Spatial Distribution of Mule Deer (Odocoileus hemionus): Effects of
Plant Palatability and Habitat Structure
Alessandra Pistoia Mentors: Mary Price, Nick Waser, and Dan Blumstein
Advanced Independent Research Summer 2011
Abstract: ScientistsattheRockyMountainBiologicalLaboratory(RMBL)haverecordedmoremuledeer(Odocoileushemionus)activitydensitywithintheGothictownsiterelativetooutsidethetownsite.OnehypothesistoaccountforincreaseddeeractivityinGothicisthatcoyotes,whopreyondeerfawns,avoidareasofconcentratedhumanactivity,causingdeertoaggregateinthese“safesites.”Toputthishypothesisunderscrutiny,wewantedtotestanalternatehypothesisthatdealswithvegetationasaninfluencefordeerspatialdistribution.IhypothesizedthatifGothicisrichinpreferredplantspeciesandhabitatsthenhighdeeractivitydensitywillcorrelate.TofollowupwithpreviousobservationsondeeractivitydensityinGothic,twopredeterminedroutes,insideandoutsideGothic,werewalkedforafour‐weekperiodtoobservedeerspatialdistribution.Tolookattheavailabilityofpreferredplants,multipleplanttransectswereconductedtoobserveaveragepalatabilityofacertainarea.Tolookattheavailabilityofhabitats,GISArcMapwasusedtodefinedifferenthabitattypesandcalculatearea.WeobservedthatthereweremoredeeroutsideGothicthaninside.Second,wefoundthattheabundanceofpalatableplantsissimilarinsideandoutsideGothic.Last,wefoundthattheproportionofhabitatstypeswassimilarinsideandoutsideGothic,howeverthereweremorewetmeadowhabitatoutsideGothicandmoremeadowhabitatinsideGothic.Fromtheseresults,weconcludethatplantpalatabilitydoesnotinfluencethespatialdistributionofmuledeer;also,sincehabitatstructureissimilarinsideandoutsideofGothic,thishaslittle,ifany,influenceondeerspatialdistribution.Abetterunderstandingofinfluencesofmuledeerspatialdistributionwillleadtoimprovedmanagementofmuledeeroverabundanceinanecosystem. Keywords: muledeer•Odocoileushemionus•spatialdistribution•vegetation•palatability•habitatstructure•trophiccascade
Introduction: Theconceptofspeciesoverabundanceisbestdefinedintermsofmanagementobjectives(UnderwoodandPorter1991).Muledeerareconsideredasoverabundantwhentheylimittheabundanceoroccurrenceofotherresourcesorhinderanecosystemprocessorhumanactivity(Healyetal.1997).Sincemuledeerselectivelybrowseherbaceousplants,theypotentiallyinfluenceecosystemsbyalteringnutrientcycles,reducingplantcover,plantcommunitycomposition,andplantspeciesdiversity(Tilghman1989,Arozqueta2005).Thus,studyingthedistributionofmuledeerpopulationisimportantforecosystemmanagement,restoration,andconservationpolicies(Paceetal.1999).Thenumberofindividualsinapopulationofmuledeerisconstantlychanginginresponsetotheenvironment.Individualsmayrespondtotheenvironmentbychoosingtostayinageographicareaiftherearefavorablelivingconditions,orchoosingtoleavetheareainsearchforamorefavorableenvironment.
Inpreviousyears,researchersfoundmuledeeractivitydensitytobegreaterinsideafencethatsurroundstheRockyMountainBiologicalLaboratory(RMBL)in
Gothic,Coloradothanoutsidethefence(Arozqueta2005,CastroEscobar2010,Pickens2010).ThefencesurroundingRMBLservestoexcludecattle;italsoencompassestheareaofgreatesthumanactivityinthearea.Thehighactivitydensityofdeerwithinthetownsitecouldreflectatrophiccascadeinvolvinghumans,coyotes,muledeer,andplants(CastroEscobar2010,Pickens2010).Accordingtothishypothesis,coyotes,whopreyondeerfawns,avoidtheareaofconcentratedhumanactivity,causingdeertoaggregateinthese“safesites.”Itfollowsthatplantbrowsingisgreaterinthetownsitewherethereisahighdensityofherbivores.Thetermtrophiccascadedescribestherelationshipbetweenprimaryproducersandhigherlevelsofconsumers,andtheeffectthisrelationshiphasontheabundanceandproductivityofpopulationsataspecifictrophiclevel(Paceetal1999).Insuchsystems,predatorscanhaveindirectanddirecteffectsonpreycommunitystructureandpopulationdynamics,preyhabitatselection(Hebblewhiteetal2005),andultimatelyontheplantsconsumedbyherbivores(Schmitzetal2000).Ecosystemswithouttoppredatorsarearguedtohavehighherbivoredensitieswithassociatednegativeimpactsonplantcommunities,eventuallyleadingtoareductioninplantbiodiversity(Terborghetal1999).Indeed,trophiccascadesappeartobefrequentwithinterrestrialecosystems(Paceetal.1999).
AlthoughthetrophiccascadehypothesismayexplainhighdeerdensitiesinsideRMBL,Iexploredtwoalternativepossibilities:palatableplantsandfavorablehabitats.Deerabundancemayreflectthespatialdistributionofpalatableplants,orthedistributionoffavorablehabitats.Ihypothesizethatifpreferredplantsandhabitatsinfluencedeerdensity,thentheareawherethereisahighabundanceofpreferredplantsandhabitatsshouldcorrelatewithahighdeeractivitydensity.Palatableplantsrefertothespeciesofplantsthatmuledeerconsume.Theabundanceofpalatableplantspresentinanareacouldplayacriticalroleinwhetherornotanherbivorewillchoosetoliveintheareabecausetheremustbeenoughpotentialenergyavailableforoptimallivingconditions.Habitatstructurereferstothedominantlifeformofvegetationinanarea(e.g.herbaceousplants,shrubs,ortrees).AtRMBLthereare3predominanthabitats:willow(largeshrubs),meadow(herbaceousplants),andforests(aspenandconifertrees).Thehabitatstructureofanareamayinfluencetheriskofpredationandaffectaprey’sallocationoftimeandenergytobeingvigilantaswellasitsabilitytoescapeattack(HalofskyandRipple2008).Itfollowsthatamuledeerwillchooseahabitatstructurewithhighvisibility,aswellasanareatohidefrompredators(HalofskyandRipple2008).
ThegoalofthisresearchwastocontinueandfurtherpreviousresearchbyaddressingthefollowingquestionsaboutmuledeerinsideandoutsideRMBL:(1)AremuledeermorecommoninsideoroutsidetheRMBLfence?(2)Aremuledeermorecommoninwillow,meadow,orforesthabitat?(3)IstheabundanceofhighlypalatableplantsgreaterinsideoroutsidetheRMBLfence?(4)Istheabundanceofhighlypalatableplantsgreaterinwillow,meadow,orforesthabitat?(5)Whatistherelativeavailabilityofwillow,meadow,andforesthabitatinsideandoutsideoftheRMBLfence?
Methods: StudySite
RMBLisanindependentfieldstationlocatedinGothic,CO,USAatca.2900melevationintheEastRivervalleyoftheWestElkmountains(Inouyeetal.2000).AfenceenclosestheRMBLtownsiteforthepurposeofkeepinggrazingcattleout.Thefencecreatesadistinctionoftwodifferentenvironments:insideRMBLthereishumanactivityandnocattle;outsideRMBLthereisverylittlehumanactivityandnumerousgrazingcattle.Thestudycomparesthesetwoenvironments.DeerObservationAremuledeermorecommoninsideoroutsidetheRMBLfence?
Iestimatedthespatialdistributionofdeeractivitybywalkingstandardroutes,the401(NorthGothic)andDeerCreek(SouthGothic),andnotedwhereIobserveddeer.Thedatawerethenexpressedasdeerobservedperminuteperunitofareaobserved.
IwalkedthetworoutestoestimatetheactivitydensityofdeerinsideandoutsideRMBL.TheseroutesincludedareasoutsideaswellasinsideoftheRMBLfence(Map1).IwalkedtheroutesduringthemonthofJuly,atdawnanddusk.Inordertoavoidconfoundingestimateswithtime‐of‐dayorseasonalactivitypatterns,Iblockedobservationsbyweek,walkingeachrouteequalnumbersoftimesatdawn(05:30‐7:00)anddusk(19:00‐20:30)eachweek.Ialternatedstartingatthebeginningandtheendofarouteonsuccessiveobservations,toavoidconsistentlysamplingparticularpointslaterthanothers.Iwalkedeachroutetwelvetimesduringthesamplingperiod.
Altogether,therewere24observationpointsalongtheroutes(Map1):halfofthesepointswereinsidetheRMBLfenceandhalfwereoutsidetheRMBLfence,alsohalfofthesepointswereinmeadow/willowhabitatandtheotherhalfwereinforesthabitat.Bothhabitatsweresampledinequalnumbersinsideandoutsidethefence.Inotonlyobserveddeerattheseobservationpoints,butalsoatthesegmentsbetweenpoints.IwalkedeachrouteatastandardpaceandrecordedanydeerthatIsaw.Istoppedateachofthe24pointsandscanned(360°scan)fordeerforaminute.
ForeachroutesegmentandeachobservationpointIestimatedtheareaIcouldsee(viewshedarea)bynotinglandmarksthatmarkedtheboundaryofthevisiblearea.IusedanaerialphotographoftheGothicareatomaptheselimitsofvisibility,creatingalayerinGISArcMap(v.9.3.1;ESRIInc.,Redlands,CA,1999‐2009)fromwhichIcouldcalculatetheviewshedarea(Map1).The24observationpointshadagreaterradialviewshedthanthesegmentsduetothefactthatIwasstandingatthepointsforaminute,asopposedtowalking.
InotedthetimeIarrivedateachobservationpointwhilewalkingroutes.FromthisIcouldcalculatetheaveragetimespentwalkingeachroutesegmentandobservingateachpoint.Thesedata,togetherwithviewshedareas,allowedmetoexpressdeeractivitydensityas#deer/(minute)*(hectare).ThedataisexpressedinthisunitinordertoavoidpotentialbiasinthecasethatIsawmoredeerinoneplacethananothersimplybecauseIsampledlargervisibleareathereorspentmoretimelookingthere.To
comparedeeractivitydensityinsidevs.outsidethefenceIusedanANOVAtestonJMP(v.9.02;SASInstitute,Inc.,Cary,NC,2010).
Aremuledeermorecommoninwillow,meadow,orforesthabitat?
WheneverIobservedadeerwhilewalkingtheroutes,Irecordedthehabitatthatitwasin.Itotaledthenumberofdeersightingsineachhabitat,regardlessifthedeerwasfoundinsideoroutsideRMBL.Usingacontingencytable,Iestimatedtheexpectednumberofdeersightingsunderthehypothesisthattheyhavenohabitatpreference.TodothisIestimatedtherelativeareaofthehabitatsalongtheroutesusingArcMap(moreinformationunderGISMappingofHabitatTypes).PlantTransectIstheabundanceofhighlypalatableplantsgreaterinsideoroutsidetheRMBLfence? Inordertomeasuretherelativeavailabilityofhighlypalatableplantsformuledeer,apinwasdroppedevery1meteralongasingle50metertransectestablishedateachofmy24observationpoints.Anyplantsthatthepinhitwererecorded;ifthepinhitthesamespeciesmorethanoncethenthenumberofhitswastotaled.Thismethodestimatestheproportionofthegroundareacoveredbytissueofagivenspecies.Thesetransectswereconductedonceoverthesummer,overasevendayperiod,inlateJuly(beforethecowscameupvalley).Thedirectionoftheplanttransectwaschoseninarandommanneraroundthepoints,howeveritwasguidedbyinterpretationsofmydeerobservations. Wesummedthenumberofhitsforeachspeciesforeachplanttransect.Thenwemultipliedeachspecies’relativeabundance(hitsofeachspeciesdividedbyhitsofallspecies)byitspalatabilityrating(Table1).Thescaleis1‐3(1=frequentlyeaten,2=sometimeseaten,3=rarelyeaten);theratingsofeachplantspeciesarebasedonpreviousdatafromArozquetaandCastroEscobar,aswellasobservationsfromscientistsatRMBL.Afterthespeciesrelativeabundanceweremultipliedbytheirrespectivepalatabilityratingwesummedtheproductoverallthespeciestoderiveanoverallpalatabilityscoreforeachtransect.WethenusedANOVAtoassesswhetherscoresdifferedfortransectslocatedinsidevs.outsidetheRMBLfence.Istheabundanceofhighlypalatableplantsgreaterinwillow,meadow,orforesthabitat? Thedatafromabovewasusedtoaddressthisquestion,howeverthefocusisonthehabitat(willow,meadow,orforest)oftheallobservationpoints,insideandoutsideRMBL.WethenusedANOVAtoassesswhetherpalatabilityscoresdifferedamonghabitats.GISMappingofHabitatTypesWhatistherelativeavailabilityofwillow,meadow,andforesthabitatinsideandoutsideoftheRMBLfence? ArcMapwasusedtoquantifytheareaofhabitatsinthevisibilityviewshedofdeerroutesinordertoestimatemysamplingeffortinthedifferenthabitatsinsideand
outsidethefence.IoverlaidamapoftherouteviewshedontoanaerialphotoofRMBLtakenin2006.ThenItracedtheboundaryofeachhabitattypewithintherouteviewshedandusedArcMapsoftwaretocalculatetheareaofeachhabitattype.
Inordertocategorizethedifferenthabitats(meadow,wetmeadow,willowandforest),numerouspolygonswerecreatedusingArcMaptodefinetheareaofthesehabitatsintherouteviewshed(Map2).Oncethepolygonswerecreated,IusedArcMaptooutputthecumulativeareaofeachhabitattype.FortheanalysisIdecidedtodistinguish“meadow”and“wetmeadow”habitatsbecausethelatteraredominatedbyVeratrumcalifornicum,whicharemuchtallerthanothermeadowplantsandsogivedistincthabitatstructure.TheareaofthehabitatswasfoundusingArcMap,andthesumofeachhabitatinsideandoutsidewasputontoagraphinordertocomparethetwo.Results: AremuledeermorecommoninsideoroutsidetheRMBLfence? Overall,muledeerwereslightlymorecommonoutsidetheRMBLfencethaninside(Figure1).However,itshouldbenotedthatthisoveralldifferencewasnotsignificant(P=0.4682).Therewasamarginallysignificant(P=0.0712)differenceindistributionofdeerinthetworoutes(Figure2):alongtheNorthGothicroutemoredeerwereobservedinsidethefencethanoutsidethefence,whereastheoppositewastruefortheSouthGothicroute(Table2).Wefounddoesandfawnssignificantlymorecommoninsidethefencethanoutsiderelativetobuckscontingencyhabitatinvs.out(x2=4.180493,D.F.=2,P<0.05)(Table3).Aremuledeermorecommoninwillow,meadow,orforesthabitat? Muledeerappearedtopreferforestovermeadoworwillowhabitat(Figure3),astheywereseeninforestmoreoftenthanexpectedbasedontherelativeareaofforestalongroutes,andlessofteninwillowormeadowhabitatrelativetotheareaofthosehabitatsalongtheroutes(Table2).Wefoundbuckstobemorecommoninforestsrelativetodoesandfawns;wealsofounddoesandfawnstobemorecommoninmeadowsrelativetobucks(x2=4.3586,D.F.=1,P<0.05)(Table5).IstheabundanceofhighlypalatableplantsgreaterinsideoroutsidetheRMBLfence? Therewasnosignificantdifferenceinplantpalatabilityinsidevs.outsidethefence(Figure4,Table6).However,thetworoutesdiddifferinpalatabilityinsidevs.outsidethefence(Figure5);intotal,plantsalongtheDeerCreekroutewereonaveragemorepalatablethanalongthe401route.(Figure6,Table6).Istheabundanceofhighlypalatableplantsgreaterinwillow,meadow,orforesthabitat? Habitatsdidnotdifferinpalatabilityofvegetation(Figure7,Table6).
Whatistherelativeavailabilityofwillow,meadow,andforesthabitatinsideandoutsideoftheRMBLfence? Theproportionofallmeadows,forest,andwillowhabitatsareaboutequalinsideandouttheRMBLfence.However,therewasrelativelymoremeadowinsidethefencethanoutsidethefenceandmorewetmeadowwassampledoutsidethefencethaninside.Forestandwillowhabitatsweresimilarlyrepresentedinsideandoutsidethefence(Figure8). Discussion: Thepurposeofthisstudywastoobservemuledeerspatialdistributionandalsotoobservetheeffectsofplantpalatabilityandhabitatstructureondeeractivitydensity.Wefoundthatdeeractivitydensitywasnotobviouslyassociatedwiththedistributionofpreferredplantspeciesorhabitats.Thus,myhypothesiswasrejected.Thisstudystrengthensthetrophiccascadehypothesisbecauseiteliminatesanalternatehypothesis.
First,Theplantpalatabilityfindingsforbothinsidevs.outsidethefenceandthedifferenthabitattypeswasanexcitingfindingbecauseitsuggeststhatpalatableplantavailabilitydoesnotinfluencewheremuledeerprefertolivebecauseitissimilareverywhereinandaroundthetownsite.Palatabilitydifferencesdidvaryamongthetwotransects(Figure6).Thispossiblyexplainsthedeeractivitydensitypatternsobservedontheseparatetransects(Figure2).However,tobetterassessplantpalatability,moretransectsinsideandoutsideofthefence,andalsoindifferenthabitats,shouldsampleabroaderspatialscaleinordertoobtainabetterideaofwhatspeciescoverthosedifferentareas.
Second,therelativeavailabilityofwillow,meadow,andforestwereproportionateinsideandoutsidetheRMBLfence.However,whenwetmeadowisdistinguishedfrommeadow,wefindthattherelativeavailabilityofwetmeadowwasgreateroutsidetheRMBLfence.Theimplicationsthismayhaveondeerspatialdistributioncanonlybeconcludedfromnaturalhistoryobservations.ResearchersatRMBLhaveobservedthatdoeshidetheirfawnsinwillowhabitats,possiblybecauseitisviewedasavisuallyprotectedareafrompredators.Wetmeadowcouldalsoprovidethesametypeofvisualcoverinmid‐latesummerduetothelargeplantspecie,makingitapreferredhabitatformuledeer.Ifthishypothesisistrue,thenitcouldbeonereasonwhywesawmoredeeractivityoutsidetheRMBLfencethaninside.
Althoughtheresultofmuledeeractivitydensityinsidevs.outsidetheRMBLfencewasnotconsistentwithpreviousresearch(Arozqueta2005,CastroEscobar2010,Pickens2010),thiscouldbeduetotemporalfactors.Thissummerwasuniquetoalllivinginthetownsitebecausetherewasalargeconstructionprojectinthecenteroftown.Thenewlaboratorybeingbuiltbroughtnoise,destructiontosurroundingareas,andadrasticincreaseinhumanactivity.Thisextramovementcouldpossiblybescaringthealreadyskittishdeerfartherfromthetownsite.Anotherpossibleexplanationforthedifferenceinresultsisthenewsamplingmethodusedforobservingdeer.WhereasIwalkedtransectstogatherobservations,previousresearcherswereisolatedtothree
observationspointswheretheobserversatforaperiodoftimeandnotedanydeer(CastroEscobar2010).WhencomparingsimilarpointsfrommytransectswiththepointsfromCastroEscobar’spoints,IfoundthatwebothhadsimilardeeractivitydensityforNorthGothic(outsidethefence).However,wehadoppositefindingsfordeeractivitydensityforintownandSouthGothic(outsidethefence)(Table7).
MoredoesandfawnswereobservedinsidetheRMBLfencerelativetobucks.Sincefawnsarethemostvulnerabletocoyotes,thissexdifferencecouldbesignificanttotrophiccascadehypothesis. Torecapitulate,weconcludethatthedistributionofpreferredplantspeciesandhabitatstructuredoesnotinfluencethespatialdistributionofmuledeerinGothic,CO.Therejectionofmyhypothesisstrengthensthehuman‐coyote‐deer‐planttrophiccascadehypothesis.EvelynStrombom,astudentatRMBLthissummer,producedpreliminaryresultsoncoyotescatspatialdistributionandfoundthatcoyotesactivitydensityisgreater500+metersawayfromcabins,relativetoactivitydensity0‐500metersfromacabin(Figure9).Thisinformationalsostrengthensthetrophiccascadehypothesis.Moreresearchisneededinordertogainabetterunderstandingofinfluencesofmuledeerspatialdistribution.
Thebasicmeansbywhichdeerinfluenceanecosystemareknown:nutrientcycling,netprimaryproduction,anddisturbanceregime(Hobbs1996).However,knowledgeabouthowdeerandplantsrespondtomanagementplansremainslimited(Healyetal.1997).Interactionsbetweendeerandtheenvironmentarecomplex,involvingdirectandindirecteffects,feedbackmechanismsandtime‐lags(Pastoretal.1988,PastorandNaiman1992,Ostfeldetal1996).Amodelofthesecomplexrelationshipsshouldbeagoalforfutureresearchonmuledeerspatialdistribution.Thisknowledgewillleadtoimprovedmanagementofmuledeeroverabundanceinageographicarea.
Literature Cited: ArozquetaR(2005)ImpactsofMuleDeerHerbivoryonHerbaceousVegetationin
theGothicarea,withfocusonAquilegiacoerulea.RMBLStudentPaper.
Castro‐EscobarB(2010)Theeffectofanintroducedpredatorscentonmuledeer
(Odocoileushemionus)browsingactivitiesinmeadowhabitatsinGothic,
Colorado.REUFinalPaperRMBL.
HebblewhiteM,CAWhite,CGNietvelt,JAMcKenzie,TEHurd,JMFryxell,SEBayley,
PCPauet(2005)Humanactivitymediatesatrophiccascadecausedby
wolves.Ecology86:2135‐2144
HealyWM,DSdeCalesta,SLStout(1997)Aresearchperspectiveonwhite‐taileddeer
overabundanceinthenortheasternUnitedStates.WildlifeSocietyBulletin25(2):
259‐263
HobbsNT(1996)Modificationofecosystemsbyungulates.JournalofWildlife
Management60:695‐713
InouyeDW,BarrB,ArmitageKB,InouyeBD(2000)Climatechangeisaffecting
altitudinalmigrantsandhibernatingspecies.PNAS97:1630‐1633
OstfeldRS,CGJones,JOWolfe(1996)Ofmiceandmast.Ecologicalconnectionsin
easterndeciduousforests.Bioscience46:323‐330
PaceML,JJCole,SRCarpenter,JFKitchell(1999)Trophiccascadesrevealedin
diverseecosystems.TrendsinEcology&Evolution14:483‐488
PastorJ,andRJNaiman(1992)Selectiveforagingandecosystemprocessesinboreal
forests.AmericanNaturalist139:690‐705
PastorJ,RJNaiman,BDewey,PMcInnes(1998)Moosemicrobes,andtheborealforest.
Bioscience38:770‐777
Pickens(2010)Thelandscapeoffearandtrophiccascades:doeshumanpresenceat
RMBLaffectdeerbehavior?RMBLStudentPaper.
SchmitzOJ,PAHamback,APBeckerman(2000)Trophiccascadesinterrestrial
systems:Areviewoftheeffectsofcarnivoreremovalsonplants.American
Naturalist155:141‐153
TilghmanNG(1989)Impactsofwihte‐taileddeeronforestregenerationin
northwesternPennsylvania.JournalofWildlifeManagement53:524‐532
UnderwoodHB,WFPorter(1991)Valuesandscience:white‐taileddeermanagementin
easternnationalparks.TransactionoftheNorthAmericanWildlifeandNatural
ResourcesConference56:67‐73
Map1:AerialMapofGothic,CO.Thebluelinerepresentsrouteswalkedtoestimatedeeractivitydensity.Thelightorangerepresentstheareasthatcouldbeseenfromtheroutes(viewsheds).ThewhitedashedlinerepresentstheRMBLfence.Thelightgreendotsrepresentobservationpointsinmeadoworwillowhabitatinsidethefence(6total).Thedarkgreendotsrepresentobservationpointsinmeadowandwillowhabitatoutsidethefence(6total).Thepinkdotsrepresentobservationpointsinforesthabitatinsidethefence(6total).Thereddotsrepresentobservationpointsinforesthabitatoutsidethefence(6total).
Map2:AerialmapofGothic,COshowingthemosaicofhabitatswithintheareavisiblefromtheroutes.Orangeiswillowhabitat.Pinkisforesthabitat.Blueiswetmeadow.Tealismeadow.
Plant Species %browse Betsy
%browse Renee Palatability score
Achillea lanulosa -0.064 0.000 3 Aconitum columbianum . . 3 Agoseris auranticaca 0.103 0.000 2 Agoseris glauca . 0.000 2 Aquilegia coerulea 0.147 0.500 1 Artemisia dranunculus -0.064 0.000 3 Campanula rotundifolia . . 2 Castilleja sulphurea . 0.333 1 Delphinium barbeyi -0.038 0.000 3 Delphinium nuttallianum -0.064 . 3 Dugaldea hoopesii -0.041 0.000 3 Epilobium angustifolium -0.064 0.125 1 Erigeron elatior -0.064 0.000 3 Erigeron formosissimus . . 3 Erigeron speciosus -0.060 0.075 2 Frasera speciosa -0.064 0.000 2 Fragaria virginiana -0.064 0.000 3 Galium septentrionale -0.064 0.028 2 Geranium richardardsonii . 0.245 1 Geum macrophyllum -0.005 . 2 Heliomeris (Viguiera) multiflora -0.064 0.000 3 Helianthella quinquenervis 0.132 0.121 1 Heracleum lanosum -0.064 0.034 2 Heterotheca villosa . . 3 Heuchera parviflora -0.039 0.625 2 Hydrophyllum capitatum -0.064 0.000 3 Hydrophyllum fendleri -0.064 0.000 3 Ipomosis aggregata -0.064 0.500 1 Lathyrus leucanthus -0.064 0.019 2 Ligusticum porteri -0.016 0.052 1 Linum lewisii -0.064 0.048 2 Lomatium macrocarpum -0.064 . 2 Lupinus sp. -0.064 . 3 Mahonia repens . 0.000 3 Mertensia ciliata . 0.100 2 Oreochrysum (Pyrrochoma) parryi 0.936 . 2 Osmorhiza occidentalis -0.064 0.000 3
Pedicularis bracteosa -0.064 0.500 1 Pneumonanthe parryi . 0.000 2 Polemonium foliosissimum . . 3 Potentilla hippiana . . 2 Potentilla pulcherrima -0.039 0.113 1 Pseudocymopteris montanus -0.064 0.091 2 Senecio bigelovii -0.064 0.000 3 Senecio integerrimus . 0.286 1 Senecio serra . 0.000 3 Senecio triangularis . . 2 Solidago multiradiata . 0.037 2 Taraxacum officionale 0.036 0.111 1 Thalictrum fendleri -0.051 0.018 3 Tragopogon pratensis 0.811 . 1 Valeriana edulis 0.054 0.667 1 Valeriana occidentalis 0.023 0.245 1 Veratrum californicum . . 3 Vicia americana -0.064 0.000 3 Viola adunca . 0.000 3 Viola nutalli . 0.093 2 small annuals . . 2 small perennials . . 2 Amelanchier alnifolia . . 3 Artemisia tridentata . . 3 Loniceria involucrata . . 1 Paxistima myrsinites . . 3 Picea engelmannii . . 3 Populus tremuloides . 1.000 1 Potentilla fruticosa . . 2 Ribes sp. . . 2 Rosa woodsi . 0.143 1 Symphoricarpos rotundifolia . . 2 Grasses or sedges . . 3 bare ground . . . unknown herb . . .
Table1:Speciespalatabilityranking.%browseBetsyistheaveragedeviationofaspeciesfromoverallbrowsingrateatalocation(CastroEscobar2010).%browseReneeistheproportionofstalksofaspeciesthatwerebrowsed(Arozqueta2005).Palatabilityscore:1=frequentlyeaten,2=sometimeseaten,3=rarelyeaten.
Figure1:Averagedeeractivitydensity(deer/(minute*hectare))forroutesegmentsandpointsinsideandoutsidetheRMBLfence.Verticalbarsindicate+/1standarderrorofthemean.
DF Sum of Squares F Ratio P value Model 3 5.9176 1.4636 0.2233 Route 1 0.8914 0.6614 0.4164 Segment IN or OUT 1 0.7102 0.5269 0.4682 Route*Segment IN or Out 1 4.4010 3.2655 0.0712 Error 651 877.3595 . . C. Total 654 883.2771 . .
Table2:SummaryoffactorialANOVAanalysisfordeerdensity(deer/(minute)*(hectare))insideandoutsidetheRMBLfence.
0
0.05
0.1
0.15
0.2
0.25
IN OUT
Dee
r/(m
in)(
ha)
Relation to RMBL Fence
Figure2:Deerdensity(deer/(minute*hectare))onthe401route(NorthGothic)andtheDeerCreekroute(SouthGothic),insidevs.outsidetheRMBLfence.
Table3:SummaryofGoodnessofFitanalysisfornumberofobserveddoesandfawnsobservedcomparedwithexpecteddeerinsidevs.outsidetheRMBLfence.Thesameanalysisappliestothenumberofbucksobservedandexpected.
‐0.1
‐0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
IN OUT
Dee
r/(m
in)(
ha)
Relation to RMBL Fence
401
DeerCreek
Does&FawnsObserved
DoesandFawnsExpected
BucksObserved BucksExpected
IN 53 47.25 11 15.75OUT 35 39.75 18 13.25
Figure3:Thetotalnumberofdeerobservedinwillow,forestandmeadowhabitat.Thetotalnumberofdeeriscomparedwiththeexpectednumberofdeer,basedontherelativeareaofhabitat.Therewassignificantdeviationbetweenobservedandexpected(x2=6.034,DF=2,P<0.005).
Habitat
Average Deer of All Deer Observed Total Deer Expected Deer Proportion of Habitat Total Area
Willow 0.08 9 15.39314928 0.130450418 96877.821 Forest 0.26 31 22.38335717 0.189689468 140871.2 Meadow 0.66 78 80.22349355 0.679860115 504891.956
Table4:SummaryofGoodnessofFitanalysisfornumberofdeerobservedcomparedwithexpecteddeerinwillow,forestandmeadowhabitat.
Table5:SummaryofGoodnessofFitanalysisfornumberofdoesandfawnsobservedcomparedwithexpecteddoesandfawnsinforest,meadow,andwillowhabitat.Thesameanalysisappliestothenumberofbucksobservedandexpected.
Does&FawnsObserved Does&FawnsExpected
BucksObserved BucksExpected
Willow 6 6 2 2Forest 19 23.25 12 7.75Meadow 62 57.75 15 19.25
0102030405060708090
Willow Forest Meadow
Num
ber
of D
eer
Obs
erve
d
Habitat Type
TotalDeer
ExpectedDeer
Figure4:Averagepalatabilityscoreofvegetationatthe12observationpointsfoundinsidetheRMBLfencevs.the12observationpointsfoundoutsidethefence.
Figure5:Averagepalatabilityscoreofvegetationinsidevs.outsidethefenceonthe401andDeerCreekroutes.
2.15
2.2
2.25
2.3
2.35
2.4
2.45
IN OUT Ave
rage
Pal
atab
ility
Sc
ore
Relation to RMBL Fence
0
0.5
1
1.5
2
2.5
3
IN OUTAve
rage
Pal
atab
ility
Sc
ore
Relation to RMBL Fence
401
DeerCreek
Figure6:Averagepalatabilityscoreofthe401routevs.theDeerCreekroute.VegetationalongtheDeerCreekroutewassignificantlymorepalatablethanthevegetationalongthe401route(Table4).
DF SumofSquares FRatio Pvalue
Model 7 0.6346 1.4841 0.2418Route 1 0.4477 7.3286 0.0155Habitat 1 0.0435 0.7122 0.7062Route*Habitat 1 0.0090 0.1472 0.7062INorOUT 1 0.0001 0.0012 0.9734Route*INorOUt 1 0.0264 0.4322 0.5203Habitat*INorOUT 1 0.0533 0.8724 0.3642Route*Habitat*INorOUT 1 0.0221 0.3611 0.5563Error 16 0.9773 . .C.Total 23 1.6119 . .Table6:SummaryofANOVAoftheaveragepalatabilityscoreinsideandoutsidethefence.
0
0.5
1
1.5
2
2.5
3
401 DeerCreekAve
rage
Pal
atab
ility
Sc
ore
Transect
Figure7:Averagepalatabilityscoreofthe12observationpointsinforesthabitatvs.the12observationpointsinmeadow/willowhabitat.
Figure8:Theproportionofareaofhabitats(forest,willow,wetmeadow,andmeadow)intheviewshedsoftheroutes.
2.1
2.15
2.2
2.25
2.3
2.35
2.4
2.45
2.5
forest meadowAve
rage
Pal
atab
ility
Sc
ore
Habitat Type
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
IN Out
Pro
port
ion
of V
iew
shed
Relation to RMBL Fence
Forest
Willow
Wet Meadow Meadow
LocationCastroEscobar(deer/(scan)(ha))
Pistoia(deer/(min)(ha))
Gothic 0.095 0.062SouthGothic 0.006 1.105NorthGothic 0.000 0.000Table7:Comparisonof2010(CastroEscobar)and2011(Pistoia)muledeeractivitydensitydataforthreeobservationpointsthatwerecommontothetwostudies.GothicisinsidetheRMBLfence.SouthandNorthGothicareoutsidetheRMBLfence.
Figure9:Thetotalnumberofcoyotescatobserved0500metersand500+metersfromacabin.Thetotalnumberofcoyotescatiscomparedwiththeexpectednumberofcoyotescat,basedontherelativeareafromthenearestcabin.
0
5
10
15
20
25
30
0‐500 500+Num
ber
of C
oyot
e Sc
at
Distance from Nearest Cabin (meters)
Observed
Expected