Fine-scale movement of the European hedgehog: an ......SHANAHAN ET AL.: SPOOL-AND-THREAD TRACKING OF...
Transcript of Fine-scale movement of the European hedgehog: an ......SHANAHAN ET AL.: SPOOL-AND-THREAD TRACKING OF...
160 NEWZEALANDJOURNALOFECOLOGY,VOL.31,NO.2,2007
New Zealand Journal of Ecology(2007)31(2): 160-168©NewZealandEcologicalSociety
Availableon-lineat:http://www.newzealandecology.org/nzje/
Fine-scalemovementoftheEuropeanhedgehog:anapplicationofspool-and-threadtracking
DanielleF.Shanahan1*,RenaudMathieu3,4andPhilipJ.Seddon21SchoolofIntegrativeBiology,UniversityofQueensland,StLucia4072,Australia2DepartmentofZoology,UniversityofOtago,POBox56,Dunedin,NewZealand3SpatialEcologyResearchFacility,SchoolofSurveying,UniversityofOtago,POBox56,Dunedin,NewZealand4CSIR–NRE,EarthObservationResearchGroup,Building33,POBox395,Pretoria0001,SouthAfrica*Authorforcorrespondence(Email:[email protected])
Publishedon-line:3December2007
___________________________________________________________________________________________________________________________________
Abstract: The European hedgehog is a significant predator species of rare and endangered ground-nesting birds in theriverbedsoftheWaitakiBasin,SouthIsland,NewZealand.Studiesfocusingonthemovementsandhabitsofthisspecieshavegenerallybeenlimitedtobroad-scaleradio-trackingstudiesorincidentaltrap-catchdata.Withinourstudy, we aimed to investigate the finer scale movement patterns of the hedgehog in relation to vegetation structure byusingspool-and-threadtracking.Wecaptured30hedgehogs(15female,15male)withinthestudyarea,andspool-and-thread-trackedthemovementsofeachoverasinglenight.Onlytwoofthe30animalsmovedontothegravelareasoftheriverbedswherebirdsnest–hedgehogsmaythereforenottargetbirds’nestsasaprimarypreysource,butratherasasecondarypreyitem.Themovementpathswereallnon-random,andmalesdemonstratedparticularlinearityintheirtracks.Thisstraighterandmoredirectedmovementmaybeduetomoreintensivematesearchatthistimeoftheyear.Wealsoassessedhabitatuseusingaveryhighresolutionhabitatmap(derivedfromIkonos4-m-resolutionsatelliteimage).Densegrasslandwasthemostselectedhabitattype,perhapsbecauseinsectpreyareatahigherdensityinthisvegetationtype.Hedgehogs(particularlymales)alsousedboundariesofallhabitattypes significantly more than the centre of habitat patches. We found the spool-and-thread tracking technique does have limitations: (1) it could be inappropriate for animals exhibiting a significant escape response; (2) the data do notincludeatemporaldimension.However,theseproblemswerenotconsideredrelevantforthisstudy.Fine-scalestudiessuchasthiscanprovideincreasedpowerwheninvestigatingtheecologyofspeciesatascalerelevanttotrapplacement.___________________________________________________________________________________________________________________________________
Keywords: animal movement; Erinaceus europaeus occidentalis; Ikonos; introduced predators; Waitaki Basin
IntroductionIntroducedmammalianpredatorshavebeenimplicatedinthedeclineorextinctionofmanyofNewZealand’sendemic species (King 1984; McLennan et al. 1996; Murphyetal.2004).TheEuropeanhedgehog(Erinaceus europaeus occidentalis) preysuponmanyground-nestingbirdspecies.Onesuchspeciesisthecriticallyendangeredblackstiltorkaki(Himanotopus novaezelandiae),whichnestsonthebraidedriverbedsoftheWaitakiBasin,SouthIsland, New Zealand (BirdLife International 2003).DespiteextensivecontroloperationstargetingintroducedmammalsintheWaitakiBasin,hedgehogscontinuetohaveanadverseeffectonthenestingsuccessofground-nestingriverbirds(Keedwelletal.2002).Feralcats(Felis catus),hedgehogsandferrets(Mustela furo)havebeenheldresponsiblefor43,20and18%,respectively,ofnestlossesofkakiandotherbirdsintheriverbeds(Sanders
& Maloney 2002). Further information on the fine-scale movementbehaviourofspeciessuchasthehedgehogmayenableustoimproveourunderstandingoftheecologyofthespecieswithreferencetotherisktoground-nestingbirds. This information has been identified as essential for determiningoptimumplacementoftrapsandpoisonbaitstations; but such knowledge may also support alternative managementstrategies(suchashabitatmanipulations)thatmaydiscouragepredatorsfromimportantconservationareas (Norbury et al. 1998; Ragg & Moller 2000).
Fewstudieshaveinvestigatedmovementbehaviourof hedgehogs in relation to landscape and vegetationstructure and composition; those that have, have generally usedbroad-scaletechniquessuchasradiotracking(e.g.Morris 1986; Baker 1989; Moss 1999), or incidental trap-catchdataanalysedpost hoc(Cameronetal.2005).Suchstudieshaveprovidedinformationthatcanaidpredatormanagement at a broad scale (e.g. Berry 1999; Ragg &
161SHANAHANETAL.:SPOOL-AND-THREADTRACKINGOFHEDGEHOGS
Moller2000).However,thereissomeindicationfromthese broad-scale studies that fine-scale habitat structure may play an important role in determining the areasindividualswillmovethrough.Forexample,Doncasteret al. (2001) used radio-telemetry to demonstrate thatdisplaced hedgehogs showed an attraction to habitatedges,whichwerebelievedtoactascorridorstobetterlocations.Also,inanincidental-trap-catchstudyCameronet al. (2005) found that hedgehogs were most oftencaughtintrapswithin2.5moflowvegetativecover,andalongpredatorpathways.Thesestudiesraisequestionsas to whether these structures are used continuouslyduringforagingtrips,orwhetherthereareotherareasorlandscapestructuresthattrappingeffortsshouldfocusonmore to improve trap efficacy. This is the basis for the currentresearch.
Refining the scale at which animal movement studies arecarriedoutpresentstechnicalandlogisticalproblems.Highlyaccurateanimalmovementdataincombinationwithextremelyhigh resolutionhabitat informationarerequired.Options forgathering theappropriateanimallocationdataincludeGlobalPositioningSystem(GPS)collars,asusedformedium-tolarge-sizedmammalssuchas ungulates and wolves (Merrill 2000; Johnson et al. 2002).CommerciallyavailableGPSdevicesare,however,not yet light-weight enough for use on many of NewZealand’sintroducedpredators,andareoftentoocostlyforwildliferesearchprojects.Spool-and-threadtrackinghasbeendevelopedasaninexpensiveyeteffectivemethodoftracking fine-scale movements of small mammals (Miles et al. 1981; Woolley 1989). In an environment such as theWaitakiBasin,thethreadprogressivelyunravelsandsnags on boulders or vegetation as the animal moves,recording every turning point. In combination withGPS technology, spool-and-thread tracking can yieldaccuratemovementdatainaformcompatiblewithGISsoftware.High-resolutionsatelliteimagery(upto1m)isnowavailablecommercially(e.g.Ikonos,Quickbird),andthisprovidesauser-friendlyoptionfordevelopingprecisehabitatmaps.
Thisstudyaimstobegintodevelopandtestpossibleresearch techniques for investigating the fine-scale habitat use of small mammals in New Zealand braided riversystems,whilebuildingonknowledgeofhedgehoghabitatuse.Weemployspool-and-threadtrackingtoobtainhigh-accuracymovementdata,andcombinethesewithhabitatmapsgeneratedfromhigh-resolutionsatelliteimagery.Indoingsowealsoaimtoaddressthreemainquestionsinrelationtohedgehogmovementpatterns:(1) Is a very fine-scale movement study meaningful? i.e. at a very fine scale, do hedgehogs move in a random fashion or are there obvious patterns in this movement?(2)Whichhabitatorvegetationtypesappeartobeimportantfor foraging hedgehogs? How often do hedgehogs frequent the gravel/rocky areas where river birds nest? (3) Do hedgehogs exhibit expected use of landscape
structure,includingvegetationedgesandlinearvegetationstrips?
MethodsStudy areaThestudywascarriedoutintheOhauandTekapobraidedriverbeds, Waitaki Basin, South Island, New Zealand(44°16–20´ S; 170°5–13´ E; Fig. 1). The river areas used inthisstudycompriselongstretchesofrockyandgravelareas interspersed with low-flow waterways. The water flow is heavily controlled by hydro dams located at the head of each river, and hence flooding events are rare. Both rivershaveheavyinvasive-weedinfestations,particularlysweetbriar(Rosa rubiginosa)andwillow(Salixspp.).Ninespeciesofnativebirdusetheareaasbreedinghabitat(Byrom2002), severalofwhich requireopenareasofgravelsorrockstonestsinglyorincolonies.
Habitat mapThehabitatmapusedinthisstudywasderivedfromaclassified Ikonos satellite image, with 4-m resolution. The final habitat map was composed of six vegetation classes, consideredtorepresentthemajorstructuralelementsofthevegetationinthebraidedriverbeds.Thisnumberofclasses also provided a very high level of classification accuracy.TheIkonosimagewascapturedon1December2002. The vegetation structure was unlikely to havechanged significantly in the period between the image capture and the tracking element of this study (springand summer 2003), as no flood events were experienced intheriverbedsintheyearsbetweenand,asmentioned,water flow is heavily controlled. The final classes of the
Figure 1.Locationofthestudysite–WaitakiBasin,SouthCanterbury,SouthIsland,NewZealand.
162 NEWZEALANDJOURNALOFECOLOGY,VOL.31,NO.2,2007
habitat map were: water; rocks and gravels; low-density vegetation on bare soil; dense grassland; medium- to high-density shrubs; high dense willow trees.
Animal trapping and trackingThirty hedgehogs (15 males, 15 females) were live-trappedusing20Holden live-capturebox traps (MikeHolden; www.traps.co.nz) and 50 Grieves cage traps (GrievesWroughtIronandWirework,302TravisRoad,Christchurch, New Zealand). A total of 1500 ha wastrapped between September 2003 and January 2004(spring/summer).Blocksofapproximately300haweretrappedfor10-dayperiods,withtrapsspaced150–250mapart.Trapswerebaitedwithfreshrabbit(Oryctolagus cuniculus),andcheckedwithin3hofsunrise.Trappedhedgehogswereprovidedwithdrygrassandwater,andleftinsituuntildusk.Non-targetspecieswereimmediatelyreleased.
Spool-and-thread tracking devices were attached,andtheanimalswerereleasedatdusk.Aspool-and-threadbackpack (weight 12.4 g, average 1.9% of hedgehogbodyweights)wasgluedtothespinesontherearofeachhedgehog(Fig.2).Backpacksweremadebyplacingtwospoolssidebysideandenclosingthemwithinheat-shrinkplastic. Each spool (manufactured by Danfield (UK), distributedinNewZealandbyTexspecNZ,Auckland)containedapproximately450mofthread.Itisunknownwhatproportionoftheanimal’sentirenightlymovementthis length encompasses, but no hedgehogs were everfound still attached to the string. The total availablethreadlengthwasapproximately900m,whichisclosetoestimatesofhome-rangelengthforhedgehogsintheregion(Moss1999).Theendofthethreadwastiedofftoa fixed object and the animal was released at dusk. The resultingtrailwasfollowedandmappedthenextmorningusing a GeoExplorer 3 GPS unit (Trimble). Locationfixes were recorded every five seconds. Tracks were then downloadedanddifferentiallycorrectedwithinTrimble
Figure 2. European hedgehog with a spool-and-threadbackpackattached.
GPS Pathfinder Office 2.90, using the base station from theSchoolofSurveying,UniversityofOtago,Dunedin,NewZealand(190kmfromthestudysite).Themaximumhorizontalerrorofthetrackswas3m.
Analysis of animal movement pathsThreeaspectsofhedgehogmovementpatternandhabitatselectionwereinvestigatedtoaddressthethreequestionsposed in the introduction: (1) test for random movement; (2) preferred habitat types; and (3) use of habitat edges.
Test for random movementThis analysis was carried out to determine whetherhedgehogsmove through their landscapeat randomatthis fine scale, or whether the animals make directional choices due to other, potentially measurable fine-scale factors(e.g.habitatstructure).Thelinearityofalltrackswas compared with that of random walk tracks. Onethousand random tracks were generated for each realtrackusingAnimalMovementAnalysis1.1(Hooge&Eichenlaub1997),anextensiontoArcView3.2®(ESRI):eachdistancetravelledbetweenturningpointswasequaltothatobservedinthetrueanimaltracks,andrandomturninganglesweresubstitutedfortheactualanglestaken.Thisprocessprovideddatathatenabledustocomparerandommovementpatternsdirectlywithrealpatterns.Thestatisticscomparedbetweenrealandrandomtrackswere(a)R2:theaveragestraight-linedistanceeachconsecutiveturningpointmovedawayfromthegeometriccentre(startpointof the track); and (b) LI: the total linear distance between thestartandendpointsofananimal’spath,dividedbythe total length of the path (linear paths yield LI = 1; and LI<1isnon-linear).
Foreachindividual,observedR2andLIvalueswerecomparedwiththerangeofvaluesgivenbytherandomtracks.Observedvaluesthatfellintherangeofthetop5%ofvaluesfortherandomwalktracksindicatedthemovement path was significantly more linear than would be expected if turning angles were made at random.Alternately,iftheR2andLIofanobservedanimaltrackwasinthelower5%oftherangeofvaluesfortherandomwalk tracks, then the observed track was consideredsignificantly more constrained than would be expected from a random track (i.e. it exhibits higher site fidelity) (Hooge&Eichenlaub1997).
Preferred habitat typesCompositionalanalysis(Aebischeretal.1993)wasappliedtothedatatodetermineifthehedgehogspreferredmovingwithin certain habitat types. This statistical techniquecomparestheproportionofeachhabitatavailabletoananimalwiththeproportionactuallyused.Inthisstudythehabitat‘used’wasconsideredthatwithinabufferof15meithersideofthetrack.AsrecommendedbyEricksonetal.(2001),thisbuffersizeaccountsforalltheeffortassociated
163SHANAHANETAL.:SPOOL-AND-THREADTRACKINGOFHEDGEHOGS
with the habitat map and animal tracks.This includesmaximumerrorinthegeo-referencingoftheIkonosimage(± 6 m); horizontal precision of the animal tracks recorded ontheGPS(± 3 m after differential correction); and a pixelsizeof4m(satelliteimageresolution).Thehabitat‘available’toananimalwasconsideredthecircularareaaroundthestartpoint,witharadiusofthestart-to-endpointlengthoftherealanimaltrack(i.e.thetotalareathatthespool-and-threadtrackcouldhaverecorded).Rawdataonthetotalareaofhabitattypesboth‘used’and‘available’toeachanimalwereconvertedintoapercentageoftheareacovered.Compositionalanalysisrequiresthateveryhabitat isused, sozerovaluesweresubstitutedwithasmallpositivevalue,smallerthanthesmallestproportionof used habitats (Aebischer et al. 1993); this substitute valuewas0.001.
Use of habitat edgesThis section of the analysis aims to provide quantifiable evidence for fine-scale preferential boundary use as described by broad-scale habitat-use studies (Baker 1989; Pascoe 1995; Ragg & Moller 2000; Doncaster et al. 2001). A landscape index (edge-density index; edge length (km) per hectare; Haines-Young & Chopping 1996) was used to quantifytheamountofeachhabitatboundarywithinthe15-mbufferzoneofeachanimaltrack.ThismeasurementwasgeneratedwithinPatchAnalyst1.0,anextensiontotheArcView®GISsystem(Elkieetal.1999).Anedge-densityvaluewasgeneratedforeachhabitattypeinthebufferzoneofeachobservedanimaltrack,andtocreatepaireddata,theindexwasalsogeneratedinthesamewayfor10randomwalktracks(i.e.10foreveryrealtrack).Tentrackswereusedbecausetheasymptoteofthemeanwasreachedbythispointinmultiplesimulationsofedge-densitymeasuresforeachrealtrack,indicatingthatthisnumberprovidedacceptablerepresentationofedgeuseexpectedfromrandommovement.
The mean index values for the real tracks werecomparedwiththemeanvaluesfromtherandomtracks,usingpairedt-tests,onet-testforeachhabitattype.Thoughrepeatedt-testscanincreasetheTypeIerror,otheranalyseswereconsideredinappropriateforthistypeofdata.Asthelengthoftheedgeofonehabitattypeisnotindependentofthelengthofotherhabitattypes(i.e.edgesbetweentwoside-by-sidehabitattypesaremeasuretwice:onceforeachhabitattype),themeasurementscannotbeconsideredindependent. We attempted to carry out a RepeatedMeasures anova to deal with this non-independence; however,asthedegreesoffreedomwithinthistestarereduced, the sample sizes in this studywerenot largeenough.Thet-testswerethusappropriatefordetectingoveralldifferencesbetweentherealandtherandomdataforeachhabitattype,anddifferencesbetweenthehabitattypeswasnotinvestigated.Summarystatisticsdetailingthedistanceofeachturningpointwithinthetracksfromthenearesthabitatedgewerealsogenerated.
ResultsGeneral observationsHedgehogs often did not follow existing animal runs; rathertrackswouldoftenamblethroughdensegrass(Fig.3). Short-term nests (where the animal had obviouslyremainedforaportionofthenight–stringwasballedup and a flat hollow had been made in the vegetation) werecommonlyfoundwithindenselonggrass,andmorepermanentnests(twicefoundwithyoung)weregenerallylocatedinlongergrassrightatthebaseofsweetbriarorothershrubs.
Test for random movementTheR2valuesfor12females(80%offemales)and13males (87.5% males) were significantly higher than for the corresponding random tracks.The LI values for 7females (46.7%females)and14males (93.7%males)were significantly higher than the corresponding random values.Therefore, there is evidence that generally themovementsofbothmaleandfemalehedgehogsaremoredispersed(non-random)thanwouldbeexpectedhadtheybeenmovingrandomly.
Preferred habitat typesOnlytwomalehedgehogswererecordedasmovingoutontotherocky/gravellyareaswhereriverbirdsnest.Thewaterwasneverenteredbyanimals,andisonlyrecordedinthehabitatrankingsbelowasitappearsasahabitattype in the buffer zone when animals walk alongsideriversorlakes.
Male hedgehogs showed non-random habitatselection(WilksΛ =0.2353,χ2
5=21.70,P<0.001).Arankingmatrixorderedthehabitattypesinthesequence:densegrassland>medium-tohigh-densityshrubs>highdensewillowtrees>low-tomedium-densityplantsonbare soil >>> rocks/gravels >>> water (Table 1; >>> denotes a significant difference between two consecutively rankedhabitattypes).Rocks/gravelsandwaterwereusedsignificantly less than all other habitat types.
Forfemalehedgehogs,comparisonofhabitat-usewithhabitatavailabilitygaveWilksΛ =0.5022(χ2
5=10.33,P=0.066).Arankingmatrixorderedthehabitattypesinthesequence:densegrassland>medium-tohigh-densityshrubs>highdensewillowtrees>low-tomedium-densityplantsonbaresoil>water>rocks/gravels(Table1).
Use of habitat edgesThemeanedge-densityvaluesformalehedgehogsweresignificantly higher (t-tests, P < 0.001; Fig. 4b) than the values for the random tracks for five habitat types. All habitat types produced a P-value at this level ofsignificance. The habitat edges used more than randomly were:rocks/gravels,highdensewillowtrees,medium-tohigh-density shrubs, dense grassland, and low-density
164 NEWZEALANDJOURNALOFECOLOGY,VOL.31,NO.2,2007
Figure 3. Examplesof female and malehedgehog spool-and-threadtracksmappedwith the six-classhabitat map derivedf rom an Ikonossatellite image of4-mresolution.
Table 1. Resultsfromthecompositionalanalysisofhedgehogspool-and-threadtracksfromSeptember2003toJanuary2004.Habitat types are ranked in order of relative preference (1, highest; 6, lowest) with significant differences between consecutively rankedhabitattypesmarked*(P<0.05,fromrandomisationtests).___________________________________________________________________________________________________________________________________
Rankingofhabitattypesforeachgroup
Femalehedgehogs Malehedgehogs
Habitattype n=15 n=15___________________________________________________________________________________________________________________________________
Water 5 6*Rocks/gravels 6 5*Low-densityvegetationonbaresoil 4 4Densegrassland 1 1Medium-tohigh-densityshrubs 2 2Highdensewillowtrees 3 3___________________________________________________________________________________________________________________________________
165SHANAHANETAL.:SPOOL-AND-THREADTRACKINGOFHEDGEHOGS
plantsonbaresoil.Thisindicatesthatmalehedgehogsusehabitatedges,orecotonalboundaries,morethanexpectedhadtheirmovementsbeenrandom.
Themeanedge-densityvaluesforfemalehedgehogswere significantly higher (t-tests,P < 0.01; Fig. 4a) than thevaluesfortherandomtracksforthreehabitattypes.Thesehabitatedgeswere:highdensewillowtrees,low-tomedium-densityplantsonbaresoil,androcks/gravels.Thisindicatesthatfemalehedgehogsusetheboundariesofonlysomehabitattypesformovement.
DiscussionSpool-and-thread trackingThis study successfully used fine-scale movement data from spool-and-thread tracking and high-resolutionsatellite imagery toquantifyhabitatand landscapeuseby hedgehogs. This methodology provided quantifiable evidenceforhedgehogmovementpatternsthatpreviouslycould only be inferred from broad-scale movement
Figure 4.Bargraphsofedgeuseofdifferenthabitattypes(W=water,R=rocks/gravels,T=highdensewillowtrees,S=medium-tohigh-densityshrubs,G=densegrassland,B=low-densityvegetationonbaresoil)for(a)femalehedgehog(darkgrey)andcorrespondingrandomwalktracks(white),and(b)malehedgehog(darkgrey)andcorrespondingrandomwalktracks(white).Edgeuseisrepresentedbyedge-densityvaluesofeachhabitattypeina15-mbufferzonearoundeachtrack.n= 15 for all groups. * denotes a significant difference between themeansatP<0.05.Errorbarsarestandarderrors.
studies.Theseresultsmustbeinterpretedalongsidethelimitationsofthesetechniques,andinthissensethisstudyalsoprovidesessentiallessonsontheinterpretationandmethodology of future fine-scale movement studies.
Theinterpretationofthemovementdataispotentiallyconfoundedbyescapebehaviourexhibitedbytheanimal.Afterhandling,thehedgehogsusedinthisstudygenerallyburrowedintonearbygrassyareasandremainedthereatleastuntiltheresearcherleft.Tracksdidnottendtoexhibitlonglinearmovementsaswewouldexpecttoobserveifanimalswereattemptingtoescape,thusweconsideredtheeffectoftheescaperesponsetobenegligible.Thisstudyalsooriginallyattemptedtousethespool-and-threadtechniqueonferrets(Mustela furo),amoremobilespecies.Whenreleasedtheseanimalstendedtorundirectlyintotheundergrowthawayfromtheresearcher.Longstraightmovements were initially observed in the tracks. Asthe length of the escape behaviour was very difficult to quantify,thetrackswereconsiderednon-representativeofnormalferretmovement.Thesewerethereforediscardedfromthestudy.Thisexperiencedoesprovideavaluablelessoninunderstandingthelimitationofthespool-and-threadtechnique.Thesuitabilityofthetechniqueforthespeciesunderstudyshouldbeconsideredcarefully.
The lack of a temporal dimension to the animalmovementpathsinthisstudycouldalsoplacelimitationson the interpretation of this data, though the powerand fine-scale information gained can be considered a beneficial compromise. The spool-and-thread tracks give noindicationastotheamountoftimeananimalspentineachhabitatpatch,sotheimplicitassumptioninthisstudy is thatpresence in ahabitat type is an adequatemeasureofrelativeintensityofuseforactivepurposes(suchas foragingormate search).Recordingonly themovementpathofananimalmayhoweveravoidamajorbiasthatcanexistinconventionalmethodswhererestinghabitattypesmaybecomeover-represented.Thismaybeofparticularproblemwhenusing, for example, radio-trackinginahedgehogstudy.Thecaptureresponseofahedgehogistoremaininonepositionuntilthethreathasbeenremoved,andtheanimalsarenocturnal,commonlyrestinginoneplaceforthelengthofaday.Radio-trackingstudies commonly obtain only a single-position fix over thecourseofa24-hourperiod,potentiallyonlyobtaininglocationdatafornestingorrestinghabitattypes.
Thespooland thread techniquedidprovideveryfine-scale data for this study, and was inexpensive and accurate. There are currently no modern trackingtechniques that can provide the researcher with moreaccurate information. The spool-and-thread techniquealsohasthepotentialtoallowtheresearchertoanswerfurther specific questions on how the animal physically movesthroughthevegetation,andwhattrackingmarks(e.g.characteristicburrowing throughvegetation)mayidentifythepresenceoftheanimalinanarea.
166 NEWZEALANDJOURNALOFECOLOGY,VOL.31,NO.2,2007
Hedgehog movement patternsOurstudyprovidesevidencethatinvestigatingmovementatthescaleofthisstudyisappropriate,aspatternsandnon-randommovementwereobservedacrossthelengthofthetracks.Thisanalysisalsorevealedotherinterestinginformationrelevanttohedgehogecology.Malehedgehogtracks tended to be more linear than female tracks.Previous observational (non-quantitative) studies onhedgehogmovementpatternshavealsoshownadifferencebetweenthesexes.Wroot(1984),foundthatover70%offemalehedgehogmovementswereshort,withfewrapidmovements.Howevermalestendedtomoveinamorerapidandlinearstylethanfemales,allowingthemtocovermoreground.ThehedgehogsinthisstudyweretrackedbetweenSeptemberandJanuary,muchofwhichisthehedgehog-breedingseasoninNewZealand(Novemberto January in the Waitaki Basin; Moss & Sanders 2001). Thismayexplainthedifferencesinmovementpatternsbetweenthesexes.Campbell(1973)suggestedthatduringthebreedingseasonmalehedgehogsarelikelytosearchactivelyformates,andarethereforemorewide-rangingthanfemales.Femalehedgehogsalsooftenhavelittersinanestatthistimeofyear(twofemalesinthisstudyweretrackedbacktonestscontaininglitters),andmaybelesslikelytotravellongdistancesawayfromtheiroffspring.Thiscould reduce themeansquareddisplacementandlinearitymeasurements,resultinginvaluesclosertothoseoftherandomtracks.Thistendencytostaywithinamoreconfined area than males is supported by the smaller home-rangesizeoffemalehedgehogsobservedinmanystudies(Moss&Sanders2001).
The hedgehogs in this study tended to remainpreferentiallyindensegrassland.Cameronetal.(2005)also found that hedgehogs were best trapped in lowvegetation rather than bare substrate, supporting ourfinding. The most obvious potential explanation for the preference for grassland is a possible higher relativeabundance of insect prey species in this habitat type.PreviousstudiesinEuropehavefoundstronglinksbetweenhedgehog distribution and the density of invertebrateprey species (Micol et al. 1994; Cassini & Föger 1995). Further studies on invertebrate distribution within thestudyareawouldberequiredtotestthis.Thepresenceofsuitablenestsiteshasalsobeenlinkedwithhedgehogdistribution(Parkes1975),andMoss(1999)foundthatshrubsandtussockorgrasslandwasthepreferrednestinghabitat.Theselectionofdensegrasslandbyhedgehogsinourstudythereforesupportsthis.
Finally,thisstudyprovidedsupportforthehypothesisthat fine-scale landscape structure is important tohedgehogs. In this case habitat boundaries were usedpreferentiallyovermorecentralareas,thoughthispatternwasmuchstrongerinmalesthaninfemales.Wesuggestthat the use of edge habitat is linked with directionalmovementorsearchbehaviouratahigherlevel(e.g.inthesearchformatesratherthanforinsectprey)inwhichanimals may be required to move greater distances at
greaterspeeds.Theoverallhigherlinearityofthemalehedgehogtrackssupportsthis.Doncasteretal.(2001)andMorris(1986)alsodocumentedhedgehoguseofhabitatboundaries, suggesting that the edgesof densehabitattypesallowlessencumberedmovement.
Interestingly,inthisstudyonlytwomalehedgehogswererecordedtomoveoutintotherockyorgravelareaswhere ground-nesting river birds have been known tonest. As the proportion of nests taken by hedgehogsis high (20%; Sanders & Maloney 2002), our result is unexpected.Thiswouldsuggesthedgehogsrarelyseekbirds’eggsasprey.Astudyofthedietofhedgehogsintheareaalsodemonstratedthatonly4%ofthegutsstudiedcontainedtheremainsofbirds’eggs(Jonesetal.2005).ThisresultandthelowfrequencyofanimalstravellingontothegravelriverbedssupportthehypothesisproposedbyJonesetal.(2005)thatalargehedgehogpopulationcouldbesupportedprimarilybyinvertebratebiomass,andsmallpopulationsofnativebirdsandlizardsmaysufferasasecondarypreysource.
ConclusionThecombinationoftheveryhighresolutionhabitat-mapandtheaccuratespool-and-threadtrackingdataenabledquantification of fine-scale habitat-use by hedgehogs. This research generally supported habitat-use findings fromotherradio-trackingandincidental-trap-catchdata.This,initself,addstothebodyofinformationonhabitatpreferences of the European hedgehog that could beusedtoenhancetrappingprocedures.Wedemonstratedthat it is worth following hedgehogs at this fine scale, andthismayalsobeimportantforotherspecies.Radio-trackingdatacanleavetheresearcherwithanumberof‘unknowns’resultingfrominaccuraciesanderrorswithinthetechnique.Spool-and-threadtrackinghasenabledustofollowmomentbymomenteachdecisiontheanimalhasmadewithregardstoitshabitatandenvironment.Thispoweraloneallowsustosupportobservationsmadeusingothertechniques,butalsoenhancesourunderstandingoftheecologyandbehaviourofthespeciesatamuchmoreintimatescale.
AcknowledgementsWethanktheDepartmentofConservation,Twizel,forlogisticalsupportandLandcareResearch,Alexandra,fortheloanofequipment.ChrisJonesprovidedadviceondrafts and fieldwork, and numerous field assistants made the field season possible. Des Smith provided statistical adviceandreadmanuscriptdrafts,andEsbenKristensonandClairCameronalsoprovidedstatisticaladvice.IanJamiesonandGrantNorburycommentedonearlyversionsofthemanuscript.ThisprojectwasfundedbyanOtagoUniversityResearchGranttoPJS.
167SHANAHANETAL.:SPOOL-AND-THREADTRACKINGOFHEDGEHOGS
ReferencesAebischer NJ, Robertson PA, Kenward RE 1993.
Compositionalanalysisofhabitatusefromanimalradio-trackingdata.Ecology74:1313–1325.
BakerG1989.Aspectsofmammalianpredatorecologyco-inhabitinggiantskinkhabitat.UnpublishedMScthesis,UniversityofOtago,Dunedin,NewZealand.97p.
Berry CJJ 1999. European hedgehogs (Erinaceus europaeus L.) and their significance to the ecological restoration of Boundary Stream Mainland Island,Hawke’s Bay. Unpublished MSc thesis, VictoriaUniversityofWellington,NewZealand.129p.
BirdLifeInternational2003.BirdLife’sonlineworldbirddatabase: thesiteforbirdconservation. Version2.0.Cambridge,UK,BirdLifeInternational.Available:http://www.birdlife.org(accessed11July2004).
ByromAE2002.Dispersalandsurvivalofjuvenileferalferrets Mustela furo in New Zealand. Journal ofAppliedEcology39:67–78.
CameronBG,vanHeezikY,MaloneyRF,SeddonPJ,Harraway JA 2005. Improving predator capturerates:analysisofrivermargintrapsitedataintheWaitakiBasin,NewZealand.NewZealandJournalofEcology29:117–128.
CampbellPA1973.Thefeedingbehaviourofthehedgehog(Erinaceus europaeus,L.) inpasture land inNewZealand.ProceedingsoftheNewZealandEcologicalSociety20:35–40.
Cassini MH, Föger B 1995. The effect of food distribution onhabitatuseofforaginghedgehogsandtheidealnon-territorialdespoticdistribution.ActaOecologica–InternationalJournalofEcology16:657–669.
DoncasterCP,RondininiC, JohnsonPCD2001.Fieldtest for environmental correlates of dispersal inhedgehogsErinaceus europaeus.JournalofAnimalEcology70:33–46.
ElkiePC,RempelRS,CarrAP1999.Patchanalystusersmanual:Atoolforquantifyinglandscapestructure.NWSTTechnicalManualTM-002.Canada,OntarioMinistryofNaturalResources.
EricksonWP,McDonaldTL,GerowKG,HowlinS,KernJW 2001. Statistical issues in resource selectionstudieswithradio-markedanimals.In:MillspaughJJ, Marzluff JM, eds Radio tracking and animalpopulations. San Diego, Academic Press. Pp.209–242.
Haines-Young R, Chopping M 1996. Quantifyinglandscapestructure:areviewoflandscapeindicesandtheirapplicationtoforestedlandscapes.ProgressinPhysicalGeography20:418–445.
Hooge PN, Eichenlaub B 1997. Animal movementextensiontoArcView,version1.1.Anchorage,AK,USA,AlaskaScienceCenter–BiologicalScienceOffice, U.S. Geological Survey.
JohnsonCJ,ParkerKL,HeardDC,GillinghamMP2002.Movement parameters of ungulates and scale-specific responses to the environment. Journal ofAnimalEcology71:225–235.
JonesC,MossK,SandersM2005.Dietofhedgehogs(Erinaceus europaeus)intheupperWaitakiBasin,NewZealand:Implicationsforconservation.NewZealandJournalofEcology29:29–35.
KeedwellRJ,MaloneyRF,MurrayDP2002.Predatorcontrolforprotectingkaki (Himantopus novaezelandiae):lessons from20yearsofmanagement.BiologicalConservation105:369–374.
KingC1984.Immigrantkillers:introducedpredatorsandtheconservationofbirdsinNewZealand.Auckland,OxfordUniversityPress.
McLennan JA, Potter MA, Robertson HA,Wake GC,ColbourneR,DewL,JoyceL,McCannAJ,MilesJ,MillerPJ,ReidJ1996.Roleofpredationinthedeclineofkiwi,Apteryxspp.,inNewZealand.NewZealandJournalofEcology20:27–35.
Merrill SB 2000. Details of extensive movements byMinnesotawolves(Canis lupus).AmericanMidlandNaturalist144:428–433.
MicolT,DoncasterCP,MackinlayLA1994.Correlatesof local variation in the abundance of hedgehogsErinaceus europaeus. Journal ofAnimal Ecology63:851–860.
Miles, M.A., de Souza, A.A. & Povoa, M.M. (1981)MammaltrackingandnestlocationinBrazilianforestwith an improved spool-and-line tracking device.JournalofZoology195:331–347.
Morris PA 1986. Nightly movements of hedgehogs(Erinaceus europaeus) in forest edge habitat.Mammalia50:395–398.
MossK1999.Diet,nestingbehaviour,andhome-rangesizeoftheEuropeanhedgehog(Erinaceus europaeus)inthebraidedriversoftheMackenzieBasin,NewZealand. Unpublished MSc thesis, University ofCanterbury,Christchurch,NewZealand.
Moss K, Sanders M 2001.Advances in New Zealandmammalogy1990–2000:Hedgehog.JournaloftheRoyalSocietyofNewZealand31:31–42.
MurphyEC,KeedwellRJ,BrownKP,WestbrookeI2004.DietofmammalianpredatorsinbraidedriverbedsinthecentralSouthIsland,NewZealand.WildlifeResearch31:631–638.
Norbury GL, Norbury DC, Heyward RP 1998. Spaceuseanddenningbehaviourofwildferrets(Mustela furo)andcats(Felis catus).NewZealandJournalofEcology22:149–159.
ParkesJ1975.Someaspectsofthebiologyofthehedgehog(Erinaceus europaeusL.)intheManawatu,NewZealand.NewZealandJournalofZoology2:463–472.
PascoeA1995.Theeffectsofvegetationremovalonrabbits(Oryctolagus cuniculus) and small mammalianpredators in braided riverbeds of the Mackenzie
168 NEWZEALANDJOURNALOFECOLOGY,VOL.31,NO.2,2007
Basin.UnpublishedMScthesis,UniversityofOtago,Dunedin,NewZealand.118p.
Ragg JR, Moller H 2000. Microhabitat selection byferal ferrets (Mustela furo) in a pastoral habitat,EastOtago,NewZealand.NewZealandJournalofEcology24:39–46.
SandersMD,MaloneyRF2002.Causesofmortalityatnestsofground-nestingbirdsintheUpperWaitakiBasin,NewZealand:a5-yearvideostudy.BiologicalConservation106:225–236.
WoolleyPA,1989.Nestlocationbyspool-and-linetrackingofdiasyuridmarsupialsinNewGuinea.JournalofZoology,London218:689–700.
Wroot AJ 1984. Feeding ecology of the Europeanhedgehog, Erinaceus europaeus. PhD thesis,UniversityofLondon,UK.
EditorialBoardmember:KayClappertonactively