3/10/2015
BIO150Y:CensusingPopulations
CensusingPopulationsinAlgonquinPark
AppendixA.CountingPopulations
Thecensusingofplantandanimalpopulationsiscentraltopopulationecology,whichis
basicallythestudyofthepatternsoforganismabundanceanddistribution.Ecologistsseek
explanationsforwhyspeciessuchasBlackRat(Rattusrattus)canlivealmostanywhereand
areabundant,whentheKarnerBluebutterfly(Lycaeidesmelissasamuelis)israreandonly
foundinoaksavannahswithLupinbushes(Lupinusperennis).
Censusingapopulationisusuallyjustthestartofbroaderstudies,butitisnevertheless
importanttodoitright.Sincetotalanddirectcountsofpopulationsarerarelypossible(except
fortherarestofspecies),acensusentailsusingsamplingmethods.Thismeansthatweare
estimatingpopulationsizefromwhatwehopearerepresentativesamples.Byusingamethod
toenumerateapopulationtheobserverisinfluencingtheveryindividualstheyaretryingto
count.Howthemethodusedinfluencestheestimatesobtainedshouldbeunderstoodbythe
biologist.
A.1.SamplingMethods
Biologistsuseavarietyofsamplingmethodsdependingonthetypeoforganismanditsnatural
abundanceanddistribution.Inthisexerciseweuseavarietyofmethodswhichfallintotwo
broadcategories:
Plotbased(quadrat)methods,whereindividualsarecountedinaknownarea.Thistype
ofsamplingiswidelyused,especiallyforplants.
Capturebasedmethods,whichincludesmarkrecapture,andremoval.Thistypeof
samplingisusedforverymobileorelusivespecies.
A.2.QuadratSamplingMethod
Thequadratmethodhasbeenwidelyusedinplantstudies.Aquadratisafoursidedfigure
whichdelimitstheboundariesofasampleplot.Thetermquadratisusedmorewidelyto
includecircularplotsandothershapes.
Acommonshapeusedinstudiesofbirdsandlargemammalsisalong,skinnyquadratin
http://bio150.chass.utoronto.ca/sampling/book/sampling.html
1/5
3/10/2015
BIO150Y:CensusingPopulations
whichtheobservertravelsatransectandcountsindividuals
withinaknownareaeithersideoftheline.Thesetypesof
quadratscanbecalledtransectstrips.(Therearevarious
distancebasedmethodswhichusetransectsandthedistanceoff
thelinethatindividualsareseen,butthesemethodsarenotdealt
withhere.)
Quadratsamplinginvolvescountingallindividualswithina
knownarea(orvolume).Sincedensity(D)andpopulationsize
(N)arerelated,asN=Dxarea,wecanestimatethedensityfor
thesampleandfromthiscomputethetotalpopulation.Thisassumesthattheareathe
populationoccupiesisfiniteandknown.InthecaseoftheAlgonquinParkspecies,wetakethe
Parkboundaries(inthecaseofMoose),theLakeOpeongoarea(LakeTrout),orthestudyarea
boundaries(WhitethroatedSparrow,SugarMaple)asthetotalarea.
Normally,aseriesofsamples(quadrats)arecounted.Wedonotknowwhich,ifany,ofthe
sampleestimatesgivethetrueestimate.Instead,wecancalculatetheaverageofthesesamples.
Theaverageormeanestimate,whilenotlikelytobethesameasthetruepopulationsize,is
neverthelessunbiased.Associatedwiththemean,wecancomputeameasureofthevariability
ofthesamples(thestandarddeviation),andthisisameasureofthereliabilityofourpopulation
estimate.
A.2.1.AssumptionsofQuadratSampling
Thequadratmethodhasthefollowingassumptions:
1. Thenumberofindividualsineachquadratiscounted.
2. Thesizeofthequadratsisknown.
3. Thequadratsamplesarerepresentativeofthestudyareaasawhole.
A.2.2.WorkedExample:EstimatingPopulationAbundancefromQuadrat
Samples
YouhaveusedMethodAtocountMooseinthreequadrats(each25squarekilometres)in
AlgonquinPark(totalarea=7,500squarekilometres).>Fromtheseresults,youwantto
determinebothameanestimateofthetotalMoosepopulationinthePark,andameasureofthe
reliabilityofyourestimate.Thefollowingcalculationsarerequired.
Note:Intheonlinesimulations,themeanandstandarddeviationarecomputedautomatically
foryouforeachmethod.
Step1:CalculateMeanPopulationAbundance
http://bio150.chass.utoronto.ca/sampling/book/sampling.html
2/5
3/10/2015
BIO150Y:CensusingPopulations
Eachofthethreecountsisanobservation(n)andanestimateofthepopulationdensity(D.
WhenmultipliedbytotalParkarea,weobtainestimatesoftotalpopulationsize(N)inthe
Park.Sincewedonotknowwhichisclosesttothetruepopulationsize,wecancalculate
themeanofthethreeobservationswhichgivesusanunbiasedestimateofthetrue
populationabundance.
Table1.Samplecalculationusingquadratdata
Density
(Moose/25km2)
Count1
2100
Count2
11
3300
Count3
12
3600
TOTAL
9,000
Mean=(2100+3300+3600)3=3,000
Step2:ComputingtheStandardDeviation
Wenowwanttocalculateastatisticofthereliabilityofourmeanestimate.Wedothisby
measuringtheamountofscatterordeviationoftheobservations(counts)aroundthemean
estimate.Thisstatisticisthestandarddeviation.
Table2.Samplecalculationofstandarddeviation
Deviations
y=Nmean
y^2
sample1
2,100
2,1003,000=900
810,000
sample2
3,300
3,3003,000=300
90,000
sample3
3,600
3,6003,000=600
360,000
TOTAL
9,000
1,260,000
Thestatisticsthatarederivedfromthiscanbecalculatedasfollows:
Variance=1,260,0002=630,000
Standarddeviation=squarerootofthevariance=794
MoreinformationaboutthisandothermeasuresofdispersioncanbefoundinyourLab
Manual(Chapter3:IntroductiontoStatistics).
http://bio150.chass.utoronto.ca/sampling/book/sampling.html
3/5
3/10/2015
BIO150Y:CensusingPopulations
A.3.MarkRecapture
Themarkrecapturemethodwasfirstusedinthe1890sbyC.G.Petersontoestimatefish
abundance.Todayitiswidelyusedinstudiesofmobileorganisms,suchasmammalsand
insects,andithasbeenappliedtohumanpopulationsaswell.
Themethodinvolvestakingasamplefromthepopulation,
markingthoseindividualsandreleasingthembackintothe
population.Aftertheindividualshavemixedfreelywith
unmarkedindividuals,newsamplesaretaken,andforeach
sampletheratioofmarkedtounmarkedindividualsisrecorded.
Thekeyprincipleofsamplingappliestothismethod:theratioof
unmarkedtomarkedindividualsinasamplewillbethesameas
theratiointhepopulationasawhole.
AsampleofManimalsiscollectedfromthepopulation,marked,thenreturnedtothe
population.Aftermarkedandunmarkedanimalshavemixed,asecondsampleofnanimalsis
capturedandthenumberofmarkedrecaptures,r,istallied.Weassumethattheratioofmarked
tounmarkedintherecapturesample,isrepresentativeoftheratioofmarkedtounmarkedin
thewholepopulation.HencewecanobtainanestimateofN.
A.3.1.Assumptions
Themarkrecapturemethodhasthefollowingassumptions:
1. Thepopulationisclosed(seeexplanationinboxedtextbelow).
2. Markedindividualshavethesameprobabilityofcaptureasunmarkedindividualsinthe
resamplingphase.
3. Markedandunmarkedindividualsmixrandomlybetweenthetimeofmarkingandthe
timeofresampling.
4. Marksarenotlostandarealwaysrecognizable.
ClosureofaPopulation
Closureofapopulationmeansthatitssizedoesnotchangeovertime.Apopulationisin
geographicclosurewhenthereareboundarieswhichlimittheindividualsspatially.An
exampleisalakepopulation.Apopulationisindemographicclosurewhenthereisnonet
changeinpopulationsizeduringthesamplingperiod(nobirths,immigration,emigration,
deaths).Inpractise,thismethodcanaccommodatelossesfromthepopulation,providedthe
ratethatmarkedandunmarkedindividualsleaveisthesame.However,inputsfrom
emigrationorbirthswillallbeofunmarkedindividualsandsowillbiastheestimate.
http://bio150.chass.utoronto.ca/sampling/book/sampling.html
4/5
3/10/2015
BIO150Y:CensusingPopulations
A.3.2.WorkedExample:EstimatingPopulationAbundanceFromMark
Recapture
Assumethat50Moosehavebeenmarkedwithradiocollars(m=50)andreleasedtomixinthe
population.Later,yousamplethreeareasoftheParkandcounttheMoose(n)ofwhicha
subsetarerecaptures(r).Fromthesedatayouestimatepopulationsize,N.Herearesome
sampledataandhowpopulationsizeiscalculated.
CalculatingPopulationSizeFromMarkRecapture
Table3.Samplecalculationusingmarkrecapturedata
Count1
180
3,000
Count2
120
6,000
Count3
100
2,500
TOTAL
11,500
Meanpopulationsize=11,5003=3,833
Standarddeviation=1,893
Calculationofthemeanabundanceandstandarddeviationformarkrecapturesamplesuses
thesameprocedureasforquadratsamplingwhichisdescribedindetailabove.
2002UniversityofToronto.Allrightsreserved.
CommentstoBIO150Ystaff
http://bio150.chass.utoronto.ca/sampling/book/sampling.html
5/5
