estimating quail and pigeon populations in nsw for use in ... · aerial picture of forbes survey...
TRANSCRIPT
ESTIMATING QUAIL AND PIGEON POPULATIONS IN NSW FOR USE IN 2017 NATIVE GAME BIRD HARVEST QUOTA DETERMINATION
FINALREPORT
(MAY2017)
Prepared for: The Game Licensing Unit, Department ofPrimaryIndustry
Reportcompiledby:AssociateProfessorPaulMcDonaldandDrGrahamHall,ZoologyDiscipline,UniversityofNewEngland.
2|P a g e
SUMMARY
• ALLREQUIREDPERMITSANDAUTHORITIESTOUNDERTAKE
THERESEARCHWEREOBTAINED.
• SITEIDENTIFICATIONOBJECTIVE:COMPLETEDASTASKED.FOURSURVEYSITES,CORRESPONDINGTOADMINISTRATIVE
BOUNDARYZONESIDENTIFIEDBYTHEGAMELICENSINGUNIT
(GLU),WERESURVEYED.PRECISESAMPLINGAREASWITHINTHESE
WERECHOSENBASEDONSATELLITEDATA,ON-GROUNDASSESSMENTSANDTHEUSEOFLOCALCONTACTS.
• CONDUCTTRADITIONALSAMPLINGOBJECTIVE:PROTOCOLSWEREDEVELOPEDFORALLSITES,WITHDATACOLLECTION
COMPLETEDATEACHBYTHEENDOFMARCH2017.
• NON-TRADITIONALSURVEYTECHNIQUEEFFICACYOBJECTIVE:CAMERATRAPPINGAPPROACHTRIALEDANDSUCCESS
QUANTIFIED,HOWEVERTHISAPPROACHISNOTRECOMMENDED
FORFURTHERDEPLOYMENT.ACOUSTICCENSUSGROUNDTRUTHINGCOMPLETEDANDSHOWSHIGHPROMISE.ASSESSMENT
INOTHERHABITATSNOWAPRIORITY,HOWEVERTECHNIQUEIS
RECOMMENDEDFORFURTHERSURVEYWORKONBOTHQUAIL
SPECIESSURVEYEDHEREIN.
• SUSTAINABLEHARVESTOBJECTIVE:DENSITYESTIMATESIN
EACHOFTHEFOURZONESHAVEBEENCOMPLETEDFOREACHOF
THETARGETSPECIESTOENABLETHEGLUTOMAKEINFORMED
DECISIONSCONCERNINGPOTENTIALHARVESTLIMITS.
3|P a g e
OVERVIEW
STAFFINGANDCAPACITY:
InadditiontoMcDonaldandHall,toensureanappropriateWHSenvironmentinthefield, and tomaximisedata collection andanalysis efficiency,Ms SigridMackenziewasappointedtotheprojectinthecapacityofaResearchAssistant.SigridhasafirstclassHonoursdegree focusingonthenatural sciences, specialising inzoology.Thisstaff person assisted in carrying out fieldwork, collecting survey data, andpreliminaryanalysisofsonograms.
To further augment the project, a current PhD student at UNE, Ms Ionie-LeeSmallwood and a final year Bachelor of Zoology student, Ms Alison Cash, alsoassistedindatacollection.MsCashhasseveralyearsexperienceactinginaresearchassistantcapacitywithavian fieldresearchprojects,andassisted indeploying fieldequipment.MsSmallwoodiscurrentlyinherfirstyearofaPhDthatwillinvestigatethe behaviour,movement andmolecular population substructure of both Stubble(Coturnix pectoralis) andBrownQuail (C. ypsilophora) across several sites inNSW.Given this focus and obvious overlap with the objectives of this project, MsSmallwood'sexperiencewasusedtoassistindeployingfieldequipmentandvisuallysurveyingquailduringtheacousticcensusassessmenttrialsconductedatInverell.
Together,thisstaffingportfolioprovidedexpertiseandsufficientlabourinallareasrequiredtocompletetheproject,whilstalsokeepingtobudgetaryconstraintsandreportingtimelines.Higherthanexpectedrainfallinearly2017ledtoslightdelaysinpreparing this report,as fieldsiteswere inaccessible tovehicles forseveralweeks,and therefore recordingequipmentand theironboarddata couldnotbecollectedforassessment.Bydelayingthereleaseofthisfinalreport,wehaveensuredthatallcollected data has been incorporated and analysed to provide the most robustpopulationsizeestimatespossible.
4|P a g e
OBJECTIVE1:IDENTIFYTHEFOURGLUZONESMOSTLIKELYTOHOLD
SIGNIFICANTPOPULATIONSOFTHEFOURTARGETSPECIES
RESULTS: Weidentifiedthefoursitesthatweremonitoredunderthisprojecttogainanindex ofabundance,andthussustainableharvestlevels,forthefourtargetspecies(Fig.1). These sitesweredistributedso that theycovered fourof theGameLicensingUnit (GLU)administrativezones,namely:
o Site1:Armidale/Inverellregion:NorthernTablelandsGLUZone
o Site2:PlainssouthwestofDubbo:WesternGLUZone
o Site3:JerilderieandHay:SouthWestRiverinaGLUZone
o Site4:CoffsHarbourregion:NorthCoastGLUZone
Figure1:TheGLUadministrativeboundariesandfouridentifiedregionsofresearch asmentionedthroughouttheproposal.
5|P a g e
These regions were identified through a combination of assessment of past siteoccupancydataintheliteratureandrelevantdatabases(e.g.,BirdlifeAtlasandAtlasof LivingAustraliadata), aswell as assessmentof current conditionsusing remotesensingdata.Thelatterwaspossiblethankstoasystempreviouslydevelopedbyourteam(McDonaldetal.2015)thatidentifiedtheamountofstandingwaterpresentataresolutionof250x250mpixelsacrossNSW.Webothvisitedandcarriedouton-groundassessmentsof potential specific survey siteswithin all of these regions inlate2016tovisuallyconfirmthepresenceofmosttargetspecies.Weaugmentedourown personal observationswith information from local GLU staff and landownersaboutbirddistributionandabundanceintheirareatoassistindeterminingthefinalsurveyandmonitoringsiteswithineachofthefourGLUadministrativezones.
Objective1isthereforecomplete.Detailsofthefinalareassurveyedwithineacharepresentedbelow.PigeonsurveystargetingCrestedPigeons(Ocyphapslophotes)andCommon Bronzewings (Phaps chalcoptera) consisted of a minimum of 4 x 50kmsurveyroutes,asoutlinedanddetailedbelowunderObjective2,withineachzone.Surveyworktargetingthetwoquailspeciesutilisedspotlightingandacousticsensordeploymentfocusingonkeypropertiescontaininglikelyhabitatineachofthezones.The locationofeach isoutlinedbelow.BothStubbleandBrownQuailwereeitherobserved in these areas directly by members of the research team or had beenreportedasbeingpresentbylandholderswithinamonthofthesurveybeginning.
SITE1:NORTHERNTABLELANDSGLUZONE
Location1:Armidale,‘PointsView',30.43700S,151.98386E.
Habitat:Irrigatedimprovedpasturewithtree-linedcreeknearby(Fig.2).
Location2:Inverell,‘Cayuga’,29.58377S,150.96936E.
Habitat: Both barley and chickpea crops in paddocks, often interspersed withimprovedpasture.SparseprimarilyEucalypttreeoverstoreyinsomeareas(Fig.3).
SITE2:WESTERNGLUZONE
Location:Forbes,'LawsonGrains',33.49490S,147.96259E.
Habitat:Largewheatstubblepaddockwithdistant,sparseEucalypttreeborders(Fig.4).
6|P a g e
SITE3:SOUTHWESTRIVERINAGLUZONE
Location1:Jerilderie,‘OldCorrie’,35.34127S,145.54840E.
Habitat: Large wheat stubble paddocks, with sparse Eucalypt tree lines aroundcroppedarea(Fig.5).
Location2:Hay,‘Mungadal’,34.54256S,144.81169E.
Habitat:Largenativegrasslandpaddockswithnoemergenttrees(Fig.6).
SITE4:NORTHCOASTGLUZONE
Location1:Bellingen,30.47472S,152.88185E.
Habitat: Dense Eucalypt forest with native grassland in cleared zones throughout(Fig.7).
Location:NanaGlen,30.11796S,152.98561E.
Habitat: Dense Eucalypt forest with native grassland in cleared zones throughout(Fig.8).
7|P a g e
Figure 2. Aerial picture of Armidale survey site 'Points View', with location ofbioacoustic recorders deployed during surveys as part of the Northern TablelandsGLUZoneQuailSurveyindicated.
Figure3.AerialpictureofInverellsurveysite'Cayuga',withlocationofdifferentcroptypessurveyedaspartoftheNorthernTablelandsGLUZoneQuailSurveyindicated.
8|P a g e
Figure 4. Aerial picture of Forbes survey site 'Lawson Grains', with location ofbioacoustic recorders used during surveys as part of theWesternGLU ZoneQuailSurveyalsoindicated.
Figure 5. Aerial picture of Jerilderie survey site 'Old Corrie', with location ofbioacoustic recordersusedduring surveys aspart of the SouthWestRiverinaGLUZoneQuailSurveyalsoindicated.
9|P a g e
Figure6.AerialpictureofHay survey site ‘Mungadal’,with locationofbioacousticrecorders used during surveys as part of the SouthWest RiverinaGLU ZoneQuailSurveyalsoindicated.
Figure 7. Aerial picture of Bellingen survey site, with location of bioacousticrecordersusedduringsurveysaspartoftheNorthCoastGLUZoneQuailSurveyalsoindicated.
10|P a g e
Figure 8. Aerial picture of Nana Glen survey site, with location of bioacousticrecordersusedduringsurveysaspartoftheNorthCoastGLUZoneQuailSurveyalsoindicated.
11|P a g e
OBJECTIVE2:CONDUCTTRADITIONALLINETRANSECTANDPOINTCOUNTSURVEYSINREPRESENTATIVEAREASOFTHEIDENTIFIEDGLUREGIONSTO
BUILDROBUSTESTIMATESOFFOCALSPECIESPOPULATIONSIZES.
RESULTS:
Initial survey activity focused on the first site in the New England region. Thisapproach enabled a direct comparison between established traditional surveysourceddataandinformationgatheredinthenon-traditionalapproachesoutlinedinObjective 3. Our aim was to establish an appropriate protocol in this region thatcould then be utilisedwith confidence across the other three regions to enable arobustestimationofpopulationsizesthroughoutrepresentativeregionsofNSW.
PIGEONSURVEYS:
Thedifferent survey targets requireadifferentapproach toeffectively census.Forthe two pigeon species of interest, namely Crested Pigeons and CommonBronzewings,avehicular-basedtransectsurveywasdevelopedtoidentifyindividualspresent in thestudyarea.Thesespeciesboth regularlycongregatenear roadsandarerelativelyeasilyidentifiedbyanobservertravellinginaslowmovingvehicle.Anadvantage of these vehicular based surveys is that a larger area is able to besurveyed than otherwise possible on foot, an important consideration given thewidelydispersed,clumpednatureofthepopulationsofbothspeciesinthesurveyedareas.
During each survey, a transect of at least 50kmwas coveredwhilst driving at lessthan 40km/hr. The driver counted each species to the right of the road out to adistance of 50 meters from the roadside edge, while the passenger countedindividualspresentfromtheroadto50meterstotheirleft.Thiseffectivelyenableda 100 meter wide transect to be continuously surveyed, therefore for everykilometretravelled,surveyareawas0.10km2.Binocularswereusedforclarificationofspeciesidentitywhereneededduetovegetativeobstruction,andthevehiclewasstoppedtoensurelargerflocks(>5individuals)wereaccuratelyassayed.
Transect routeswere repeated 4 times per site,with survey effort spread equallyover the morning and evening periods to account any diurnal variance in pigeonnumbers. Morning surveys were completed within two hours of sunrise, whileevening surveyswere initiatednomore than twohours prior to sunset. Transectswerealsoconducted inanalternatingdirection (e.g., ifagivensurvey initiallywasdriven in a clockwise loop, the second survey would be driven in the oppositedirection,i.e.,ananti-clockwiseloop).Inthisway,eachtransectwasdrivenatleastonceineachdirectionforeachtemporalperiod.
Before commencing each transect, potential confounds such as cloud cover andclimaticvariablessuchaswindspeed,precipitationandambienttemperaturewererecorded.Thesedidnotmeaningfullyimpactresults,soarenotpresentedhereinforbrevity.
12|P a g e
Theroutetakenforeachtransectwasmappedbeforehandusingsatellite imagery.An odometer reading was taken at each encounter of the focal species, enablinglocations of birds to be mapped along transect routes for identifying potentialclustersorhabitatsofgreaterthantypicalabundance.
Atotalof28surveyswerecompleted,coveringatotalareaof155.08km2(Table1).Inall,1128CrestedPigeonsand67CommonBronzewingsweredetectedduringthesurvey. It should be noted that the Northern Tablelands GLU Zone and surveyedregions of the SouthWest RiverinaGLU Zone are consideredmarginal habitat forCommon Bronzewings, following perusal of past atlas records of the species.However, Crested Pigeons are considered common to abundant in these regionswithalowerrelianceonwoodedareas,leadingtoussurveyingthesezones.
PerusalofTable1quicklyhighlightsthatbothspecieswereencounteredinrelativelylow numbers throughout many surveys. Further, when the location of eachobservationwasmapped,manyrecordsofbothspeciesweretightlyclusterednearfeedlots or other anthropogenic structures on survey routes. In all, 55%of the 67Bronzewingindividualsand18%ofthe1128CrestedPigeonrecordsweredetectedwithin 2km of a feedlot or town centre on surveys. Further, when flocks wereconsidered(flockheredefinedasgroupsof>1bird),thisincreasedto56%ofthe9Common Bronzewing flocks and 38% of the 69 Crested Pigeon flocks. A clearconclusion from these data is that both species are tightly clustered aroundtownships in the current landscape, presumably to access the anthropogenicresourcesthattheseprovide,suchasfoodandwater.
Tofurtherreinforcethesedifferencesbetweenruralsurveysandtownareasinthesurveyzones,weundertookasinglesurveyof50kmthroughthestreetsofArmidaleon 14 February 2017 where we counted a total of 182 Crested Pigeons. Surveyroutes of the same distance encompassing rural landscapes in the NortherntablelandsGLUAdministrativeZoneaveragedjust20.3±12SDbirdsforcomparison.Together, these data highlight the clustered nature of both species around theseanthropogenicresources.
Overleaf:Table1.LocationofpigeonsurveysandnumbersofCommonBronzewing(BWP)andCrestedPigeon (CP) recordedoneach survey conducted ineachof thefour GLU administrative zones surveyed. Region numbers correspond to SitenumberspresentedunderObjective1.
13|P a g e
GLUZone
Location DateDistance(km)
Area(km²)
Time #BWP #CP#BWP/km²
#CP/km²
1 Inverell 13.09.16 52.4 5.24 am 5 40 0.95 7.63
1 Inverell 13.09.16 52.4 5.24 pm 6 44 1.15 8.40
1 Inverell 14.09.16 71.6 7.16 am 9 50 0.13 7.00
1 Inverell 12.10.16 71.6 7.16 pm 5 53 0.07 7.40
1 Inverell 13.10.16 71.6 7.16 am 3 51 0.04 7.12
1 Inverell 13.10.16 71.6 7.16 pm 5 78 0.07 10.89
1 Inverell 14.10.16 71.6 7.16 am 6 58 0.08 8.10
1 Armidale 25.10.16 45.5 4.55 am 0 9 0 1.98
1 Armidale 25.10.16 45.5 4.55 pm 0 50 0 11.00
1 Armidale 26.10.16 45.5 4.55 am 0 20 0 4.40
1 Armidale 26.10.16 45.5 4.55 pm 0 17 0 3.74
1 Armidale 1.11.16 54.5 5.45 am 0 18 0 3.30
1 Armidale 1.11.16 54.5 5.45 pm 0 15 0 2.75
1 Armidale 2.11.16 54.5 5.45 am 0 25 0 4.59
1 Armidale 2.11.16 54.5 5.45 pm 0 14 0 2.57
1Armidale
City14.02.17 49.6 4.96 am 0 182 0 36.69
2 Forbes 6.03.17 57 5.7 pm 4 9 0.7 1.58
2 Forbes 7.03.17 57 5.7 am 1 15 0.18 2.63
2 Forbes 7.03.17 57 5.7 pm 1 3 0.18 0.53
2 Forbes 22.03.17 57 5.7 am 0 24 0 4.21
3 Jerilderie 22.03.17 46.6 4.66 pm 0 76 0 16.31
3 Jerilderie 23.03.17 46.6 4.66 am 0 64 0 13.73
3 Jerilderie 23.03.17 46.6 4.66 pm 0 62 0 13.30
14|P a g e
Table1.(continued)
GLUZone
Location DateDistance(km)
Area(km²)
Time #BWP #CP#BWP/km²
#CP/km²
3 Jerilderie 24.03.17 46.6 4.66 am 0 111 0 23.82
4Coffs
Harbour12.01.17 56 5.6 pm 1 10 0.18 1.79
4Coffs
Harbour13.01.17 56 5.6 am 0 4 0 0.71
4Coffs
Harbour20.01.17 56 5.56 pm 0 4 0 0.71
4Coffs
Harbour30.01.17 56 5.6 am 21 22 3.75 3.93
15|P a g e
QUAILSURVEYS
Thetwofocalquailspecies,theStubbleandBrownQuailwerealsomonitoredusingmethods established for other Coturnix and similar ground bird speciesinternationally (Hall 2013). Given the cryptic nature of these two target species,smaller, foot-based transects being walked through likely habitat to detect theirpresence was considered, analogous to the techniques used to monitor pigeons.However, time and labour constraintsmeant that this approachwould have beenunviable,coveringaninsufficientareaofinteresttosatisfyourobjectives(Fig.9).
Instead,nocturnalspotlighttransectsofrecentlyharvestedpaddocksweretraversedusingacombinationofvehicularandfoottransportaccordingtoaccessrestrictions.Despitethis,itrapidlybecameapparentthatthelowdensityofbirdswouldpreventanadequatesurveybeingconductedgiventhelogisticalparametersofthisproject.Wethereforechoosetousethemanualsurveytechniqueofspotlightingtoidentifyareaswithknownquailpopulations,andthengroundtruththesuitabilityofamoreefficientacoustic-basedsurveyapproachforwiderdeployment.
Figure9:FemaleStubbleQuaildetectedandphotographedduringatransectwalkedduringdaylighthours.Notethelowtogroundposturethatmakesdetectionofthisspeciesdifficultinthickervegetation.
VISUALSURVEYS:NORTHERNTABLELANDSGLUADMINISTRATIVEZONE
We tookadvantageof thecropharvestactivitiesbeingundertakenon the Inverellproperty 'Cayuga' (Fig. 3) to ground-truth the acoustic-surveydata.At the timeoftheseactivities,OctoberthroughtotheendofDecember2016,twocroptypeswere
16|P a g e
actively being harvested: chickpea and barley. The landholders kindly allowedresearcherstositinthecabinoftheharvesterasittraversedeachpaddock,enablingthenumberofquailofeachspeciesflushedoutofthecroptoberecorded(Fig.10).
TheHarvesterfirstmovedaroundtheperimeterofeachfocalpaddockandtheninincreasinglysmallerconcentric,approximatelycircular,shapessothatthepaddockswereharvested 'from theoutside in'. This approachwasused tohelp ensure thatbirdspresentinagivenpaddockatthestartofharvestwereflushedandrecorded,rather than simply leaving thepaddockundetected through thick cropareasgiventheaversionofbothquailspeciestomoveoveropenground.Observers identifiedthespecies (BrownorStubbleQuail)andage (adultor juvenile) flushedfromeachpaddock.Individualbirdsthatflushedbackintotheunharvestedcropareawerenotrecounted.
Figure 10. The harvester used to identify number of birds present in surveyedpaddocks, here harvesting the outer limits of the crop.Observers sat in the cabinand counted the number of birds flushed from the crop as it transversed thepaddock.
Thisapproachfacilitatedameansbywhichwecouldexhaustivelysampleaknownareaofhabitattogenerateperhectareoccupancyforeachpaddock,withpaddockareacalculatedusingsatelliteimagery.ThisthenprovidedthecomparisondatausedtogeneratepopulationestimatesfromthecallratedatagatheredfrombioacousticrecordersdeployedasoutlinedinObjective3below.
Thedifferentcroptypesyieldeddifferentnumbersofflushedindividuals,withbarleycrops (n=10) and chickpea crops (n=11) differing in their relative attractiveness to
17|P a g e
eachspecies,althoughStubbleQuailweremorecommoninbothcrops(Fig.11).Asbarley crops were naturally harvested earlier in the year (late October-earlyNovember) than chickpea crops (late December) these differencesmay be due tocrop-specificmicrohabitat differences, seasonal shifts in the relative abundanceofthedifferentspecies,and/ordifferences inphysicalhabitatstructure-barleycropsat the time of harvest were 60-70cm above ground and relatively uniform indistribution,whereaschickpeacropsvariedbetween60-100cmandsometimesalsocontainedweedy speciesof varyingdensitywithin thecrop.Furtherelucidationofthe fine scale impact of these microhabitat differences on quail distribution isbeyondthescopeofthisreport,asannualmonitoringacrossanentireyearwouldberequired,howeverthisisanareathatwillbepursuedbyMsSmallwood'sprojectforfutureverification.For thepurposesof this report, crop typeshouldbe takenasaproxyvariablethatispotentiallyindicativeofalloftheabove.
Figure11.Mean (± 1 SE) densities of a) Brownandb) StubbleQuail flushed frombarleyandchickpeacropssurveyedduringharvestnearInverell,NSW.
Interestingly, despite Brown Quail being at higher density in barley crops, andStubbleQuailmoreoftenencountered in chickpeacropsonawithin-species level,therewasastrongpositivecorrelationof0.952betweenthetwospeciespresenceinapaddock.Thiscorrelationdifferedsignificantlyfromzero(t=13.526,df=19,p<0.05),indicatingthatnumbersofthetwospecieswerepositivelycorrelatedinallpaddockssampled,regardlessofcroptype.
18|P a g e
OBJECTIVE3:TESTTHEEFFICACYOFNON-TRADITIONALSURVEYTECHNIQUESTOFACILITATEFASTERANDMORECOMPREHENSIVE
SURVEYSOFTARGETSPECIESINTHEFUTURE
CAMERATRAPS:
Atotalof10cameratrapunits(BushnellTrophyCamHDAggressor)weredeployedat the Inverell site near stations baited with seed likely to attract all four targetspecies.Intotal,thecamerascapturedmorethan10,000images;mainlyoflivestock,rabbits,haresandfoxes.Thecamerasprovedverylimitedintermsoftheirabilitytorecordimagesofbirdsperse,andwereparticularlypoorincapturingimagesofquailandpigeons.
Despite success in capturing a range of non-target species, the method has notprovedtobeasuccessfuloneformonitoringthespeciesofinterest.Thesametwoindividual StubbleQuailwere capturedacross a seriesof 16 images (e.g., Fig. 12),and no pictures were recorded of pigeons, despite them being common at theInverellsite.Giventhis,theuseofcameratrapswasdeemedunsuitableasasurveytechniquethatwasnotpursuedfurther.
Figure12:TwoStubbleQuailidentifiedatasitenearInverellusingaBushnellCameraTrophyHDcameratrap.Thetwoindividualsarestandingontheorangesoiltothecentrerightofframe.
19|P a g e
BIOACOUSTICSURVEYS:
Toeffectivelysurveybothquailspecies inacostandtime-effectivemanner,anewsurvey techniquewasrequired.Ourprevious researchhasestablishedthatpassivebioacousticrecorderscanrecordquailvocalisations,andthatthepeakcallingperiodisduringthetwohoursfollowingcivildawn(McDonaldetal.2015).
Usinga sound levelmeter (DigitechQM-159)wewereable todetermine that theaverageamplitudeofvocalisationsproducedbyBrownQuail,whenstandardisedtolevelsat1musingtheformulaoutlinedinNaguibandWiley(2001),were68.6dB±5SD (n=7 individuals). Stubble Quail were louder, averaging 86.9dB ± 8SD (n=17).Standardised playbacks of recorded vocalisations at these levels indicated thatobservers viewing spectrograms could reliably discern target vocalisations fromtypical background noise at a radius of 100mof the recorder (BAR units, FrontierLabs,Australia).
Next, to generate amodel utilising detected call rates of birds an hour after civildawn, we deployed BAR recorders throughout the centre of paddocks to beharvested(Fig.13).Twodifferentcroptypeswereharvested:barleyandchickpeaasper the results above, with 28 recorders deployed in each. The call rate of birdsvocalisingwithin100mofeachrecorderthemorningofharvestswasthenquantifiedbyvisualisingcallspectrogramsusingRavenPro1.5(Charifetal.2010).Ifpaddockshadcallingbirdsandquailwereflushedfromthepaddockduringharvest,wethencomparedthenumberofadultbirdsflushedperhectarewiththemeannumberofcallsdetectedforthatspeciesperhour.Thus,whilerecordersaresamplinganareaofradius100maroundeachunit,thesedatawereusedtogenerateanestimateofthenumberofquailpresentperhectareateachrecordersite.
Itshouldbenotedthatthesemodelsarepreliminary innature,astherewassomeevidenceofpotentialseasonaldifferences inoccupationand/orhabitatdifferencesasoutlinedabove. Further, as this is a snapshot surveyundertakenat the timeofharvest, call rates of either quail species may differ if surveys are undertaken atdifferent times of the year, and/or in different habitats. Obviously the use of aharvester to flush birds from areas is not an approach that is feasibly applied toevery site, so insteadMs Smallwood's PhD projectwill continue to quantify theseissuesandcomparecallingratestothedensityofbirdsidentifiedthroughspotlight-based nocturnal surveys of key areas. These data will help inform ongoing fine-tuningofthesemodelsformoreaccurateestimatesofquaildensity.
20|P a g e
Figure13:ABioacousticRecorder (BAR; Frontier Labs) setup ina likelyhabitat formonitoringtargetquailspecies.Callswererecordedforseveraldaysduringknowncalling periods for the target species, prior to the area being harvested and thenumber of quail flushed recorded. BARs were placed at least 200m apart in thecentreofpaddockstobeharvested.
21|P a g e
Despitethis,preliminarymodelswerehighlysuccessfulinexplainingahighdegreeofvariance in the number of animals flushed from a given paddock. Brown Quailshowed a positive relationship between the number of calls detected and thenumber of individuals flushed during harvest later that samemorning, accountingfor 32% of variation in this dataset although the model did not reach statisticalsignificance(R2=0.32,F1,9=4.268,p=0.07).However,asapreliminarymodelthenumber of Brown Quail flushed per hectare surveyed (Fig. 14) could beapproximatedbytheformula:
BrownQuailNumberflushed/hectare=0.00314(Meancallrateperhour)+0.677
Figure 14. A model comparing the number of adult Brown Quail flushed frompaddockswhere recorders identified themean hourly calling rate detected in thetwohoursfollowingcivildawnonthemorningofharvest.Bestfitline:y=0.00314x+0.677.
Similarly,althoughtherewasoneoutlierpaddockremovedfromanalysis,indicatingmoreresearchisrequiredtoelucidatefine-scalefactorsinfluencingtherelationship,neverthelesstherewasagainapositiverelationshipbetweencallandflushrateforStubbleQuailinourdata(Fig.15).Again,arelativelyhighproportionofvariationinthedatawasexplainedusingthisapproach,althoughtherelationshipdidnotreachstatistical significance (R2 =0.27, F1,5 =1.892,p=0.23).However,asapreliminarymodel the number of Stubble Quail flushed per hectare surveyed could beapproximatedbytheformula:
22|P a g e
StubbleQuailNumberflushed/hectare=0.0261(Meancallrateperhour)+1.672
Figure 15. A model comparing the number of adult Stubble Quail flushed frompaddockswhere recorders identified themean hourly calling rate detected in thetwohoursfollowingcivildawnonthemorningofharvest.Bestfitline:y=0.0261x+1.672.
Together thesedata demonstrate thepotential of the technique that canbe fine-tuned with additional data quantifying the relative importance and impact ofcovariates such as season, climatic variables, habitat type and area across thesurveyedlandscape.Thatsaid,wehaveusedthesepreliminarymodelsasindicativemeasures of these likely final relationships to generate population estimations forbirdsrecordedinthefourGLUAdministrativeZonesunderObjectiveFourbelow.
23|P a g e
OBJECTIVE4:COMBINELINESOFEVIDENCETOPROVIDESUSTAINABLE
HARVESTESTIMATESFOREACHOFTHEFOURFOCALSPECIES
THROUGHOUTNSW.
RESULTS:
Toachievethegreatestefficacyofmonitoring,amixedapproachofvehicular-basedlinetransectstargetingthepigeonspecieswascombinedwithapassive,bioacousticmonitoring regime for the two focalquail species.Results fromcamera trap-basedsurveysweresufficienttoallowustoruletheseoutaseffectivemeansofmonitoringthe four target species throughout NSW, and as such these are not consideredfurther.
PIGEONPOPULATIONESTIMATES
The pigeon surveys revealed small numbers of Common Bronzewing pigeonsthroughout all four surveyed GLU Administrative Zones. We have not proceededfurtherwiththisspeciesincalculatingsustainableharvestestimates,asthenumbersdetectedaresufficiently small thatanyharvestwouldnotappear tobewarrantedcurrently.
Whilst we observedmore Crested Pigeons throughout the surveys than CommonBronzewings, the clumped nature of the observations around anthropomorphicresources suggests that a harvest of Crested Pigeons is also not currently viable.Consequently we have not provided harvest estimates for either of the targetedpigeonspecies.
QUAILPOPULATIONESTIMATES
AfterestablishingpassiveacousticmonitoringasaviabletechniqueforbothBrownand Stubble Quail, and building preliminarymodels for estimating the number ofindividualspresentperhectare,wedeployedatotalof76additionalrecordingunitsthroughout the four GLU Administrative Zoneswe surveyed for aminimum of 14days (Table 2; barley crops were the exception given the unpredictable timebetweenplacementandharvest initiation).Collectively, theserecorderssampledatotalof4981hoursofsamplingforminimallabourrequirements.Ultimatelytheuseof BARswill bemore labour effective and thus facilitate broader scalemonitoringthanothertechniquescurrentlyused(Hall2013).
We choose 4 days across these 14-day sampled periods and assessed the meannumberofcallspresentforeachrecorderateachsite.Usingthosefigures,wewereable to extract amean call rate per recorder for eachGLU Zone,whichwas thenusedtocalculateanestimateofadultbirddensityinareasofsuitablehabitatusingthepreliminarymodelsdevelopedunderObjective3.ThesearepresentedbelowinTable3.
24|P a g e
Table2.LocationandnumbersofpassivebioacousticrecordersdeployedduringthesurveyineachofthefourGLUAdministrativeZonessurveyed.
Number TotalMeters Number ofhours hours
GLUZone Location Habitat deployed ofdays recorded deployedNorthernTablelands Inverell Barley 28 3 7 161 Chickpea 28 18 4 2016
ArmidaleImprovedpasture 6 14 4 336
NorthCoast CoffsHarbour BellingenForest 4 19 4 304
NanaGlenForest 6 19 4 456Western Forbes Wheatstubble 9 14 4 504SouthWestRiverina Jerilderie Wheatstubble 9 14 4 504 Hay NativeGrassland 11 14 4 616Total 104 4981
25|P a g e
Table3.ThemeannumberofcallspersamplingperioddetectedatrecordersplacedineachofthefourGLUAdministrativezonelocationsduringthestudyfora)BrownQuail and b) Stubble Quail. These figures have been used in conjunction withpreliminarymodelsabovetogenerateestimatesofadultbirddensityperhectareofsuitable habitat in each zone,with a 95% Confidence Interval for eachmean alsopresented.
GLUZone
Meanhourlycall
rate(± 1SE)Number
daysofsamplingassayed
Estimateddensity
perhectareofsuitablehabitat
(Meanand95%CI)
a)BrownQuail
NorthernTablelands 36.9±6.93SE 47 0.793
(0.765–0.821)
Western 0 32 Nonedetected
SouthWestRiverina 0 77 Nonedetected
NorthCoast 0 39 Nonedetected
b)StubbleQuail
NorthernTablelands 5.3±3.31 47 1.810
(1.700–1.921)
Western 10.3±2.99SE 32 1.940
(1.863-2.016)
SouthWestRiverina 44.7±14.75SE 77 2.837
(2.453-3.222)
NorthCoast 0 39 Nonedetected
FromperusalofTable3,itisclearthatthereareinsufficientQuailofeitherspeciesintheNorthCoastGLUAdministrativeZonetosupportaseason.Despitehearingandseeingquailduringpreliminaryvisitsatthesesites,andlandholdersstatingthatquailvisitedtheseareasregularly,ourpassiverecordersdidnotdetectasinglecallacross39samplingdaysfor10differentrecorderlocations.
ForBrownQuail,nobirdsweredetectedintheWesternorSouthWestRiverinaarearecording data, whichwould also indicate that a harvest of BrownQuail in theseareaswasalsonotjustified.TherewereonlymodestnumbersofBrownQuailonthe
26|P a g e
NorthernTablelandsites,withestimateslowerthana1birdperhectareinsuitablehabitat. Given this low density, again it would seem prudent to not harvest thisspeciesinthisGLUZoneatthecurrenttime.
Stubble Quail were detected at three GLU Administrative Zone recorder sites:Western,SouthWestRiverinaandtheNorthernTablelands.Densities rangedfrommeanestimatesof1.8birdsto2.8birdsperhectareofsuitablehabitat.Giventhesehigherdensities, particularly in the SouthWestRiverina, a small, closely regulatedandmonitoredharvestmaybefeasibleforthisspeciesinthosethreeGLUZones.
Given thepreliminarynatureof theacousticmodels,westrongly recommendthatfurther verification of call rate data relative to habitat, time of year andAdministrativeZonebeundertakentoensurerobustpredictivemodelsarefinalised,allowinganyharvestplannedtobesustainable.Keytothisaimwillbedeterminingcritical information on reproductive success, growth rates and basic populationecologydataofbothquailspeciesandenablingthistobefurtherconsideredwhencalculatingsustainableharvestlevels.
27|P a g e
RECOMMENDATIONSFORFUTUREINVESTIGATION
PIGEONSPECIES:
ForCommonBronzewings,other regions shouldbe surveyed thatmayhave largerpopulationsthanthefourGLUAdministrativeZonesmonitoredherein.
CrestedPigeonsurveys inmorewesternGLUAdministrationZonesmaybeabletodetermine regions where their distribution is more differentiated from humanhabitationthanintheareascurrentlysurveyed.Thisshouldbeexamined.
QUAILSPECIES:
Current knowledge of Coturnix movements and population ecology inAustralia is very poorly known, largely due to the difficulties ofmonitoringthesespecies.
Hereinwehavedemonstrated the efficacyof the acoustic survey approach,howeverthecomplexityofhabitatandseasonalvariationshasyettobefullyelucidated.
Ideally,aminimum12monthsurveyingperiodacrossarangeofhabitatsandregions of NSW need to be visually and acoustically surveyed regularly sothat the influence of habitat type on calling behaviour, and any effect ofchanging call rate with season, can be fully built into models. This wouldfacilitate region wide surveys using acoustic recorders alone in subsequentyears, potentially after identifying themost appropriate time of year to doso(e.g.,earlySpring).
ALLSPECIES:
Key to building definitive models for generating sustainable harvest data areinformation on reproductive rates, density dependence and the covarianceintroducedtothesebyfactorssuchasclimaticvariables.Thespeciesvaryintermsofcurrent knowledge on these fronts, but all would benefit from a greaterunderstandingofthesefactors.
Finally, anyharvest programmustbe accompaniedbyongoingmonitoring so thatfutureharvests canbeadaptedaccordingly in termsof raisingor loweringharvestestimatesinfollowingyears.
28|P a g e
ACKNOWLEDGEMENTS Manypeopleassisted inthisproject,andallarethankedfortheirenthusiasticandwilling input. Primarily, the dedication and professionalism of Sigrid Mackenziethroughouttheproject isamajorreasonfor itssuccess. AlisonCashandIonie-LeeSmallwood assisted in quail and pigeon surveys and the installation of BARs withgoodhumourunderhotandtryingconditions.
DPIstaffincludingDrAndrewMoriarty,PatrickO’BrienandMikeWelsharesincerelythanked for their interest, advice and knowledgeof local bird populations. PatrickO’BrienfacilitatedaccesstopropertiesintheForbes,HayandJerilderieareas,whichgreatlybenefittedtheproject.
The many landowners who generously allowed us access to their properties areacknowledgedandthankedwithgratitude.ThesepeopleincludedChristineandJoeHall,‘Cayuga’,Inverell;NarelleJarryandJohnFinlayson,‘PointsView’,Armidale;NilsJacobson, Lawson Grains, Forbes; Russell Ford, ‘Old Corrie’, Jerilderie; and JockCampbell,‘Mungadal’,Hay.ChristineandJoeHallalsoprovidedaccommodationandaccesstotheirharvestedcropsatInverell.
Finally, JoshHall isthankedfortoleratingourmanyquestionsandcompanyduringtheharvestingofthebarleyandchickpeacropsat‘Cayuga’!
REFERENCES Charif RA, Strickman LM & Waack AM(2010).RavenPro 1.4User's Manual. The
CornellLabofOrnithology,Ithaca,NewYork.
HallGP(2013).EcologyandmanagementofquailinAustralia.SSAAInc,Brookvale,NSWAustralia.
McDonald PG, McEvoy J & Hall GP (2015). Monitoring native gamebirds forsustainableagriculturalmanagementpurposes.FinalreportpreparedforGameLicensingUnitofDepartmentofPrimaryIndustry,May2015.
Naguib M & Wiley RH (2001). Estimating the distance to a source of sound:mechanisms and adaptations for long-range communication. AnimalBehaviour62:825-837.