Use of camera traps for wildlife studies. A review

Abstract

As human threats continue to impact natural habitats, there is an increasing need to regularly monitor the trends in large vertebrate populations. Conservation efforts must be directed appropriately, but field work necessary for data collection is often limited by time and availability of people. Camera traps are used as an efficient method to insure continuous sampling and to work in difficult to access areas. In the present study, we illustrate how this instrument is serving a diverse field of studies, such as animal behavior, population monitoring and fauna-flora interaction. By looking at the material and technical aspects of various models of camera trap for implementation in different field studies in animal ecology, we highlight the need to choose appropriate camera trap models for the target species and to set up solid sampling protocols to successfully achieve study objectives.

Full text

BA
SE Biotechnol. Agron. Soc. Environ.201418(3),446-454 Focus on:
Useofcameratrapsforwildlifestudies.Areview
FranckTrolliet(1),Marie-ClaudeHuynen(1),CédricVermeulen(2),AlainHambuckers(1)
(1)UniversitédeLiège.UnitédeBiologieduComportement.22,QuaiVanBeneden.B-4020Liège(Belgique).E-mail:
franck.trolliet@ulg.ac.be
(²)UniversitédeLiège-GemblouxAgro-BioTech.UnitédeGestiondesRessourcesForestièresetdesMilieuxNaturels.
LaboratoiredeForesterietropicaleetsubtropicale.PassagedesDéportés,2.B-5030Gembloux(Belgique).
ReceivedonMarch13,2013;acceptedonFebruary11,2014.
As human threats continue to impact natural habitats, there is an increasing need to regularly monitor the trends in large
vertebratepopulations.Conservationeffortsmustbedirectedappropriately, but eld work necessary for data collection is
oftenlimitedbytimeandavailabilityofpeople.Cameratrapsareusedasanefcientmethodtoinsurecontinuoussampling
andtowork in difcult toaccessareas.In the present study,we illustratehowthisinstrumentis serving a diverseeldof
studies,suchasanimalbehavior,populationmonitoringandfauna-orainteraction.Bylookingatthematerialandtechnical
aspectsofvariousmodelsofcameratrapforimplementationindifferenteldstudiesinanimalecology,wehighlighttheneed
tochooseappropriatecameratrapmodelsforthetargetspeciesandtosetupsolidsamplingprotocolstosuccessfullyachieve
studyobjectives.
Keywords.Wildlifemanagement,populationcensus,animalbehaviour,photography,traps,surveillancesystems.
Utilisation des pièges photographiques pour l’étude de la faune sauvage (synthèsebibliographique). Alors que les
pressions anthropiques continuent de dégrader les habitats naturels, le besoin de suivre régulièrement les tendances des
populationsdegrands vertébrés augmente.Lesefforts de conservationdoiventêtrede plus enplusciblésmais les travaux
deterrainsnécessairesàlarécoltededonnéessontsouventlimitésparletempsetlenombredepersonnesdisponibles.Les
piègesphotographiquesapparaissentainsicommeuneméthodeefcacepourassurerunéchantillonnagecontinuetdansdes
zonesdifcilementaccessibles.Nousillustronsicilamanièredontcetoutilestutilisépourunediversitédethèmesd’études
deterraintels que lecomportementanimal,lesuivi de populations etlesinteractionsfaune-ore. En analysant lesaspects
techniques et matériels permettant d’assurer différents types de travaux d’écologie animale, nous mettons en évidence la
nécessitédesélectionnerdumatérieletdemettreenplaceunprotocoled’échantillonnageadaptéàl’espèce etauxobjectifs
xésdel’étude.
Mots-clés.Gestionde la fauneetdela ore sauvages,recensementdelapopulation, comportement animal,photographie,
piège,systèmedesurveillance.
1. INTRODUCTION
Theobservedrapiddeclineinbiodiversity,particularly
among large vertebrates, throughout the world and
the degradation of natural habitats hosting their
populationsare nowadays widelyaccepted as fact.It
has therefore never been so important to understand
how animal populations respond to modern threats
and to document the functioning of ecosystems and
intra-communityinteractions(Barrowsetal.,2005)as
tobeable to implementappropriatemanagementand
conservation strategies. Regular updating of data on
animalpopulationdensity and on the degreeofinter-
speciesinteractionsisthuscrucialtoassessthespatio-
temporal variations in populations and communities
(Bouché et al., 2012). Camera traps are increasingly
beingused to study wildlife behaviorand to conduct
populationestimations(Cutleretal.,1999;Longetal.,
2008;O’Connell etal.,2011;Roveroetal., 2013).In
thepresentstudy,weundertookaliteraturereviewon
camera trapping studies, to present some technical
aspectsofcommerciallyavailablecameramodelsand
provideanoverviewofsamplingproceduresanduses
ofcameratrappingdata.
2. MATERIALS AND METHODS
We conducted a general literature review on camera
trapping using the SciVerse Scopus® database and
GoogleScholar®.Thelistofscienticpapersconsulted
isnotexhaustiveandwedonotclaimtodocumentall

Useofcameratrapsforwildlifestudies.Areview 447
thestudiesdealingwithcameratrapping.However,the
list of documents consulted has enabled us to gain a
goodoverviewofthediversityofusesofcameratraps
over recent decades and of the main issues regarding
sampling and data analysis. To conduct our study on
thetechnicalaspectsof cameratraps, we searchedfor
cameratrapbrandssoldandadvertisedontheInternet,
as wellas those usedin recent scientic publications.
WenallyconsultedTrailCamPro.com®(TrailCamPro.
com, 2013) and Camera Traps cc®’s (Camera Traps
cc,2013)websitestoretrievetechnicalinformationon
the different models. Those two companies distribute
together 18brands of camera traps, which, to our
knowledge, include the vast majority of camera trap
models on the market. We could get the price for
61differentmodels(15brands).
3. RESULTS AND DISCUSSION
3.1. Diversity of uses of camera traps
Whileremote photographies have been used for more
thanacentury,aspresentedbyO’Connelletal.(2011),
theautomatedcameratrapasitisnowknowncameonto
themarketattheendofthe1980s.Savidgeetal.(1988)
usedalmcameraconnectedtoaninfraredtransmitter,
whichwasabletoshootapictureassoonasthebeamwas
interruptedbyananimal.Thesystemwasautomatic;after
apicturehadbeentaken,thelmwasreloadedandthe
camerawasreadytotakemorepictures.Thistechnique
wasusedtoidentifypredatorsvisitingbirdnests.Some
years later, Carthew et al. (1991) and Kucera et al.
(1993) listed the advantages of the automated camera
trap system for anarray of different eld applications
suchasthestudyofactivitypatterns,intra-community
interactionsandlargecarnivorespopulations.
Therststudiesusingcameratrapsforthepurpose
of large mammal conservation appeared in the 1990s
andfocusedonthetiger,Panthera tigris(e.g.,Grifths,
1993; Karanth, 1995). Following the designation of
P. tigris as endangered (Chundawat et al., 2011), one
of thefew “agship”species listed on the IUCN red-
listasearlyas1986,thesestudiesaimedat estimating
homerangespanand populationsize.Inthis way, the
useofcameratrapstoestimatepopulationsizegreatly
helpedtowardstheconservationstrategyforthespecies,
andmoregenerally, themonitoringofotherthreatened
populationsandcommunities.Thisuseofcameratraps
was highlighted in a study on the activity patterns
of mammal communities in Indonesian rain forests
(van Schaik et al., 1996). The aforementioned early
studies oftheuseofcameratrapsclearly illustrate the
major advantages of using the technique, including
beingableto observecrypticorelusiveanimalsliving
in difcult to access habitats such as dense tropical
forests.Theuseofcameratrapshasbeenrevolutionary
for studying the behavior of carnivores, as they are
difcult toobserveintheirnaturalhabitat duetotheir
solitarynature.Thetechniquehasalsobeenthesubject
ofmanyother scientic paperssincethe beginningof
the 21st century, revealing more about the ecology of
rare,nocturnalanimals,aswellasthosehighlysensitive
tothepresenceofhumansorthoselivinginlargehome
ranges.A goodexample isthestudyofMoruzziet al.
(2002), whichpromotestheuseof thistechnology for
estimating carnivore distribution over large area and
documentingspecies-specichabitatpreferences.
A large proportion of conservation projects aim at
managingthreatenedspecies,whichimpliestomonitor
populationsovertimeandspace.Thus,themajorityof
studiesusingcameratrapsnowadaysappeartodealwith
theestimationofpopulationdensity (e.g.,Kalleet al.,
2011;Garroteetal.,2012;Oliveira-Santosetal.,2012)
or simply with the presence of species in given areas
(e.g., Gil-Sanchez et al., 2011; Gray et al., 2011; Liu
etal.,2012).Populationcharacteristicsare,toagreater
or lesser extent, related to habitat use behaviors and
habitatselection.Cameratrapsareusefulformonitoring
theseaspectsastheyallowtheestimationofhomerange
size(e.g.,Gil-Sanchezetal.,2011).
Some studies also deal with activity budget (e.g.,
vanSchaiketal.,1996;Azlanetal.,2006;Grayetal.,
2011;Oliveira-Santosetal.,2012)andasmallernumber
withmorespecicbehaviors.Forinstance,Soleyetal.
(2011) reportedthe storingbehavior of non-ripefruits
byamustelideae,allowingthefruitstomatureandtobe
consumedonfutureoccasions;thisisaspecicbehavior
thatisveryhardtoreportwithoutcameratraps.Blake
etal.(2010)studiedtheimportanceofsaltlicksforan
animalcommunityinaneotropicalforest.Otherstudies
havedealtwithanimalinfantcare(e.g.,Charruauetal.,
2012) or social interaction (Lopucki, 2007; Srbek-
Araujoetal.,2012).
Camera traps are also increasingly being used to
study plant-animal interactions such as seed dispersal
andpredation(e.g.,Babweteeraetal.,2010;Nyiramana
et al., 2011; Campos et al., 2012; Koike et al., 2012;
Pender et al., 2013). Moreover, focal observations
needtobeconductedinthestudyoftheseeddispersal
capacity of a given plant species, to listthe frugivore
species interacting with the plants and to dene the
quantitativecontributionofeachspeciesintheprocess
of seed dispersal. Camera traps are revolutionary in
thisregard,astheyallowthe identication of diurnal,
nocturnal, and shy species that would not be seen
using other methods such as direct observation. This
isexempliedbythestudyofNyiramanaetal.(2011),
whodiscoveredthataspeciesofrodent,theforestgiant
pouchedratCricetomys emini(Wroughton,1910),was
responsibleforthesecondarydispersaloflargeseedsin
anAfro-tropicalforest.

448 Biotechnol. Agron. Soc. Environ. 201418(3),446-454
3.2. Various technical aspects
Morethanadecadeago,Cutleretal.(1999)reviewed
the advantages and disadvantages of using different
lm camera trapping equipment depending on the
researchobjectives. Giventherapidadvancesinsuch
technology,andthegreatvarietyofcameratrapbrands
anddigital models existing on the market nowadays,
lmcameras arecompeted.Wepresentherethemost
important characteristics to take into account when
choosing digital equipment. Characteristics such as
trigger speed, detection zone, recovery time, night
detection and battery consumption can vary greatly
andhaveasignicantimpactonthetypesofdatatobe
collected,such asthenumberof speciesdetectedand
photographicrates(Hughsonetal.,2010).Therefore,
the choice of the most appropriate equipment is an
importantconsideration.
Trigger speed. Trigger speed is the time delay
necessary for the camera to shoot a picture once an
animal has interrupted the infrared beam within the
camera’s detection zone. This delay can vary from
between 0.197seconds for the Reconyx HC500
model to 4.206seconds for the Stealth Cam Rogue
IR model. Given the relatively narrow eld of view
of most camera trap lenses (42mm), a slow trigger
speeddoesnotallowthephotographingoffastmoving
animals(Scheibeetal.,2008).Thus,dependingonthe
study goals and the target animal species, this time
delaycouldbeacrucialcharacteristictoconsider.For
example,ifacameraissetupatarandomlocationfor
a wildlife survey (Pereira et al., 2012), fast moving
animalsarelikely to pass in frontof the camera trap
withoutstopping. In thiscase,avery reactivecamera
(with a fast trigger speed) would be necessary so it
couldshootpicturesofthedetectedanimalbeforeitleft
thecamera’seldof view.Intheircomparativestudy
ofmotion-activatedcamerasforwidlifeinvestigation,
Hughson et al. (2010) showed that some camera
models (such as the fast Reconyx) can detect up to
86% more animal species. If the trigger speed is too
slow,thecameramayframeonlyapartoftheanimal
ormayeventakeemptypictures(picturesnotshowing
what the beam has detected). Hughson et al. (2010)
observedthat, incomparisonwithother models,Leaf
River cameras took the highest percentage of empty
pictures.In the case of acamera installed in front of
abirdnest,abait,oralure,visitinganimalsaremore
likely to stay longer (to either depredate the nest or
interactwiththebait)andtotriggermorephotographs
(Garrote et al., 2012; Trolle et al., 2003) even if the
camerahasarelativelylongtimedelay(lowreactivity).
Usinglurestoattractlargecarnivorescanalsoallowa
betteridenticationofindividuals(Gil-Sanchezetal.,
2011).Thisriskoftakingemptypicturesdoesnotonly
dependonthespeedofthecameraintakingapicture;
thedetectionzoneaswellastheeldofviewarealso
primarycriteriatoconsider.
The detection zone. The detection zone is the zone
covered by the camera’s infrared beam in which
movementcan be detected. The zone variesin width
and depth, depending on the model (Table 1). This
criterionisprobablythemostimportantindetermining
detectionrate(Rowcliffeetal.,2011)andthereforethe
numberofpicturesthatwillbetakeninagivenevent.
The eld of view.Theeldofviewisthezonecovered
bythecameralens,andwhichappearsonthepictures.
The eld of view is generally 42° but there are rare
exceptionssuchaswiththeLeupoldbrand,whichgoes
upto54°(Table 1)andtheMoultriepanoramicmodel,
whichcoversanangleof150°.Thedetectionzonecan
varygreatlyaccordingtothebrandandthemodel.We
thusndmodelswith adetectionzonewiderthanthe
eldofview(e.g.DLCCovertExtreme)andmodels
withthedetectionzonenarrowerthantheeldofview
(e.g.CuddebackAmbush).Wherethedetectionzoneis
widerthantheeldofview(Figure 1a),theadvantage
liesinbeingbetterabletocapturefastmovinganimals.
The limitation in this case is that the camera is also
likely to take empty pictures when animals enter the
detectionzone(thuspassingthroughtheinfraredbeam
andtriggering the camera)butwithoutmaking itinto
theeldofview.Wherethedetectionzoneisnarrower
thantheeldofview(Figure 1b),thedetectionzoneis
centeredrelativetotheeldofviewofthecamera,and
sotheadvantagecanbeseeningainingwellcentered
pictures.Thiscanbeveryusefulfortheidentication
oflargemammals.However,thelimitationinthiscase
isthatrelativelyfewerpicturespervisitcanbeshot,as
animalsarelikelyto occupytheeldofviewwithout
crossing the detection zone. As presented in table 1,
the detection zone can be described with a given
width(angle)andagivendistancefromthecameraat
whichit willdetectan animal.Thedetectiondistance
of a camera is an important aspect to consider when
focusingonanimalspeciesofeitherlargeorsmallbody
mass.Largeranimals will bemore easily detected at
furtherdistancesthansmalleranimals.However,speed
ofmovementseemstobelesscorrelatedwithdetection
distance(Rowcliffeetal.,2011).
Recovery time.Recovery timeistheamountof time
necessaryforthecamerato prepare to shoot the next
pictureafterthepreviousonehasbeenrecorded.Given
the wide differences in recovery time for different
models,thischaracteristicmustbetakenintoaccount,
as it can be a very important aspect for some study
goals. The fastest camera can take a picture every
0.5second(ReconyxHC500model)whiletheslowest

Useofcameratrapsforwildlifestudies.Areview 449
needsupto60secondsbeforetakinganewpictureof
ananimalstilloccupyingthedetectionzone(Moultrie
I-35smodel).Acameraabletotakenumerouspictures
withina few seconds is very usefulwhen needing to
recordacompletesequenceofafeedingbehaviorand
tonotethenumberoffruitsmanipulated(Seufertetal.,
2010). Also, having different views of a species of
carnivorecangreatlyhelpintheprocessofidentifying
individuals(Trolleetal.,2003).Bycontrast,whenthe
aimisonlytocarryoutadiversitycensus,andonlyone
pictureperspeciesisneeded,aslowrecoverytimewill
belessproblematic(Lantschneretal.,2012).
Nighttime pictures. Nighttime pictures are very
useful, as a wide range of taxa exhibit exclusive
nocturnalactivity.Twomethods existforcameratrap
night photography: incandescent ash and infrared
light. Incandescent ash allows color pictures to be
taken, which are generally of better resolution and
quality.Inthismethod,theamountoflightcapturedis
greaterthanwithinfraredlight,andthiscanbecritical
for individual animal identication with the use of
tagsornaturalmarks.Thelimitationofthismethodis
that the ash has a strong risk of scaring the animal
(Sequinetal.,2003;Weggeetal.,2004).Theinfrared
method is much more discrete, and is consequently
veryuseful.Indeed,infraredcamerasaremorewidely
used by wildlife researchers than incandescent ash
(Meek et al., 2012). The infrared light emitted by a
seriesofLightEmittingDiodes(LEDs),whichallows
thecameratotakeblack-and-whitepictures,is hardly
visible,althoughtheredlightoftheLEDsisslightly
visible. The most discrete and best solution to avoid
scaringwildlife is to use a camera with a “no-glow”
infraredash(e.g.,BushnelTrophyCamBlack,Covert
Black 60, Reconyx Hyperre SM750, etc.). These
camerasbasicallyfunctioninthesamewayasnormal
infrared cameras, shooting black and white pictures,
butusingLEDsthatemitnovisiblelightatall.
Battery consumption. Battery life can also be a
crucial point to consider when preparing eld work
withcameratraps,especiallyinremoteareas.Several
= Detection zone = Field of view
ab
Figure 1. Diagramoftheeldofviewandofthedetection
zone for two types of camera trap — Schéma des zones
de vue et de détection d’après deux types de pièges
photographiques.
a.detectionzonewiderthantheeldofview—zone de détection
plus large que la zone de vue;b.detectionzonenarrowerthanthe
eldofview—zone de détection plus étroite que la zone de vue.
Table 1.Maintechnicalcharacteristicsofsomecameratrapmodelsfoundonthemarketatthetimeofstudy—Principales
caractéristiques techniques de modèles de pièges photographiques disponibles sur le marché au moment de l’étude.
Brand Model Detection zone
Angle Distance Total
(°)(m)area
(m²)
Field of
view (°)
Trigger
speed (s)
Recovery
time (s)
Resolution
(Mpx)
Price
range
(USD)
Cuddeback AmbushIR,V 7.6 11 8 42 5 0.25 NA 100-200
Scoutguard SG565F,V 64.7 14 116 42 8 1.31 5.1 100-200
Moultrie Panoramic150IR,V 150 NA NA 150 8 0.95 6.2 200-300
Moultrie I-35sIR 40 9 31 40 4 2.5 60 100-200
Wildgame
Innovation
micro6IR,V 15.1 24 76 42 6 1.08 NA 100-200
Uway UM562IR,V,C 60.5 16 133 42 5 1.2 NA 300-400
Leupold RCX-1IR,V 48.2 10 42.5 54 8 0.93 2.8 100-200
Reconyx HC500IR,V 33.4 18.6 100 42 3.1 0.2 1 400-500
Spypoint Live3GIR,V,I 41.9 17 110 42 8 2.7 10 400-500
Primos TruthCamX 45.2 13.7 456 42 1.3 1.2 5 200-300
Spypoint ProX 50 21 82.5 39 12 1.76 10 400-50
Boldcharactersindicateminimaandmaximavaluesfoundforeachrespectivefeature—Les caractères en gras indiquent les valeurs
minimales et maximales trouvées pour chaque caractéristique.

450 Biotechnol. Agron. Soc. Environ. 201418(3),446-454
characteristicsneed tobetakenintoaccount, suchas
thelevelofenergy consumption inmonitoringmode
(when the camera is on and ready to take pictures
if it detects movement) and the level of energy
consumptionfordayandnighttimepictureprocessing.
These variables can vary greatly depending on the
available models and will then vary in suitability
depending on the habitat, the faunal composition
presentinthehabitatandaccessibilityof the camera
for the changing of batteries. For example, in the
caseofanaridhabitatwithfewnocturnalspecies,no
diurnalanimalsspecies,anddifcultaccess,itwould
be better to use a camera that requires little energy
inmonitoringmode (asbatteryreplacementisnot as
frequent) and for nighttime picture taking (as only
nocturnalpicturesaretaken).Thus,batterylifewillbe
maximized.Bycontrast,inthecaseofastudytaking
placeinahabitatwithahighlevelofdiurnalactivity,
amodel that uses as little energy as possible for the
processing of daytime pictures would be preferred.
To extend battery life, some brands (e.g., Reconyx,
Scoutguard,Spypoint)alsoprovidesolarpanels.
Picture resolution. Picture resolution, expressed
in megapixels (Mpx), can vary more than 10fold
between models. Some Primos models take pictures
of relatively low resolution (1.3Mpx), whereas the
Spypoint Pro-X takes pictures up to 12Mpx. The
advantageof lowerresolutionimagesis thattheyare
lessheavytostore so more pictures canbesavedon
agivenmemorycard but, as having lesspixels,they
tendtohavelessdetailsandbelessprecise.Giventhe
largestoragecapacityofmemorycardsnowadays,we
would recommend to select for models with higher
resolution pictures and especially when individual
identicationis needed.Amore detailed and precise
picture can surely help being more accurate when
looking at differences in fur patterns and marks to
differentiate between individuals. However, the
number of pixels advertised by manufacturers must
be considered cautiously because it is not the only
factor affecting picture’s quality. Image sensor, the
componenthousing thepixels,isalso veryimportant
indeterminingpicturequality.Foragivensensorsize,
an increase in the number of pixels is automatically
associated with a decrease in pixel size. Yet smaller
pixelsarelesssensitiveto light, produce more noise
(unwantedsignal)andhaveanarrowerdynamicrange
(i.e. the range of light intensities being captured)
(Nakamura,2005).Itisthereforepossiblethatacamera
with fewer pixels but a larger sensor can produce
pictures of higher quality than a camera with more
pixels packed into a smaller sensor. Unfortunately,
informationonsensorsizeissofarpoorlydocumented
bymanufacturersandwouldneedfurtherinvestigation
andcomparison.
Camera cost. At the time of writting, cameras traps
cost from about USD40 to 1,200, though more than
half(54%)of themodelscomparedinthis studycost
betweenUSD100 to 200.Whilethe cheapestmodels
can have an infrared ash (Hunter GSC35-20IR;
Wildgame Innovations Red4), the most expensive
onescanprovideinstantrecoverytime(Reconyx)and
are able to transmit pictures to cell phones or email
(Reconyx,Spypoint,Covert).
In addition to these main characteristics, various
additional options serving specic research needs
deserveconsideration,suchastheprogrammableburst
modeallowingaseriesofuptovepicturestobetaken
ofthe same trigger event. Some cameras also record
video,withorwithoutsound,whichcanbeusefulfor
reportingonbehaviorrepertoires(Scheibeetal.,2008).
3.3. Sampling methods
Individual behavior. Studies aiming to report on
specicbehaviors(feeding,reproduction,territoriality,
socialinteraction,etc.)mustdirectsamplingeffortsto
placesofinterests(e.g.,saltlicksuses:Blakeetal.,2010;
carcassscavenging:Baueretal.,2005;specichabitat
use:Sequinetal.,2003).Todate,onlyfewstudiesuse
cameratrapsdatatostudyindividualrangingbehavior
andestimatehomerangesize(e.g.,GilSanchezetal.,
2011). Those often have to be completed with data
collected using other protocols such as telemetry or
indirectanimalclues(feedingresiduals,latrines,nests,
etc.),whichcouldexplaintherelativelysmallnumber
ofstudiesestimatinghomerangesize.
Population level studies. Studies dealing with
populationmonitoringusuallyneedstrongersampling
effort and more complex sampling design. To do so,
cameratrapsareincreasinglyusedasanalternativeto
other more traditional methods. However, Gompper
et al. (2006) proved camera traps to be inefcient at
detectingsmallcanids,whichwereotherwisedetected
by scat surveys, DNA analysis and/or snowtracking.
When comparing different methodologies for the
censusofpopulationdiversityandabundance,camera
trappingappearto be themostappropriatemethodin
difcult to access areas compared to line transect or
animaltracksurvey(Silveiraetal.,2003).Usingcamera
traps to estimating population density can involve
complexsampling designandbesubjecttonumerous
biases. Firstly,it is important to consider the bias of
disproportionally samples more easily accessible or
more attractive places for wildlife where detection
probabilityisincreased(Fosteretal.,2011).Thetypical
procedure to characterize an animal population in a
givenhabitatconsistsofsettingupthesamplingeffort
(cameratraps) inarandomorsystematicway(Foster
etal., 2011).Asexplained byRowcliffeetal.(2013),

Useofcameratrapsforwildlifestudies.Areview 451
cameras can be positioned in less or more attractive
places to animals as long as those are proportionally
sampledinregards to their relative occurrence in the
studied ecosystem. Thus, using a grid and a random
number generator, or following a stratied design
allow ones to select positions where to install the
camerasatrandominregardstotheanimals(Rowcliffe
etal., 2013). However,some researchers have set up
their cameras in specic places where the targeted
elusivespeciesarelikelytopass,hopingtomaximize
encounter rate (e.g., Sanderson, 2004; Weckel et al.,
2006); some have even tried to lure animals with
attractive smells or baits (e.g., Trolle et al., 2003;
Garroteetal.,2012).Indeed,placingcameratrapsina
non-randomwayisnotnecessarilyanissueas“itisthe
animalpopulationwithinanareathatisthesubjectof
samplingbyobservationstations,nottheareaitself”as
observedbyBengsenetal.(2011).Secondly,oneneeds
to consider variations in detection probability due to
the material used. The use of incandescent ash at
nightcaneasilyspookthetargetanimalsandnegatively
inuence future visitation rates in the vicinity of the
camera(Sequinetal.,2003;Weggeetal.,2004).Thus,
in the case of capture-recapture sampling or studies
on habitat use of nocturnal species, it is preferable
to avoid using camera models with an incandescent
ash.Inaddition, it is important tomakesure all set
up cameras have sufcient battery life for a given
sampling period. Due to spatial variations in animal
communityortodifferentcameramodels,thenumber
of pictures taken can greatly vary between cameras
andsome can see their batteries getting empty much
morerapidlythan others do. Cameras runningout of
batteriespossiblymissinformation(animalspassingin
theeldofdetectionwithoutbeingphotographed)and
leadtounderestimatedwildlifeestimation.Apartfrom
samplingbias,populationestimateswithlowprecision
is a common issue when using camera traps data.
Sampling design with low detection probability, due
toalownumberofcameratraps,ashortdurationofa
studyorinadequatematerialcanonlypermittoobtain
lowsamplesize,whichitselflimitsourabilitytoobtain
preciseparametersandstronglyaffectsthestrengthof
populationestimates (Foster et al., 2011).As a mean
to increase sample size, setting up two cameras at a
samestation allows obtaining pictures of both anks
formarkedanimalsandcanfacilitatetheidentication
ofindividuals(Kalleetal.,2011;Negrõesetal.,2012).
Intra-community interactions. In the case of seed
dispersalstudies,thecameraisoftensetupsothatthe
visual eld includes the fruits or seeds of interest to
maximize the chances of photographing frugivores
(Seufertetal.,2010;Nyiramanaetal.,2011).Variables
ofinterestherearefrequencyofvisitsandtherelative
contributionofdifferentanimalspeciestoseedremoval.
From personal experience, two remaining limitations
can,however,beidentied.Therstlimitationoccurs
when the camera is positioned close to a fruit/seed
sample so observers can easily quantify the number
of items manipulated by animals. Here, the focal
distancemightbetooclosetobeingabletophotograph
all the animals visiting the area. The camera would
then record a limited number of visiting species and
individualanimals.By contrast,thesecondlimitation
occurs when the camera is positioned to sample the
widest area possible below a fruiting tree canopy,in
orderto systematically record all visitinganimals. In
thisscenario, the focal distance might be too high to
allowobserverstoseeaccuratelythenumberoffruits/
seeds manipulated. An alternative couldbe to set up
two or more cameras at a same location to sample
boththe tree canopy’sshadow and a fruit sample on
theoor.In thelatter case, an alternative toevaluate
species-specic contribution to seed removal could
be to consider visit frequencies per species in the
area.Additionally,seedremovalratecanbeindirectly
assessed with an exclusion experiment (Culot et al.,
2009).
Data analysis.Theidenticationofindividualanimals
is generally made by natural fur marks, injuries, and
coloration patterns (dots, bands). This identication
is, however, always subjective and likely to vary
accordingto theobserverand thuslikelyto affectthe
precisionofestimates.Todiminishtheriskofmistaken
identication, different computer models are able to
help matching pictures of marked individuals (Kelly,
2001;Mendozaetal.,2011).Suchtoolsallowobservers
toimprovetheirabilitytorecognizeindividualanimals
andto be more precise inmaking population density
estimates.
Individualidenticationisacrucialstepinmaking
population estimate. The spatially explicit capture-
recapture technique is increasingly used for this
purpose(e.g.,Efford,2011;Kalleetal.,2011;O’Brien
etal., 2011).This techniqueassumesthatanimalsare
independently distributed in space and that they use
denedhomeranges.Thus,amodelmustberun,which
considers, on the one hand, a population parameter
(populationdensity)and,ontheotherhand,aprocess
of individual recognition. The detection process is
itselfdrivenbyamathematicalfunctiondescribingthe
probability of detecting an animal, which decreases
asthecenterofagivenhomerangegetsfurtheraway
fromacameratrap(Kalleetal.,2011).
Camera trap data are also used to generate
abundanceindicesandgetquickinsightintopopulation
size. However, the power of such indices is limited
compared to true estimates of population density
for different reasons. Firstly, variations in indices
cannotnecessarilybe attributed to true variationsin

452 Biotechnol. Agron. Soc. Environ. 201418(3),446-454
populationsize.Indeed,touseandbeabletocompare
suchindices one needs to makethe assumption that
wildlifedetectabilityisconstantovertime,spaceand
betweenspecies,however,thisiseithernottested,nor
true (Sollman et al., 2012). Secondly, those indices
are rarely calibrated with the actual population and
thusonly givelittleinformationonthetruedynamic
ofpopulationsize(Sollmanetal.,2012).Moreover,a
toolownumberoftrapssetup(replicas)doesnotallow
the calculation of a condence interval (variance)
necessarytoestimatetheexactitudeofindices(Azlan
et al., 2006), though Bengsen et al. (2011) adapted
aGeneral Index Model able to account for variance
whencalculatingpopulationabundanceindices.
Camera traps data such as species detection/
non-detection can also be used in occupancy model
(e.g., MacKenzie et al., 2003; Long et al., 2011) to
predictspeciesoccurrenceand determine population
dynamicparameters.Suchmodelsgeneratedetection
probability data and thus prevent the recording of
falseabsence.This has veryhelpfulimplicationsfor
monitoringelusivespeciesforwhichobservationsare
scarce.
4. CONCLUSION
Dependingonthedatatobecollected,thetargetanimal
species and the type of ecosystem, it is essential to
rstchoosetheappropriate equipment to collect the
dataneeded,asnotallcameramodelswillbesuitable
foraspecicresearchobjective.Giventheincreasing
useof camera trappingbyscientists,we believethat
the available technologies should and will know
improvementsinthefuture.Higherimageresolution
resulting from larger sensor and more efcient
infraredbeam would allow a better identication of
individuals,especiallyfor marked nocturnalspecies.
Evenmorediscreteandfastercameraswouldprevent
spooking animals and get more unblurred pictures.
Next, the implementation of appropriate sampling
protocolsmustbeseriouslyconsidered.In a general
way, we believe that homogenization of detection
probability could improve the use of camera traps
data by diminishing biases and allowing stronger
inter-site and inter-species data comparison. This
couldbedone:
– atthecamerascale,byusingcameramodelshaving
 similarfeatures(detectionzone,eldofview,trigger
 speed,etc.),
– attheecosystemscalebyimplementingstandardized
 samplingscheme(numberofcameras,spacing,and
 placement).
Having a standard sampling protocol would
also permit more solid use of statistical models and
interpretationof results.Theuseof computertoolsto
improvethescienticvalueofpicturesisincreasingly
commonbutalldoesstill not agree basic assumption
requirements. Future development of computer tools
forpopulationdensity,abundance andsiteoccupancy
estimates would need to rely on empirical validated
results on individual habitat use behavior and
populationdynamics.
Acknowledgements
ThisstudytookplaceinthecontextoftheBIOSERFproject
funded by the “Politique Scientique Fédérale Belge”
(BELSPO).
Bibliography
AzlanJ.M. & SharmaD.S.K., 2006. The diversity and
activitypatternsof wild felids in a secondary forest in
PeninsularMalaysia.Oryx,40,36-41.
BabweteeraF. & BrownN., 2010. Spatial patterns of tree
recruitmentinEastAfricantropicalforeststhathavelost
their vertebrate seed dispersers. J. Trop. Ecol., 26(2),
193-203.
BarrowsC.W. et al., 2005. A framework for monitoring
multiple-speciesconservationplans. J. Wildl.Manage.,
69(4),1333-1345.
BauerJ.W., LoganK.A., SweanorL.L. & BoyceW.M.,
2005. Scavenging behaviour in puma. Southwestern
Nat.,50(4),466-471.
BengsenA.J., LeungL.K.-P., LapidgeS.J. & GordonI.J.,
2011. Using a general index approach to camera-trap
abundanceindices.J. Wildl. Manage.,75(5),1222-1227.
BlakeJ.G.etal.,2010.Useofminerallicksbywhite-bellied
spider monkeys (Ateles belzebuth) and red howler
monkeys(Alouatta seniculus)inEasternEcuador.Int. J.
Primatol.,31(3),471-483.
BouchéP.,LejeuneP.&VermeulenC.,2012.Howtocount
elephants in West African savannahs? Synthesis and
comparison of main gamecount methods. Biotechnol.
Agron. Soc. Environ.,16(1),77-91.
Camera Traps cc, 2013. http://www.cameratrap.co.za,
(20/09/2013).
CamposR.C., SteinerJ. & ZillikensA., 2012. Bird and
mammalfrugivores ofEuterpe edulisatSanta Catarina
island monitored by camera traps. Stud. Neotropical
Fauna Environ.,47(2),105-110.
CarthewS.M.&SlaterE.,1991.Monitoringanimalactivity
with automated photography. J. Wildl.Manage., 55,
689-692.
CharruauP. & HénautY., 2012. Nest attendance and
hatchlingcareinwildAmericancrocodiles(Crocodylus
acutus) in Quintana Roo, Mexico. Anim. Biol., 62(1),
29-51.

Useofcameratrapsforwildlifestudies.Areview 453
ChundawatR.S.et al., 2011.Panthera tigris. IUCN 2012,
IUCN Red List of Threatened Species. Version 2012.2,
http://www.iucnredlist.org,(15/11/2012).
CulotL.,HuynenM.-C.,GérardP.&HeymannE.W.,2009.
Short-termpost-dispersalfateofseedsdefecatedbytwo
small primate species (Saguinus mystax and Saguinus
fuscicollis) in the Amazonian forest of Peru. J. Trop.
Ecol.,25(3),229-238.
CutlerT.L.&SwannD.E.,1999.Usingremotephotography
in wildlife ecology: a review. Wildl. Soc. Bull., 27(3),
571-581.
EffordM.G., 2011. Estimation of population density by
spatiallyexplicitcapture-recaptureanalysisofdatafrom
areasearches.Ecology,92(12),2202-2207.
FosterR.J. & HarmsenB.J., 2011. A critique of density
estimation from camera-trap data. J. Wildl. Manage.,
76(2),224-236.
GarroteG. et al., 2012. The effect of attractant lures in
camera trapping: a case study of population estimates
fortheIberianlynx(Lynx pardinus).Eur. J. Wildl. Res.,
58(5),881-884.
Gil-SánchezJ.M.et al., 2011.Theuse of camera trapping
forestimatingIberianlynx(Lynx pardinus)homeranges.
Eur. J. Wildl. Res.,57(6),1203-1211.
GompperM.E.,KaysR.W.&RayJ.C.,2006.Acomparison
of noninvasive techniques to survey carnivore
communitiesinnortheasternNorthAmerica.Wildl. Soc.
Bull.,34,1142-1151.
GrayT.N.E. & PhanC., 2011. Habitat preferences and
activity patterns of the larger mammal community in
Phnom Prich Wildlife Sanctuary, Cambodia. Rafes
Bull. Zool.,59(2),311-318.
GrifthsM., 1993. Population density of Sumatran tigers
inGunungLeuserNationalPark.TigerBeat.Newsletter
Tiger Species Survival Plan,6(2),17-18.
HughsonD.L., DarbyN.W. & DunganJ.D., 2010.
Comparison of motion-activated cameras for wildlife
investigations. California Fish Game, 96(2), 101-
109.
KalleR.,RameshT.,QureshiQ.&SankarK.,2011.Density
of tiger and leopard in a tropical deciduous forest of
MudumalaiTigerReserve,southernIndia,asestimated
using photographic capture-recapture sampling. Acta
Theriologica,56(4),335-342.
KaranthK.U., 1995. Estimating tiger Panthera tigris
populations from camera-trap data using capture–
recapturemodels.Biol. Conserv.,71,333-338.
KellyM.J.,2001.Computer-aidedphotographmatchingin
studiesusingindividualidentication:anexamplefrom
Serengeticheetahs.J. Mammalogy,82(2),440-449.
KoikeS.et al.,2012.Seedremovaland survivalinAsiatic
blackbearUrsus thibetanusfaeces:effectofrodentsas
secondaryseeddispersers.Wildl. Biol.,18(1),24-34.
KuceraT.E. & BarrettR.H., 1993. In my experience: the
Trailmaster® camera system for detecting wildlife.
Wildl. Soc. Bull.,21(4),505-508.
LantschnerM.V.,RuschV.&HayesJ.P.,2012.Habitatuse
by carnivores at different spatial scales in a plantation
forest landscape in Patagonia, Argentina. For. Ecol.
Manage.,269,271-278.
LiuX. et al., 2012. Monitoring wildlife abundance and
diversity with infra-red camera traps in Guanyinshan
NatureReserveofShaanxiProvince,China.Ecol. Indic.,
33,121-128.
LongR.A., MacKayP., RayJ. & ZielinskiW., eds, 2008.
Noninvasive survey methods for carnivores.Washington:
IslandPress.
LongR.A. et al., 2011. Predicting carnivore occurrence
with noninvasive surveys and occupancy modeling.
Landscape Ecol.,26,327-340.
ŁopuckiR., 2007. Social relationships in a bank vole
Clethrionomys glareolus (Schreber, 1780) population:
videomonitoringundereldconditions.Polish J. Ecol.,
55(3),543-558.
MacKenzieD.I. et al., 2003. Estimating site occupancy,
colonization, and local extinction when a species is
detectedimperfectly.Ecology,84(8),2200-2207.
MeekP.D. & PittetA., 2012. User-based design
specications for the ultimate camera trap for wildlife
research.Wildl. Res.,39(8),649-660.
MendozaE., MartineauP.R., BrennerE. & DirzoR.,
2011. A novel method to improve individual animal
identication based on camera-trapping data. J. Wildl.
Manage.,75(4),973-979.
MoruzziT.L.etal.,2002.Assessingremotelytriggered
camerasforsurveyingcarnivoredistribution.Wildl. Soc.
Bull.,30(2),380-386.
NakamuraJ.,ed.,2005.Image sensors and signal processing
for digital still cameras. Boca Raton, FL, USA: CRC
Press.
NegrõesN. et al., 2012. One or two cameras per station?
MonitoringjaguarsandothermammalsintheAmazon.
Ecol. Res.,27,638-648.
NyiramanaA., MendozaI., KaplinB.A. & ForgetP.-M.,
2011.Evidenceforseeddispersalbyrodentsintropical
montaneforestinAfrica.Biotropica,43(6),654-657.
O’BrienT.G. & KinnairdM.F., 2011. Density estimation
ofsympatriccarnivoresusingspatiallyexplicitcapture-
recapture methods and standard trapping grid. Ecol.
Appl.,21(8),2908-2916.
O’ConnellA.F., NicholsJ.D. & KaranthK.U., 2011.
Camera traps in animal ecology: methods and analyses.
Tokyo:Springer.
Oliveira-SantosL.G.R. et al., 2012. Abundance changes
andactivityexibilityoftheoncilla,Leopardus tigrinus
(Carnivora: Felidae), appear to reect avoidance of
conict.Zoologia,29(2),115-120.
PenderR.J.,ShielsA.B.,Bialic-MurphyL.&MosherS.M.,
2013. Large-scale rodent control reduces pre- and
post-dispersal seed predation of the endangered
Hawaiian lobeliad, Cyanea superba subsp. superba
(Campanulaceae).Biol. Invasion,15(1),213-223.

454 Biotechnol. Agron. Soc. Environ. 201418(3),446-454
PereiraP. et al., 2012. Coexistence of carnivores in
a heterogeneous landscape: habitat selection and
ecologicalniches.Ecol. Res.,27(4),745-753.
RoveroF., ZimmermannF., BerziD. & MeekP., 2013.
“Which camera trap type and how many do I need?”
A review of camera features and study designs for a
range of wildlife research applications. HystrixItal. J.
Mammalogy,24(2),148-156.
RowcliffeJ.M. et al., 2011. Quantifying the sensitivity of
camera traps: an adapted distance sampling approach.
Methods Ecol. Evol.,2,464-476.
RowcliffeJ.M.,KaysR.,CarboneC.&JansenP.A., 2013.
Clarifyingassumptionsbehindtheestimationofanimal
density from camera trap rates. J. Wildl. Manage., 77,
876.
SandersonJ.G., 2004. Camera phototraping monitoring
protocol. The tropical ecology, assessment and
monitoring (team) initiative, http://www.teamnetwork.
org/files/protocols/terrestrial-vertebrate/archive/
TEAMCameraTraps-PT-EN-2.0.pdf.,(20/09/2013).
SavidgeJ.A.& SeibertT.F., 1988. An infrared trigger and
cameratoidentify predators atarticialnests.J. Wildl.
Manage.,52,291-294.
ScheibeK.M. et al., 2008. Long-term automatic video
recording as a tool for analysing the time patterns of
utilisationofpredenedlocationsbywildanimals.Eur.
J. Wildl. Res.,54(1),53-59.
SequinE.S., JaegerM.M., BrussardP.F. & BarrettR.H.,
2003. Wariness of coyotes to camera traps relative to
social status and territory boundaries. Lincoln, NE,
USA:UniversityofNebraska–Lincoln.
SeufertV., LindenB. & FischerF., 2010. Revealing
secondaryseedremovers:resultsfromcameratrapping.
Afr. J. Ecol.,48(4),914-922.
SilveiraL., JácomoA.T.A. & Diniz-FilhoJ.A.F., 2003.
Camera trap, line transect census and track surveys: a
comparativeevaluation.Biol. Conserv.,114(3),351-355.
SoleyF.G.&Alvarado-DíazI.,2011.Prospectivethinkingin
amustelid?Eira barbara(Carnivora)cacheunripefruits
to consume them once ripened. Naturwissenschaften,
98(8),693-698.
SollmannR., MohamedA., SamejimaH. & WilltingA.,
2012. Risky business or simple solution – Relative
abundanceindicesfromcamera-trapping.Biol. Conserv.,
159,405-412.
Srbek-AraujoA.C., SilveiraL.F. & ChiarelloA.G., 2012.
The red-billed curassow (Craxblumenbachii): social
organization, and daily activity patterns. Wilson J.
Ornithol.,124(2),321-327.
TrailCamPro.com, 2013. http://www.trailcampro.com,
(20/09/2013).
TrolleM.&KéryM.,2003.Estimationofocelotdensityin
thepantanalusingcapture-recaptureanalysisofcamera-
trappingdata.J. Mammalogy,84(2),607-614.
van SchaikC.P. & GrifthsM., 1996. Activity periods of
Indonesianrainforestmammals.Biotropica,28(1),105-
112.
WeckelM.,GiulianoW.&SilverS.,2006.Jaguar(Panthera
onca)feedingecology:distributionofpredatorandprey
throughtimeandspace.J. Zool.,270,25-30.
WeggeP.,PokheralC.Pd.&JnawaliS.R., 2004. Effectsof
trapping effort and trap shyness on estimates of tiger
abundance from camera trap studies. Anim. Conserv.,
7(3),251-256.
(61ref.)