ofthe world’spukuantelope( Kobusvardoni ),whichisanesti- mated50,000–60,000individuals(East,1998;Jenkins, Maliti&Corti,2003).Globally,survivalofthisspeciesis closelyrelatedtothispopulation(Rodgers,1984;Tanzania WildlifeConservationMonitoring,1999),andtherefore studiesinvestigatingchangesinpukunumberinthevalley areimportantfordevelopingconservationmanagement strategiesforthisspecies. Pukuisa‘NearThreatened’species(IUCNSSCAntelope SpecialistGroup,2008),whichishabitatspecic,occu- pyinggrasslandsnearpermanentwaterwithinsavannah ingseasonalinundationoftheoodplainduringthelong rainyseason(March–May),pukumoveintohabitatonthe oodplainperiphery,suchasmiombowoodland(Jenkins etal .,2002). Anumberofhumanactivities,includinglivestockhus- bandry(Corti etal .,2002;Bonnington,Weaver&Fan- ning,2007)andsettlement(Haule,1997),are encroachingonthesefavouredoodplaingrasslandsinthe KilomberoValley,andotheractivitiessuchaspoaching (Jenkins etal. ,2002),negativelyimpactthepukupopu- lationofthevalleydirectly. Starkey etal .(2002)statedthatthereisanurgentneed forsurveystobeconductedintheKilomberoValleyin openareasofsuitableforaginghabitatandalsoinsuitable peripheralhabitatnearhumansettlementtoassessuc- tuationsinpukunumberinareasexperiencingdifferent levelsofhumanimpacts.Inthispaper,wedescribetherst ground-basedstudyofpukuintheKilomberoValley assessingannualchangesintheabundanceofthisspecies attwosuchlocalities.Wereportasignicantdeclineinthe relativeabundanceofpukuatbothlocationsovera5-year periodanddiscussthepossibleimplicationsforthecon- servationofthisspeciesinthisstronghold. Materialsandmethods 2 , issituatedinsouthernTanzania(Fig.1)approximately 320kmfromthecountry’slargestcity,DaresSalaam. TheinnerpartofthevalleyistheKilomberoGameCon- trolledArea(KGCA),anditconsistsofaseasonalood- plainfringedbymiombowoodlandandfarmland,with fragmentsofEasternArcevergreenforestathigherele- vations(InternationalInstituteForEnvironmentand Development,1992).WithinTanzanianGCAs(e.g. KGCA),unauthorizedhuntingisprohibited,butlanduse changes(e.g.livestockhusbandryandagriculture)arenot regulated(Haule,Johnsen&Maganga,2002). OurstudieswereconductedintheKGCAbetween1999 and2003,neartheareasofMofuandMadabadaba.Mofu andMadabadabaareapproximately55kmfromthetownof Ifakara(8  08 ¢ 20.48 ¢¢ S.36  40 ¢ 54.58 ¢¢ E.)andaresituated approximately4kmapart(Fig.1).BothMofu(8  32 ¢ 43.69 ¢¢ S. 36  23 00.60 ¢¢ E.)andMadabadaba(8  32 ¢ 45.28 ¢¢ S.36  25 ¢ 27.90 ¢¢ E.)arelocatedinthesouthernregionoftheood- plain,withMadabadabaclosetoavillageandMofusituated furtherintotheoodplain(awayfromvillages). Thestudysitesweresurveyedusingdirectobservation xed-pointtransects(markedusingaGPS).Twentyxed pointswerepositionedat0.5-kmintervalsalongtwo 10-kmtransectsinMofu(closetoatouristhuntinglodge), andtenxedpointswerepositionedat0.5-kmintervals alongtwo5-kmtransectsatMadabadaba.Ateachxed pointalongthetransects,withina0.25-kmradiusofeach point,thenumberofpukuobservedwasrecordedfora periodof10minusingbinocularsandarangender. Pukuwererecordedassoonastheobserversarrivedat * Correspondence: E-mail:c.bonnington@shefeld.ac.uk  2010SocietyforEnvironmentalExploration ‡ Frontier.Journalcompilation  2010BlackwellPublishingLtd, Afr.J.Ecol.

2 1 eachpointtoaccountforpukumovementsawa
1 eachpointtoaccountforpukumovementsawayinresponsetoobservers(thusminimizingtheimpactsofpotentialchangesinpukuresponsebetweenseasonsandyears).Alltransectswerewalkedbyasurveyteamcon-sistingofveobservers.Itwasassumedthatindividualswerenotcountedattwodifferentpointsalongthesametransectaspukumovementsduringtheshorttimeateachpointcountareminimal,andevenwhenstartled,ightdistancesareshort(.25m;Estes,1991).AtMadabadaba,transectswerelimitedto5kmbyawatercoursebarrier.Thetwotransectsateachsitewerespaced3kmapart,andallwereorientatedonaNorth–Southcompassbear-ing.Transectswereconductedatsunriseorassoonaspossiblethereafter(07:00h)andwereterminatedat10:30h.Asvegetationinthesurveyareaishomogeneousinthedryseason(June–November)andasaconsequenceofshortdetectiondistancesnegatingtheinuenceofvegetationstructureonvisibility,detectionprobabilitiesremainedconstantbetweensurveyseasonsandyears.AtMofu,datawerecollectedalongtwotransectsinAugustandSeptemberin1999,2000and2001;andalongthesametransectsinOctoberandNovember1999,2001and2003.AtMadabadaba,datawerecollectedalongtwotransectsinOctoberandNovemberin1999,2001and2003;andalongonetransectinAugustandSeptemberin1999and2000.ThedatacollectedbetweenAugustandSeptemberandOctoberandNovemberateachsite,ineachyear,werepooledtogethertoproduceanoverallpukurelativeabundanceindex(N).Nisthemeannumberofpukuobservedper10minwithinthesurveyareaof0.196km(0.25km*0.25km*)foreachsurveyyear.SurveyswerenotconductedateitherlocationinDifferencesinrelativeabundancewereanalysedusingagenerallinearmixedmodelthatincorporated‘Season’asaxedfactor,‘Location’asarandomfactorand‘Year’asacovariate.‘Transect’and‘SurveyPoint’werealsoincluded,respectivelynestedwithinLocationandTran-sect.ThemodeltestedforaninteractionbetweenLocationandSeason.Abundancedatawere1(x+1)transformedtoensureresidualvalueswerenormallydistributed.AllanalyseswerecarriedoutusingthestatisticalpackageMINITABver.15(MinitabInc.,StateCollege,PA,USA).PukurelativeabundancewassignicantlyhigherinMofucomparedtoMadabadabaacrossallsurveyyears=20.01,=0.006).Pukunumbersalsoshowedasignicantdeclineoverthe5-yeartimeperiodwhencombiningthevaluesforbothlocationsacrossallsurveyyears(=10.84,=0.001),withtherelativeabundancevaluesforthisspeciesparticularlylowin2003(Fig.2).Themodelreturnedanvalueof0.3775,indicatingthatareasonableamountofvariationwithinthedataisexplainedbythesevariables.NeitherSeason=2.01,=0.389)northeinteractionbetween Fig2Relativeabundanceofpuku(N)atMadabadaba(presentedonFigureas‘Mad’)andMofuovera5-yearperiod.Dataarebacktransformed.Errorbarsshow95%condenceintervals.Datapointsareoffsetalongthe-axisforclarity. MofuIfakara00101020 km20 miDar es SalaamIringa00100100200 km200 mi Fig1LocationoftheKilomberoValleyinTanzania.Insetillus-trateslocationofstudyareasinrelationtothetownofIfakara;KilomberoRiverisshownasadarkline,roadsaslighterlines.AdaptedfromanimagekindlysuppliedbyKVTC.ColinBonningtonetal.2010SocietyforEnvironmentalExplorationFrontier.Journalcompilation2010BlackwellPublishingLtd,Afr.J.Ecol. SeasonandLocation(=0.7,=0.404)signi-cantlyinuencedpukuabundance.Wereportasignicantdeclineinrelativelocalpukuabundanceoverthe5-yearsurveyperiodatMofu,whileMadabadabahasasmallerpukupopulationbutonewhichalsoexhibitsasimilardecline.Itismostlikelythatthesedeclinesareattributedtoanthropogenicfactors.Rodgers(1982)statedthatrisinglakelevelsintheRukwaValleyinsouth-westernTanzaniaresultedinareductionoflargemammalpopulations(includingpuku)asmuchofthesuitableforaginghabitatwasooded.Unassessed,stochasticenvironmentalfactors(suchaschangesinoodingregimesandprecipitation)arelikelytohavealowerinuenceonpukunumberthangrowinganthro-pogenicpopulationpressures,asourresultsshowacon-sistentdeclineinthisspeciesoverthelengthofthestudy.Pukuadoptaexiblesocialstructure,wh

3 ichresultsina‘freepassageofanimalswithin
ichresultsina‘freepassageofanimalswithinandbetweenpatches’etal.,1998).Assuch,pukuwouldbeexpectedtomovearoundsuitablehabitatwithintheKilomberooodplainduringthesurveyperiod,withdispersalbetweenthetwosurveyareasapossibility.Furthermore,acaveatthatshouldbeconsideredisthepossibilitythatourresultsareindicativeofatemporarydeclinebecauseofindividuals(particularlyfemalesasmaleswilloftenremaininanareadefendingaterritory;Goldspinketal.,1998)movingintosuitableadjacentun-surveyedhabitat.How-ever,counteringthis,our5-yearstudysupportsrecentndings(Bonnington,Weaver&Fanning,2007)whichdocumentedthatthehabitataroundthehuntinglodgeatMofuistraditionallyheavilyusedbypuku.Thisindicatesthatdirectpopulationpressuresfromanthropogenicsourcesaremostlikelytobethemajorfactordrivingdeclinesratherthananynaturaldispersal.Therefore,anydeclineinnumberatthisoptimallocationraisesconser-vationconcerns.ThecontinuallylowpukuabundanceatMadabadabaislikelycausedbythehighhumanpopulationinthearea,whichincreasesby.2.5%peryear(Meshack,2001).Humanandlivestockencroachmentclosertotheood-plainhasbeenreportedthroughouttheKilomberoValley(Haule,1997)andthispressureislikelytobelimitingthepukunumberinMadabadaba.AsimilarsituationhasbeenreportedbyWaltert,Chuwa&Kiffner(2009)atLakeRukwa,Tanzaniawherethesmallpukupopulation(ofglobalpopulation;Jenkins,Maliti&Corti,2003)avoidlandontheperipheryoftheRukwaGameReservepossiblyasaresultofhighpastoralistactivity.Ifourresultsrepresentalong-termdecline,thedecliningpukuabundanceatMofuisofprimaryconcern.PukunumbersatMofuhavebeenmaintainedinthepastbytheinuenceoftheMofuhuntinglodge(Bonnington,Weaver&Fanning,2007),whichhashadapositivebutlocalizedeffect,evidentfromtheinitiallyhighernumbersofpukupresentinthisstudy.However,thesubsequentdeclineswereportheresuggestthateventhelocalizedprotectionofthelodgeisinsufcientinthefaceofhumanencroachmentinthearea.IfthisisasourcepopulationindeclineatMofu,similardeclinesinotheroptimalareasintheoodplaincouldhaveseriousconsequencesforpersistenceofpukuintheKilomberoValley.Unregulatedhumanencroachmentandsubsequentactivities(e.g.habitatconversion,illegaloff-takesandlivestockhusbandry)arethoughttoadverselyimpactpukuintheKilomberoValley(Jenkins,Maliti&Corti,2003),yetthestabilityofthespeciesintheKGCAisinferredfromaerialsurveydata(Jenkinsetal.,2002;Jenkins,Maliti&Corti,2003).Ourstudyisthersttohavespecicallyhighlighteddecliningpukuabundanceintwolocationswithinthevalley,andwesurmisethatanthropogenicactivitiesaresignicantdriversforthistrend.AdecliningpukupopulationhasalsorecentlybeenreportedbyWaltert,Chuwa&Kiner(2009)atLakeRukwa,TanzaniawheresimilaranthropogenicpressurestothosewitnessedintheKilomberoValley(i.e.livestockhusbandryandsubsequenthabitatdegradation)havebeenblamedforthistrend.However,Dipotso&Skarpe(2006)foundnosuchdeclineinthepukupopulationonadjacentlandtotheChobeRiverinnorthernBotswanabutdidreportashiftintheconcentrationofthepopulationeastwards,whichtheyattributedtodisturbancefromhumanactivitiesinthewestoftheriver.Humanimpacts(particularlyillegaloff-takes)havehis-toricallybeenblamedforthereductionofAfricanKobantelopespecies(e.g.lechwespp.andpuku),withearlyrecordsofKobpopulations(Hughes,1933citedfromCotterill,2005)beingdecimatedbyhumanhuntingbythemid-20thcentury(Vesey-Fitzgerald,1961citedfromCotterill,2005).Weurgentlyrecommendimplementationofalarge-scaleconservationprogrammetoquantifyhowpukuusehabitatspace,tohighlightstrongholdswithintheKilomberoValleyforthespecies.AcombinationofPukupopulationinKilomberoValley2010SocietyforEnvironmentalExplorationFrontier.Journalcompilation2010BlackwellPublishingLtd,Afr.J.Ecol. recurrentaerialsurveys(describedbyJenkins,Maliti&Corti,2003)andradio-trackingprojectswouldrevealpopulationuctuationsanddeterminethemostheavilyusedhabitat.Protection,

4 particularlyofthemostimpor-tantpukuhabit
particularlyofthemostimpor-tantpukuhabitat,alongsidecurbingexcessivepukuoff-takesarecrucialelementsinthefutureconservationandsustainabilityofthisspecieswithintheKilomberoValley.ThisstudywasproducedaspartoftheFrontier-TanzaniaSavannaResearchProgramme,whichisacollaborationbetweentheUniversityofDaresSalaam(UDSM)andtheSocietyforEnvironmentalExploration(SEE).WethankthepersonnelatUlangaDistrictDepartmentofNaturalResourcesandWildFootprintsLtdforalltheirhelpandassistance,andforpermittingeldworkonlandnearthehuntinglodge.LouisvanZylfromKilomberoValleyTeakCompany(KVTC)kindlyprovideduswithamapwhichwasusedtoproduceFig.1.ThanksalsotoPeterMsan-gameno,AbdulMwangalileandChristopherNchimbi(gamescouts)andEibleisFanning(SEE).Finally,aspecialthankstoallstaff,gamescoutsandResearchAssistantsofFrontier-Tanzaniawhocarriedouttheeldsurveysbetween1999and2003,andespeciallytoGrahamCortiwhocoordinatedthesurveys.RichardK.B.Jenkinsandthreeanonymousreviewersprovidedvaluablecommentwhichimprovedthemanuscript.,C.,,D.&,E.(2007)LivestockandlargewildmammalsintheKilomberoValley,insouthernTanzania.Afr.J.Ecol.,G.,,E.,,G.&,R.K.B.(2002)Obser-vationsonthepukuantelopeintheKilomberoValley,Tanza-Afr.J.Ecol.,F.P.D.(2005)TheUpembalechwe,Kobusanselli:anantelopenewtoscienceemphasizestheconservationimpor-tanceofKatanga,DemocraticRepublicofCongo.J.Zool.Lond.,F.M.&,C.(2006)Populationstatusanddistri-butionofpukuinachangingriverfronthabitatinnorthernBotswana.SouthAfr.J.Wildl.Res.,R.(1998)AfricanAntelopeDatabase1998.IUCNSSCAnte-lopeSpecialistGroupReport,Gland,Switzerland.,R.D.(1991)TheBehaviorGuidetoAfricanmammals:IncludingHoofedMammals,Carnivores,Primates.TheUniversityofCali-forniaPress,Berkeley,U.S.A..,C.R.,,R.K.,,G.&Stjernstedt,R.(1998)AnoteonthedistributionandabundanceofpukuKobusvardoniLivingstone,inKasankaNationalPark,Zambia.Afr.J.Ecol.,K.S.(1997)WildlifeProspectsintheKilomberoValleyGameControlledArea,Tanzania.TraditionalvsStatemanagement.CentreforInternationalEnvironmentalandDevelopmentStudiesNoragric,AgriculturalUniversityofNorway,Norway.,K.S.,,F.H.&,S.L.S.(2002)Strivingforsustainablewildlifemanagement:thecaseofKilomberoGameControlledArea,Tanzania.J.Environ.Manage.nvironmentalandTheEnvironmentalImpactoftheProposedKilomberoValleyHardwoodProject.Tanzania.AnAssessmentofaProjectProposedbytheCommonwealthDevelopmentCorporation.TheInternationalInstituteforEnvironmentandDevelopment(IIED).London,andtheInstituteofResourceAssessment(IRA).UniversityofDaresSalaam,Tanzania.IucnsscKobusvardoniiIn:IUCN2009.IUCNRedListofThreatenedSpecies.Version2009.2http://www.iucnredlist.org.Downloadedon22September,R.K.B.,,H.T.&,G.R.(2003)ConservationofthepukuantelopeintheKilomberoValleyTanzania.,R.K.B.,,G.,,E.&Roettcher,K.(2002)ManagementimplicationsofantelopehabitatuseintheKi-lomberoValley,Tanzania.,M.V.(2001)ResourceAssessmentandVillageLandUsePlan–Itete–Minazini,2002–2012.UniversityCollegeofLandsandArchitecturalStudies,UniversityofDaresSalaam,Tanza-,W.A.(1982)ThedeclineoflargemammalpopulationsontheLakeRukwagrassland,Tanzania.Afr.J.Ecol.,W.A.(1984)Statusofthepuku(Kobusvardonistone)inTanzania.Afr.J.Ecol.117–125.,M.,,N.,,A.,,R.A.,,L.,,L.,,L.,,J.,,E.,,J.,H.&,W.(2002)TheKilomberoValleyWildlifeProject:AnEcologicalandSocialSurveyintheKilomberoValley,TanzaniaKilomberoValleyWildlifeProject,Edinburgh,UK.onservationonitoringAerialcensus:SelousGameReserve,MikumiNationalParkandSurroundingArea.Dryseason,October1998.TWCMFZSWildlifeSocietySurveyReport,FrankfurtZoologicalSociety,Arusha,Tanzania.,M.,,M.&,C.(2009)Anassessmentofthepuku(KobusvardoniiLivingstone1857)populationatLakeRukwa,Tanzania.Afr.J.Ecol.Manuscriptaccepted7May2010doi:10.1111/j.1365-2028.2010.01224.xColinBonningtonetal.2010SocietyforEnvironmentalExplorationFrontier.Journalcompilation2010BlackwellPublishingLtd,Afr.J.Ec