BookPDF Available

An Introduction to Camera Trapping for Wildlife Surveys in Australia

Authors:
An introduction to
camera trapping for
wildlife surveys
in Australia
Paul Meek
Guy Ballard
Peter Fleming
An introduction to camera trapping for
wildlife surveys in Australia
Paul Meek
Guy Ballard
Peter Fleming
Vertebrate Pest Research Unit
NSW Department of Primary Industries
Forest Road, Orange
2012
An IA CRC Project
ii Invasive Animals CRC
Disclaimer: The views and opinions expressed in this report reflect those of the authors and
do not necessarily reflect those of the Australian Government or the Invasive Animals
Cooperative Research Centre. The material presented in this report is based on sources that
are believed to be reliable. Whilst every care has been taken in the preparation of the report,
the authors give no warranty that the said sources are correct and accept no responsibility for
any resultant errors contained herein, any damages or loss whatsoever caused or suffered by
any individual or corporation.
Published by: Invasive Animals Cooperative Research Centre.
Postal address: University of Canberra, ACT 2600.
Office Location: University of Canberra, Kirinari Street, Bruce ACT 2617.
Telephone: (02) 6201 2887
Facsimile: (02) 6201 2532
Email: contact@invasiveanimals.com
Internet: http://www.invasiveanimals.com
Web ISBN: 978-1-921777-57-8
© Invasive Animals Cooperative Research Centre 2012
This work is copyright. The Copyright Act 1968 permits fair dealing for study, research,
information or educational purposes. Selected passages, tables or diagrams may be
reproduced for such purposes provided acknowledgement of the source is included. Major
extracts of the entire document may not be reproduced by any process.
This document should be cited as: Meek PD, Ballard G and Fleming P (2012). An Introduction
to Camera Trapping for Wildlife Surveys in Australia. PestSmart Toolkit publication, Invasive
Animals Cooperative Research Centre, Canberra, Australia.
Front cover photo: Trial of four camera trap models using tree mounting for small mammal
investigations. Image: Paul Meek.
An introduction to camera trapping for wildlife surveys in Australia iii
Contents
Summary .................................................................................................. 1
Glossary of Terms ....................................................................................... 2
1. Introduction ......................................................................................... 4
2. Use of camera traps in Australia ............................................................... 5
2.1. Privacy, policies and the use of camera traps in Australia ....................... 6
2.2. Signage ....................................................................................... 8
2.3. Animal ethics and licensing ............................................................. 9
3. Selecting an appropriate camera trap ......................................................... 10
3.1. Decision key ............................................................................... 11
4. How Camera Traps Work ........................................................................ 15
4.1. Camera description....................................................................... 15
4.2. Camera trapping studies ................................................................ 16
4.3. Analysing camera trap data ............................................................ 16
4.4. Camera trap types ........................................................................ 16
4.5. Detection zone ............................................................................ 18
4.6. How do PIR sensors work? .............................................................. 20
4.7. Temperature signatures and differentials .......................................... 21
4.8. Non-PIR sensors ........................................................................... 22
4.9. Trigger speed .............................................................................. 22
4.10. Secure Digital (SD) and Secure Digital High Capacity (SDHC) cards ........... 23
4.11. Batteries and other power sources ................................................... 23
4.12. External Batteries ........................................................................ 26
4.13. Camera care and storage ............................................................... 26
5. Camera settings for wildlife surveys ......................................................... 28
5.1. Time, date stamp and temperature recording options .......................... 28
5.2. Sensitivity .................................................................................. 28
5.3. Trigger speed and delays ............................................................... 29
5.4. Number of photos......................................................................... 29
5.5. Flash setting ............................................................................... 29
5.6. Recovery time ............................................................................. 30
6. Field deployment of camera traps ............................................................ 31
6.1. Photographic principles ................................................................. 32
6.2. Camera trap height ....................................................................... 32
6.3. Camera trap direction ................................................................... 32
6.4. Centralising the detection zone ....................................................... 33
6.5. Attachment to poles, trees or tripods ............................................... 35
6.6. Spatial distribution of camera traps .................................................. 39
6.7. Deployment time ......................................................................... 39
iv Invasive Animals CRC
6.8. Weather recording ....................................................................... 40
6.9. Active survey designs .................................................................... 40
6.10. Passive survey designs ................................................................... 43
6.11. Animal responses to camera traps .................................................... 44
6.12. Camera trap emissions: sounds and sights .......................................... 44
6.13. Camera trap security .................................................................... 44
7. Data management and analysis ................................................................ 48
(a) Planning ........................................................................................ 48
(b) Collecting ...................................................................................... 48
(c) Data cleansing ................................................................................ 49
(d) Coding .......................................................................................... 49
(e),(f) & (g) Storing, backing-up and accessing camera trap data ..................... 50
(h) Analysing data ................................................................................ 53
(i) Reporting....................................................................................... 55
7.1. Types of camera trap surveys .......................................................... 55
7.2. Image identification and recognition ................................................. 57
8. Survey designs and methodologies ........................................................... 58
8.1. Small mammal surveys .................................................................. 59
8.2. Medium-sized mammal surveys ........................................................ 62
8.3. Introduced carnivore surveys in Australia .......................................... 64
8.4. Camera trap surveys for non-mammals .............................................. 66
9. Discussion ........................................................................................... 69
9.1. Tips and Hints ............................................................................. 70
9.2. Useful Websites ........................................................................... 71
10. Acknowledgements ............................................................................... 73
11. References ......................................................................................... 74
Appendix 1 - State legislation pertinent to the taking, storage and use of camera trap
photos .......................................................................................... 80
Appendix 2 - Checklists of equipment and set up for field surveys ........................ 82
Checklist 1 before going into the field ..................................................... 82
Checklist 2 - Setting up in the Field .......................................................... 82
Appendix 3 Camera Trapping data sheet ....................................................... 84
List of tables
Table 1. Quick reference guide: camera traps for wildlife surveys (as of June 2012). ......... 10
Table 2. Detection parameters for commonly used cameras (modified with permission from
Trailcampro). ................................................................................................. 19
Table 3. Examples of animal core body temperatures and corresponding upper and lower
ambient temperature limits for optimal PIR detection. .............................................. 21
An introduction to camera trapping for wildlife surveys in Australia v
Table 4. The average detection times from first detection to first image of 21 camera trap
models (data courtesy of TrailcamPro). ................................................................. 22
Table 5. The number of nights required for several Australian species to reach an asymptote
in detection (H = horizontal placement; V = vertical placement). ................................. 40
Table 6. A summary of camera trap settings and methods based on existing published
research or described by the authors of this document.. ............................................ 72
List of figures
Figure 1. A member of the public tampering with a camera trap, set for wildlife monitoring
purposes, within a national park (image: Guy Ballard). ................................................ 7
Figure 2. NSW OEH camera-trap sign that must be deployed in all NSW National Parks and
Wildlife Service Reserves. .................................................................................... 8
Figure 3. A stepwise process for assessing whether remote cameras are a useful tool for a
particular project (Nelson and Scroggie 2009). ........................................................ 12
Figure 4. The Intended Use Flowchart is designed to assist users in deciding on suitable
specifications and functions to suit their objectives. Non-strategic inventory refers to look-see
inventories with no scientific design or purpose. Strategic Inventory are surveys with some
planned approach and design. Resource Condition Monitoring (MER) monitors trends to detect
change. Performance Measuring (MER) refers to changes in response to a remedial action.
Research is where there is clear scientific design and a defined hypothesis. .................... 13
Figure 5. Diagram of a Reconyx HC600 showing the various components that are fairly
standard in all camera traps (courtesy of Reconyx). .................................................. 15
Figure 6. A typical infrared image of a red fox (Vulpes vulpes) traversing a road (image: Guy
Ballard). ....................................................................................................... 17
Figure 7. The detection zone in a Reconyx HC500, HC600 and HC800 camera trap showing the
warm zones (noted by pink) and sectors where animals have to cross to trigger a photo. The
deer in sector 1 would not have triggered the camera (image: Reconyx). ........................ 20
Figure 8. Image showing the hotspots of furred animals. Note the higher values associated
with the face and ears (images: by NASA/JPL-Caltech). ............................................. 21
Figure 9. The power discharge of Lithium, Alkaline and two brands (Ansmann and Tenergy) of
NiMH batteries mAH = milliamps/hour. (Data courtesy of TrailcamPro). ......................... 24
Figure 10. Battery life of three types of AA batteries (n=8) in series in a Reconyx camera trap
(a: alkaline, b: lithium and c: NiMH batteries) (data courtesy of Trailcampro). ................. 25
Figure 11. Finding a place to deploy moisture desiccant can be challenging. The best location
in a Reconyx camera where a canister can be used is in the corner of the housing (image: Paul
Meek). ......................................................................................................... 26
Figure 12. Observed pathway of the sun in the southern hemisphere (courtesy of Museum
Victoria copyright). .......................................................................................... 33
Figure 13. The Cuddeviewer (or similar device) allows you to view images taken by cameras
that use SD cards and compact flash cards (image: Paul Meek). .................................... 34
vi Invasive Animals CRC
Figure 14. Attachment of cameras using wedges and sticks to aim the camera directly at the
focal point (image: Paul Meek). ........................................................................... 35
Figure 15. Trial of three camera trap models using tree mounting for small mammal
investigations (image: Paul Meek). ....................................................................... 36
Figure 16. The Faunatech Rockpod is a sturdy and adaptable tripod-type device (image: Ross
Meggs). ........................................................................................................ 36
Figure 17. Ezi-Aim screw (top) camera trap mounting device for trees and a tripod -type
mounting bracket design (bottom) by KORA (image: Paul Meek). .................................. 37
Figure 18. Outdoor camera mounting bracket and Reconyx steel-post fitting for attaching
camera traps to pickets and posts (image: Paul Meek). .............................................. 37
Figure 19. This generic bracket can be attached to camera traps that have tripod fittings and
screwed into trees or fastened with bolts to posts. Similarly, lengths of steel strap bracing,
cut to length, can be used as a more flexible substitute. In either case, a short piece of
threaded rod is used with wing nuts to secure the device (image: Paul Meek). ................. 38
Figure 20. An example of a method for maintaining food lures used in northern NSW. A tea
infuser containing bait is wired behind a cutlery drainer and suspended on a steel picket. The
ruler can be fixed to the picket to assist with image scaling. By suspending the tea strainer
inside the cutlery drainer the setup not only excludes small- and medium-sized mammals but
ants too (image: Paul Meek). .............................................................................. 41
Figure 21. PVC Vent Cowl used for active surveys of medium-sized mammals on the south-
east coast of Australia (image: Andrew Claridge). .................................................... 42
Figure 22. A wild dog photographed at the carcase of a dead horse on private land in north-
east NSW (image: Guy Ballard and Sam Doak). ......................................................... 42
Figure 23. Pine Marten (Martes martes) bait delivery system to improve the opportunity for
capturing images of the distinctive gula pattern used in image recognition software
(Copyright: Erwin van Maanen). ........................................................................... 43
Figure 24. A security casing for the Reconyx Hypefire (photo courtesy of Reconyx). .......... 45
Figure 25. Schematic diagram of the camera trap security post. a) permanent set up b)
removable ground level plate set up (Meek et al in press)........................................... 46
Figure 26. Dry-erase or chalk boards allow a photo record to be taken at each site with
specific site location details to ensure that images correspond to the correct site (image: Paul
Meek). ......................................................................................................... 49
Figure 27. Diagrammatic interpretation of the detection area and elements of a detection
zone required to calculate chord width and detection area for analysis. ......................... 54
Figure 28. An extracted image of a Lynx being analysed using WildID to determine a known
individual (image: Fridolin Zimmerman). ............................................................... 57
Figure 29. A cairn used to protect camera traps from weather extremes in Central Asia and
limit false detections (Jackson et al 2005). ............................................................. 58
Figure 30. The ‘Mostela’ device for capturing images of fast moving mustelids such as
weasels (image: Jeroen Mos). ............................................................................. 61
An introduction to camera trapping for wildlife surveys in Australia 1
Summary
Internationally, camera trapping is rapidly being adopted for diverse monitoring purposes,
from wildlife research and management to asset protection. There are, however, myriad
cameras of multiple brands with various models, which have different functionality and are
fit for different purposes. It is difficult for any user to fully comprehend which camera trap to
select and how to use it best. Despite an array of publications about camera trapping, most
users learn from ‘doing’.
Through widespread informal consultation with private citizens, public land managers and
research groups, the Invasive Animals Cooperative Research Centre (IA CRC) and the NSW
Department of Primary Industries (Vertebrate Pest Research Unit) identified a need for a
document that brought together a range of information on wildlife camera trapping to
encourage some consistency in the collective approach to camera trapping in Australia. Based
on our collective experiences with camera trapping, this document aims to provide users with
‘one-stop-shop’ information on most aspects of camera trapping for wildlife monitoring and
research purposes, such as suggestions on selecting a fit-for-purpose camera, designing
camera trapping surveys and means of managing and analysing camera trap data. We
proposed some standards and included information on the history of camera trap use to
provide context. We also explained common terms and described how camera traps actually
work.
In preparing this document, however, it became increasingly apparent that it would likely
take many years to provide robust recommendations on specific details of the methodologies
for some surveys. Some of the information in this guide may be quickly superseded as
technology and our understanding of ecology continue to advance, and it is important to
acknowledge that we do not have all the answers. Consequently, we propose to maintain this
as a living document, to be updated as our collective knowledge of camera trapping
advances. For similar reasons, readers will note that we have tried to avoid recommending
particular brands or models of camera traps throughout the document, as they are likely to
be superseded over time.
Importantly, in developing this document we surveyed camera trap users to compose a
standard datasheet and database for site recording. The resulting documents can be
downloaded from www.feral.org.au. A Microsoft Access database can also be downloaded
from this site.
2 Invasive Animals CRC
Glossary of Terms
Burst mode
A camera trap setting that allows continuous images to be taken
following a trigger event, see also rapidfire
Camera trap
A term used to describe a heat and motion sensing camera that
captures images of wildlife automatically
Camera trap set
Connotation of a trap ‘set’ which describes the immediate area
where camera/s are placed
CF card
The acronym for Compact Flash cards, a mass storage device used
by older camera traps, virtually all new models (at the time of
publication) now use SD cards (see below)
Convert surveillance
Use of cameras set to catch illegal actions by people
Delay
A program function available on some models. This setting has
many forms but typically allows the user to set a period of time
where the camera trap is inactive or ‘hibernating’ before or
between photos.
Depth of field
Not often adjustable in remote cameras. This refers to the
aperture setting and its effect on the focus of objects in the front
and rear of the photograph.
Detection zone
The area in which a camera trap is able to detect the heat
signature and motion of a target
Event
The period of time from first trigger to the last photo in a
sequence, where the sequence is encompassed by the extent of
independent behaviour by the target/s
Focal point
Usually the centre of the image (if the photograph is composed
correctly), the subject of interest, the lure, pathway or track
centre or bait device
Field of view
The area captured in a photograph, usually between 35 and 45
degrees
Fresnel lens
A lens used by camera traps to direct infrared energy onto the
passive infrared (PIR) sensor. These lenses are commonly seen in
lighthouses and cause refraction of light.
Incandescent
A white flash used by a camera trap
Lures
A generic term referring to an attractant used to encourage
animals to investigate a specific point within the detection zone.
These may be auditory, olfactory, visual, or some combination of
these in nature.
An introduction to camera trapping for wildlife surveys in Australia 3
Non-strategic
inventory
Ad hoc use of camera traps (ie not using a standard approach to
conduct a survey ‘Look-see’ use)
Performance
measuring
A Monitoring Evaluation & Reporting (MER) term referring to
monitoring changes in response to an intervention
PIR sensor
Passive detectors of Infrared light
Rapidfire
A camera trap setting that allows images to be taken continuously
following a trigger event - see also burst mode
Resource condition
monitoring
An MER term referring to monitoring of population trends to detect
change
SD card
The acronym for Secure Digital cards. A removable digital storage
medium that is currently the standard in camera traps
Sensitivity
A setting, often adjustable, that reflects the camera’s response to
heat in motion for PIR sensors. Higher sensitivity is associated with
more images, and lower sensitivity with fewer images. Increased
sensitivity, however, does not guarantee detection of a target.
Strategic inventory
Refers to survey design. A strategic inventory has a reasoned plan
underpinning data collection.
Time lapse
A program function available on some camera traps. The time-
lapse function, or similar, typically allows a user to prescribe times
of day and/or night when the camera is inactive, regardless of
activity within the detection zone. Some time-lapse cameras (see
below) do not have a PIR and, instead, capture images at
prescribed times or intervals.
Time lapse camera
Camera traps that do not have a PIR sensor (see above) and can be
programmed to take photographs at predetermined times
throughout the day regardless of any triggers
Trigger speed
The difference between detecting heat in motion and capturing an
image. Also known as response time. Slower trigger speed (ie more
time elapsing between trigger and image capture) may decrease
the likelihood of capturing a target.
Walk test
A program function available on some camera traps. Walktest, or
similar, can be used to identify where a camera will respond to
heat in motion. Consequently, it can be used to ‘focus’ the
camera’s detection zone, as desired.
Xenon flash
An incandescent or white flash that illuminates the subject at
night in full colour
Synonyms for camera
traps
Remote cameras, remotely activated monitoring cameras, trail
cameras, spy cameras, wildlife cameras, camera traps, remote-
sensing cameras, remote sensing cameras, game cameras, photo-
trapping, sensor cameras, heat-and-motion sensing cameras
4 Invasive Animals CRC
1. Introduction
The use of camera traps in wildlife monitoring and research has escalated in Australia over
the last 10 years. Camera traps are used as a tool to conduct surveys or record general
observations, often with the inherent assumption that they will result in high quality data
with less investment by staff, thereby improving cost-benefit ratios. This has been reinforced
by claims that camera trapping provides better results than standard surveys, such as live-
trapping (Paull et al 2011). Such assumptions and claims, however, need validating for the
range of species and situation of interest. That is, considerably more research is required
before many existing methods can confidently be replaced with camera trapping alternatives.
Australian agencies currently use camera traps for two purposes:
covert human surveillance where damage or pollution is a threat to property (eg arson
or poaching)
wildlife inventory, monitoring and research (eg assessing the impact of wildlife
management interventions).
This manual is focused on the latter.
In the absence of best practice guidelines, the types of camera traps people use, how they
use them and how they store and analyse the resulting data vary considerably. In many cases,
despite good intentions, camera traps are being purchased and used in ways that are not fit
for purpose.
To this end it is important to note that the uses of camera traps are still being refined. There
is enormous progress to be made in regards to survey methods, standards and the choice of
fit-for-purpose equipment. Consequently, this document is a ‘living resource’ that provides
basic camera trap information and will be constantly updated to cater for the rapid
development of camera trap technology, survey methods and analytical techniques.
For now, this document has been prepared to:
provide useful background information on camera trap technology
describe the components and functionality of camera traps
provide a decision guide for camera trap choice
outline survey designs, methods and data management
recommend a camera trapping protocol for a range of fauna surveys.
A standardised term for the technique of using camera traps for measuring animal
populations, behaviour and activity is controversial. Throughout this document, the term
‘camera trap’ (see O’Connell et al 2011) has been adopted as a standard. It is
interchangeable with terms such as ‘remote camera’, ‘motion sensing camera’, ‘trail
camera’, ‘game camera’ and ‘sensor camera’.
An introduction to camera trapping for wildlife surveys in Australia 5
2. Use of camera traps in Australia
One of the first reported camera trap studies in Australia was conducted in 1960 by Inspector
Hanlon, Tom McMahon and Eric Guiler on the Woolnorth expedition to find the Thylacine in
Tasmania (Guiler 1985). This study trialled the deployment of a Bolex movie camera attached
to what was probably a snare cable on a pop hole in a fence. The camera produced single-
frame black and white photos using a white flash. On the third Woolnorth expedition in 1961,
five camera traps were used (Guiler 1985). These units comprise G.45 aircraft 8 mm movie
cameras with lighting and a treadle-plate trigger mechanism that were set below pop holes in
fence-lines. In 1966 another camera trapping of the Thylacine was attempted using a
modified version of the G.45 camera (Guiler 1985) but without success.
This early pioneering work resulted in the deployment of 25 camera traps between 1968 and
1972 by Jeremy Griffiths and James Malley, later to be joined by Robert Brown (Guiler 1985),
the future leader of the Australian Greens Party. Following on from this early use of camera
traps was a parallel camera trap surveys for Thylacine by Steven Smith (Smith 1981) and Eric
Guiler (Guiler 1985). Smith rigged up a Pentax 35mm MX motor drive camera with 40 mm or
50mm lenses and a bulk film magazine (250 pictures). A Metz 45 CT-1 flash was attached and
the unit was linked to a Sick Optik Elektorinik infrared source with an email relay and 5 m of
cabling. Reflectors were set up a few metres from the camera and an infrared beam was
projected across an animal path. As the beam was broken, the system automatically triggered
the camera to take photos. Camera taps were left in the field for seven days during June-
September 1980 and resulted in 111 camera-days with 420 photos of nine species but no
Thylacine. Conversely, Guiler had 15 camera traps built at a cost of AUS $25 000 in 1978 that
are probably the first step towards a modern-day designed camera trap. The device used
pulsed infrared beams, circuit boards and a delay cycle; it even had a display to show the
number of events recorded (Smith 1981; Guiler 1985). Despite the failed objectives of
detecting a Thylacine, they were the pioneers of early camera trap developments between
1960 and 1980.
Carthew and Slater (1991) were the first authors to publish scientific papers that featured the
use of camera traps in Australia. These authors used an early form of camera trap to monitor
pollinators of flowers (Carthew 1993). Throughout the 1990-2000, the use of camera traps in
wildlife research changed dramatically from primitive video cameras (Belcher 1998) to
manual treadle mechanism (Glen and Dickman 2003), heat sensor triggers (Belcher 2003) and
the first genuine infrared automated cameras distributed by Faunatech, the Digicam DC110
(Claridge et al 2004; Hayes et al 2006). In the period post-2004, the use of camera traps in
Australia changed again as mainstream devices from the USA entered the market, and
researchers saw great opportunities for the adoption of this new tool for a range of species
and purposes (Cowled et al 2006; Nelson and Scroggie 2009; Vine et al 2009; Claridge et al
2010; Meek 2010; Robley et al 2010; Paull et al 2011).
6 Invasive Animals CRC
2.1. Privacy, policies and the use of camera traps in
Australia
It is important that camera trap users understand their relevant legal responsibilities, prior to
deploying these devices. Generally the use of camera traps for wildlife-related purposes is
not legislated by Australian laws. Nonetheless, the deployment of cameras where images of
people are taken incidentally is regulated by privacy and workplace legislation (Appendix 1).
Privacy legislation varies considerably between the states, territories and within
organisations. In Australia it is not illegal to take and publish a photo of a person (see
http://photorights.4020.net). If the images are being used in a legal proceedings, however,
then law regulates how the information can be used and stored and ultimately how long the
evidence can be kept before having to be destroyed. In Australia it is widely accepted that
our laws do not provide a legal framework for the right of privacy to the individual (Butler
2005). It is essential that camera trap users seek clarification of their legal responsibilities,
relating to camera trap use, data storage, analysis and publication within their state or
territory. In NSW, for instance, there are legal obligations under the Surveillance Devices Act
2007. This Act and associated legislation govern how the agency must use cameras to protect
the privacy of the public and the community.
In the case of camera traps used for wildlife monitoring and research purposes, there is no
intent to capture images of people, meaning that these legal considerations are not
applicable. All users, however, have a responsibility to ensure that a code of practice for
deployment of camera traps for wildlife monitoring is adopted. There are two main issues of
concern from a land management agency perspective: workplace issues and public matters.
The minimum requirement before deployment of devices should be to advise local staff that
cameras are being deployed in a reserve without being specific of when and where. Secondly,
the placement of signs at main entry points to a reserve will provide a warning that camera
traps could be deployed. The exact placement of signs is critical as it is important not to
direct the public to expensive equipment in the area that may result in theft, damage and
large numbers of additional images being captured (Figure 1). Placing a sign may also attract
would-be-thieves to an area and result in damage or theft of equipment.
It is implicit that camera traps placed for wildlife purposes are not set to deliberately capture
images of people the intent must be for wildlife investigation. Nonetheless, if images of
people are captured during surveys, the onus is on the investigator to manage such images
appropriately, according to the relevant policy of your organisation. Under all of the
surveillance acts, there are legal requirements for the storage of data collected from optical
surveillance devices, including storing in a secure place and destroying the records after a
defined period (eg five years). This obligation is the responsibility of all users. The privacy of
all individuals must be respected, and where images indicate an illegal activity, appropriate
recording of the event must be formally reported to a senior officer. This form of evidence
becomes a matter for surveillance devices acts and regulations.
An introduction to camera trapping for wildlife surveys in Australia 7
Figure 1. A member of the public tampering with a camera trap, set for wildlife monitoring purposes,
within a national park (image: Guy Ballard).
For instance, under Part 1 Section 3 of the NSW Surveillance Devices Act 2007 No 64 the
following excerpt is relevant.
3 Relationship to other laws and matters
(1) Except where there is express provision to the contrary, this Act is not intended to
affect any other law of the State that prohibits or regulates the use of surveillance
devices.
Note: the Workplace Surveillance Act 2005, for instance, contains certain requirements in
relation to camera surveillance. The applicable requirements of both acts will need to be
complied with if camera surveillance is carried out.
In the event that camera trap theft is an issue and they are also being used to protect an
asset, the relevant legislation for surveillance devices is ‘switched on’ and operators must be
aware of their obligations for privacy. This will increasingly be an issue for a trapping given
the significant investment in the technology and the escalating occurrence of theft. Covert
and telephony technology already exists that can be simultaneously deployed to prevent theft
of equipment, and as it becomes more affordable, this issue will become more relevant.
In 2005 a legal decision handed down by a judge in Queensland (Skoien SDCJ) in the case of
Grosse vs Purvis, significant damages were handed down under a Tort of Invasion of Privacy.
In general terms this decision has paved a way forward for prosecution where damages to an
individual has resulted from their right to be left alone (Butler 2005). Skoien SDCJ stated
that in the above case the Tort of Invasion of Privacy was relevant under the following
causes:
8 Invasive Animals CRC
A willed act by the defendant;
which intrudes upon the privacy or seclusion of the plaintiff;
in a manner which would be considered highly offensive to a reasonable person
of ordinary sensibilities; and
which causes the plaintiff detriment in the form of mental, physiological or
emotional harm or distress, or which prevents or hinders the plaintiff from doing an
act which he or she is lawfully entitled to do.
It could be argued that images of a person, or video recordings where sound is also recorded
on a camera trap, is an impingement on an individuals privacy. In a recent review of the
Privacy Act 1988 (Australian Law Reform Commission 2006), listening devices were reviewed,
but cameras for wildlife monitoring were not considered. In light of the findings of Grosse vs
Purvis and the subsequent interest in developing legislature to recognise privacy, it is more
than likely that this issue will confront camera trap users in the future.
A useful website that summarises much of the privacy legislation in Australia can be browsed
at http://www.privacy.org.au/Resources/PLawsST.html.
2.2. Signage
Some users are reluctant to place signs to warn that camera traps have been set in the field.
The NSW Office of Environment and Heritage (OEH) stipulate that signs (Figure 2) must be
placed in areas where camera traps are being deployed and that they are to be placed at
major road entrances but not necessarily close to the device. This warns people of camera
trap deployment but not of the exact location. No other examples of equivalent signage were
available from similar agencies at the time of publication.
Figure 2. NSW OEH camera-trap sign that must be
deployed in all NSW National Parks and Wildlife
Service Reserves.
An introduction to camera trapping for wildlife surveys in Australia 9
2.3. Animal ethics and licensing
Some Animal Ethics Committees require investigators to apply to use camera traps in research
involving animals. Camera trap users are encouraged to consult with their committee to
determine requirements within their agency or institution prior to surveys. Although this may
seem onerous, there is no doubt that some types of camera traps affect wildlife behaviour
(Gibeau and McTavish 2009). Under the NSW OEH Animal Ethics Committee, for example, the
following definition of research and monitoring was developed to evaluate whether ethics
approval is required:
Use of remote cameras is research when the purpose of gathering the information is to
survey the presence/absence or relative abundance of species in a systematic way, regardless
of whether attractant or other behavioural modifiers are used. For example:
comparisons between different habitat types or between the same habitat in
different locations
comparisons over time or as part of a before-after treatment monitoring.
Use of remote cameras is not research when the purpose of gathering the information is to
identify species or individuals using a point location or small area, such as a water hole, a
camping ground, a nest, when no attractant or other behavioural modifier is used, or lethal
bait is used as a component of approved pest management activities. The information may
be used to:
determine whether to adopt a particular on-ground asset or pest management action
at a point location
determine whether a rare species or species previously not known from the area is
utilising the location (eg a bird of prey at a nest site).
Under this definition the placing of cameras on roads to determine pest animal activity:
is research when the placement of cameras will be systematic (eg many roads will be
surveyed prior to baiting, in order to determine where pest management efforts
should be put, and surveyed after baiting to determine baiting success.)
is not research when cameras will be placed along a single road or stretch of road
simply to determine if pest animals are using the road, but there is no intention to
compare abundances with other locations or remonitor after pest control activities.
At the time of publication, Victoria required animal ethics approval to use camera traps for
wildlife research (Nelson and Scroggie 2009). In Western Australia, the Department of
Environment and Conservation (DEC) Animal Ethics Committee does not require licensing or
competency standards for staff (Davis 2011a) and there are no requirements in Queensland,
Victoria, South Australia, Northern Territory or Tasmania.
10 Invasive Animals CRC
3. Selecting an appropriate camera trap
Camera traps are usually acquired for one of the following two reasons:
‘opportunistic’ - where money becomes available at short notice (eg at the end of the
financial year)
‘planned purchases’ - where a predetermined number of camera traps are bought for
a specific project.
Either way, deciding which camera traps to choose can be challenging for the novice and
managers who have to approve purchases. Several approaches can assist in making such
decisions, and a couple of examples are presented below.
The features and specifications of camera traps vary enormously and need to be evaluated
when choosing between models. The intended use will often restrict the choice of camera
trap but cost per unit must also be considered. Often the choice of camera trap to buy is
driven by cost (Meek 2012) based on the perception that more devices are better than less,
despite the consequences for data quality. Alternatively, some will buy camera traps like they
would a normal camera: buying a few, or even one, expensive camera traps to ensure they
have quality images.
Below is a quick guide to the range of prices (in 2012 AUS$) that can be paid for camera traps
and a stepwise decision keys to choosing the best camera traps for general use to survey
wildlife.
Table 1. Quick reference guide: camera traps for wildlife surveys (as of June 2012).
High end (AUS $500 1000)
Mid range (AUS ~$500)
Low end (AUS <$400)
Reconyx HC600/HC800/HC900
Reconyx HC500
Scoutguard SG560VB
Pixcontroller DigitalEye
Bushnell Trophy Cam
Moultrie M80 or M100
Scoutguard SG560
Euovision 565
Buckeye Cam Orion
Cuddeback Capture
An introduction to camera trapping for wildlife surveys in Australia 11
3.1. Decision key
Step 1. Opportunistic Purchase
I can afford to purchase:
1-3 cameras Q: Can I borrow some camera traps instead?
If not, refer to the recommendations in this manual and/or go to
http://www.trailcampro.com/trailcameratests.aspx
5-10 cameras This is an investment that requires careful consideration of the intended use.
Q: Can I pool my resources with other people and together increase our
camera trapping capacity?
Q: What are the other requirements of buying these camera traps and does
that influence my choices?
Please read the Planned Purchase decisions key.
10 + cameras This purchase should be strategic and planned.
Please read the Planned Purchase decisions key.
Step 2. Planned Purchase
Before buying camera traps, consider whether camera trapping is the best tool for the
planned survey or monitoring. Nelson and Scroggie (2009) provide a Decision Key to assist
novice camera trappers in making logical choices in selecting camera traps that are
appropriate for specific wildlife monitoring and surveys (Figure 3).
The Intended Use Flowchart below (Figure 4) is a further process aimed at assisting interested
parties in deciding whether camera traps are the most suitable survey method to carry out
the surveys you require.
12 Invasive Animals CRC
Figure 3. A stepwise process for assessing whether remote cameras are a useful tool for a particular
project (Nelson and Scroggie 2009).
Camera Traps not suitable
No
No
Yes
Do the camera traps have
infrared flash and pseudo
video options?
Yes
Yes
No
Survey objectives
Confirm species presence
Species inventory
Investigate behaviour
Are resources available to
purchase a small pool of
camera traps (e.g. 1+) +
additional equipment?
What is the target
species?
Small/medium/large mammal
Can the target
species be readily
identified from
photographs?
Camera traps suitable
Are resources available to assist
identification (i.e. field guides,
species experts)?
End of process
Are similar species likely
to be present?
START HERE
Yes
Yes
No
Cameras well suited to
behavioural observations
Cameras suitable for
confirming species presence
and species inventory
No
An introduction to camera trapping for wildlife surveys in Australia 13
Figure 4. The Intended Use Flowchart is designed to assist users in deciding on suitable specifications
and functions to suit their objectives. Non-strategic inventory refers to look-see inventories with no
scientific design or purpose. Strategic Inventory are surveys with some planned approach and design.
Resource Condition Monitoring (MER) monitors trends to detect change. Performance Measuring (MER)
refers to changes in response to a remedial action. Research is where there is clear scientific design and
a defined hypothesis.
14 Invasive Animals CRC
Choosing the right camera trap can be complicated because purchasers often intend to use
the devices for more than one purpose. Finding a camera trap that satisfies all the needs of
wildlife management and research is not possible (Meek and Pettit submitted). Few camera
traps offer both white and infrared flash with stills and video and sound capacity with quick
trigger speeds, sensitive passive infrared (PIR) sensors and a great range of settings. This
means choosing which model will be most suitable is complex and often impossible. Price is
the most important factor in camera choice (Meek 2012), followed by infrared or white flash.
These two types of camera traps are often mutually exclusive in wildlife surveys because they
have distinctly different purposes. For example, using an infrared camera trap for small
mammals where identification is difficult is pointless, as is using white flash camera traps for
surveys where it is important not to influence the behaviour of animals. In the following
sections, information will be provided on camera traps and their characteristics to help users
make informed decisions about which equipment is fit for purpose.
An introduction to camera trapping for wildlife surveys in Australia 15
4. How Camera Traps Work
The range of camera traps brands, models and types on the market is enormous, and the
functions of camera traps vary with every new model. Understanding exactly what is available
in a camera trap and how to use the functions effectively is critically important in choosing a
suitable model.
4.1. Camera description
Camera brands and models vary substantially, creating a degree of inconsistency between
them. Camera traps from popular brands, such as Scoutguard, Bushnell and Moultrie, can all
look similar and can even be made in the same factory. Camera traps with separate camera
units, such as those made by Pixcontroller, can be quite different in design, despite the use
of similar components. Common camera trap components are identified in Figure 5. Although
many key components are universal (eg lens, PIR sensor, LED arrays and light metres), others
can be identified by jargon that varies between manufacturers (see Glossary of terms).
Figure 5. Diagram of a Reconyx HC600 showing the various components that are fairly standard in all
camera traps (courtesy of Reconyx).
16 Invasive Animals CRC
4.2. Camera trapping studies
Meek (2012) found that a common limitation of camera trapping programs, internationally,
was the absence of a robust monitoring design. Furthermore, many users are implementing
camera trap surveys without even a basic understanding of the limitations of the technology
they are employing.
There are many examples of where substandard camera traps have been chosen as a tool for
detecting animals yet no calibration to evaluate detection probabilities have been
considered. This is often related to costs (Meek 2012) where researchers have a limited
budget but need many cameras, often having to sacrifice quality for quantity (Karanth et al
2011). In some studies, researchers have to move camera traps around the landscape over
long periods of time to increase sample sizes. It is in the early stages of planning that the
numbers and types of camera traps to be used must be carefully considered.
Karanth et al (2011) provide a detailed summary of considerations related to camera trap
survey designs, including season, survey duration, population closure, camera trap spacing
and placement, sample area coverage and appropriate analyses.
4.3. Analysing camera trap data
The specific design of camera trap surveys depends on many variables. It is beyond the
purpose of this document to provide specific guidance on this subject because of the
complexities and nuances that are introduced as different camera trap types are used and
different target species are selected. We strongly recommend seeking advice from a
biometrician before beginning your survey.
4.4. Camera trap types
There are two broad types of camera trap currently in use throughout Australia for wildlife
research: white flash and infrared flash. The flash power (range) determines the depth of
view and clarity of the picture taken, and there is considerable variation between camera
brands and models.
Camera traps with infrared flash
Infrared cameras use arrays of LEDs that emit infrared light, mostly in the range of 700-1000
nm. The images taken by these cameras are often in grey-scale (Figure 6) or may have a
reddish-pink tinge. Infrared flashes tend to be less obvious to wildlife than incandescent ones.
Infrared flash also uses less energy than incandescent flash, and models employing them tend
to have quicker trigger speeds. The recent development of white LED technology has
overcome this issue.
An introduction to camera trapping for wildlife surveys in Australia 17
Figure 6. A typical infrared image of a red fox (Vulpes vulpes) traversing a road (image: Guy Ballard).
Camera traps with incandescent or white flash
Incandescent flash mostly use xenon gas technology to enable cameras to record clear, colour
images during the day or night. Xenon white flash (the most common) illumination is bright
but brief. Other gases, such as Krypton and Argon, can be used but these gases have different
spectral output and some approach near infrared. The new Reconyx PC850 is the first and
only camera trap to use white LED flash for illumination providing a fast trigger speed, good
clarity and no white-out effects.
White flash cameras have an important role in wildlife research and monitoring because often
researchers rely on detailed colouration and markings to distinguish their target animals from
other individuals or species. White flash cameras can provide sharp, full colour images even
at night, often with better resolution than infrared models. Relative disadvantages of white
flash models, however, can include increased battery draw, slower trigger speed and
potential to change animal behaviour by subjecting them to bursts of visible light.
Passive Infra-Red (PIR) and Active Infra-Red (AIR) cameras
Passive infrared (PIR) detection refers to the sensing capacity of the camera and at what
point in the field of view the camera will detect heat and motion and take photos. PIR
detects the difference between the air temperature (ambient) and the animal’s body
temperature. This is the most commonly used infrared system in camera traps. Active Infra-
Red (AIR) devices rely on two units spaced apart where an infrared beam is projected across a
defined pathway. When the beam is broken, the device will take photos as programmed.
18 Invasive Animals CRC
These cameras are often more expensive and more cumbersome for remote field work but are
much more accurate (R. Meggs, personal communication, 2012).
Still or video
Video function is available on some models, either with still photos or as video only. Video
can be a useful method of capturing behavioural information, particularly if the objective is
to record a sequence of actions or movements as a part of a study (eg observing how an
animal interacts with traps or baits). Video cameras tend to use more battery power than
still-photo cameras and will not necessarily provide additional data in some instances. For
example, a series of stills, say five or 10 in succession, may be animated with computer
software to simulate video footage. Obviously, the particular issue of interest, as well as the
response time, and lag period between successive photos, will be important factors in
comparing still cameras with video cameras.
Time-lapse camera traps
Some camera traps, such as Brinno and the Wingscape BirdCam, are designed solely to take
photos using a time-lapse setting. They are not triggered by heat and movement. Other
cameras, such as Reconyx HC600, offer dual functionality, using both time lapse and heat in
motion activation. The settings will vary with each camera, but in essence, the camera traps
allow intervals to be set between photos, and some allow video and stills to be taken. These
cameras are useful where heat differentials prohibit PIR effectiveness or where the subject is
fauna such as insects and ectotherms that may not trigger a PIR. Time lapse camera traps
may also provide an additional tool in situations like the desert where heat signatures can be
masked by background heat. Some PIR camera traps also provide time-lapse options to be set
together with heat and motion sensing settings.
4.5. Detection zone
A camera trap’s detection zone is not necessarily equal to its field of view. Detection zones
vary between camera trap models, and for some, only a small proportion of the field of view
actually corresponds with the camera’s detection zone. You should check the manufacturer’s
specifications to confirm the details of the detection zones of camera traps of interest (Table
2). For instance, detection zones are not always conical in shape.
Your choice of detection zone (ie camera model) should match your needs. A narrow
detection zone requires the animal to move into a precise range and will not capture animals
that move outside of that range. Hence, cameras with a narrow detection zone are best used
in situations where the animal is being attracted to a point source of interest, such as a
feeding station. Wide detection zones will often match the width of the camera trap field of
view or just inside this area and are more suited to passive surveys or diffuse sources of
interest. Hence, wider detection zones may be likely to pick up animals sooner and capture
more images or video time.
An introduction to camera trapping for wildlife surveys in Australia 19
Table 2. Detection parameters for commonly used cameras (modified with permission from
Trailcampro).
Model
Detection
width @
9.1m
Detection
angle
Field of
view (FOV)
width
FOV
angle
Detection
range
Detection
zone area m2
Reconyx HC500
6.7
40°
6.7
40°
30.5
324.1
Bushnell Trophy
Cam
14.3
75°
7.0
42°
15.8
164.3
Scoutguard
SG550
7.3
44°
7.3
44°
15.2
89.1
Leaf River IR-5
6.4
37°
6.1
36°
17.7
100.9
Scoutguard
SG580M
7.6
45°
7.3
44°
11.6
52.7
Scoutguard
SG565
11.3
53°
7.6
45°
9.1
38.6
Moultrie I65
6.1
36°
5.8
35°
10.7
35.8
Moultrie I35
6.7
40°
6.7
40°
9.4
31.1
Recon Viper
2.4
15°
6.1
36°
11.0
15.8
Cuddeback
Capture IR
2.1
14°
6.7
40°
11.0
14.7
Predator
Traileye IR
7.6
45°
7.3
44°
14.9
87.5
Stealth Cam
Unit
11.6
63°
7.0
42°
11.6
73.7
Wildgame
Innovations X6C
9.8
56°
7.0
42°
16.2
127.5
Uway
Nighttrakker
NT50
11.0
62°
7.0
42°
13.7
101.7
Primos Truth
Cam X
11.3
53°
7.0
42°
13.7
87.0
Spypoint IR-8
8.5
50°
6.4
37°
13.7
82.0
Primos Truth
Cam 60
2.4
15°
7.6
45°
21.0
57.9
The shape and way the detection zone works is fundamental to understanding how best to use
your camera trap. The Reconyx range of camera traps has a unique detection zone in their
non-professional range (Figure 7) that requires an animal to move within a horizontal and
vertical zone before the camera will detect heat and movement. If an animal does not cross
the sectors shown below from 1-6, while moving within the pink or ‘warm’ zone, the PIR
sensor will not detect movement.
20 Invasive Animals CRC
Figure 7. The detection zone in a Reconyx HC500, HC600 and HC800 camera trap showing the warm
zones (noted by pink) and sectors where animals have to cross to trigger a photo. The deer in sector 1
would not have triggered the camera (image: Reconyx).
Detection zone type can be selected in some brands prior to purchase. For example, the
Reconyx professional range offers two options: narrow or wide. PIR settings can be changed in
some models; the Leopold RCX-2, for instance, has Dual Sensor Technology (DST). This allows
the user to set the detection zone to either 10 degrees (ideal for focusing on a bait station) or
45 degrees, where a wide detection is required (Meek and Pettit submitted).
4.6. How do PIR sensors work?
There are two types of PIR
a Ceiling sensor, which minimises dead zones
a Dual Element, where there are breaks in the detection zone bands, thereby creating
a detection zone error.
Most brands use dual element PIRs. This has implications for survey design as it changes
detection probabilities. It is especially important where camera traps are being placed facing
down or directly in the path of an animal (eg along a log or directly down a road).
A key limitation of PIR sensors is their inability or poor performance in detecting differences
between the target and the background in some situations. Where the temperature
differential between the background and target is low, some sensors may be incapable of
detecting target animals right in front of the camera (see below). This can be especially
problematic in desert or beach situations, where background heat and light mask the target,
or where reptiles are being observed. Ideally, the temperature differential between the
target and the background needs to be greater than five degrees Fahrenheit (J. Thinner,
personal communication, 2012).
An introduction to camera trapping for wildlife surveys in Australia 21
4.7. Temperature signatures and differentials
Animals have a heat signature. The intensity of heat produced by animals varies between
different parts of the body, with eyes and face literally being hot spots (Figure 8). Table 3
shows the body temperature of a number of animals and the ambient temperature where PIR
sensors may become unreliable in terms of recognising the temperature differential between
the animal and the background. These data suggest that when ambient temperature ranges
between 31.5 and 36.5 °C until 42.5 °C, camera trapping can be unreliable for some species.
Figure 8. Image showing the hotspots of furred animals. Note the higher values associated with the face
and ears (images: by NASA/JPL-Caltech).
Table 3. Examples of animal core body temperatures and corresponding upper and lower ambient
temperature limits for optimal PIR detection.
Animal
Body Temperature °C
Optimal PIR
detection below this
temperature °C
Optimal PIR
detection above this
temperature °C
Baboon
38.1
35.1
41.1
Camel*
34.5-41.0
31.5
44
Cats
39
36
42
Cattle
38.5
35.5
41.5
Chicken
42
39
45
Dogs
38.9
35.9
41.9
Elephants
36.5
33.5
39.5
Goat
39.5
36.5
42.5
Horse
38
35
41
Pig
39
36
42
Rabbits
38.3
35.3
41.3
*The camel's body temperature will vary with the time of day and water availability. When a camel is
watered daily its body temperature rises from 36.5°C in the morning to 39.5°C at noon, if the animal
has no water, the temperature range is 34.5°C to 41°C.
22 Invasive Animals CRC
4.8. Non-PIR sensors
There are alternative options to PIR sensors available for some camera traps. Pixcontroller
have manufactured their current camera traps to use seismic sensors (vibration), magnetic
circuit sensors, pressure plates and normally-open-normally-closed switch sensors. These
devices provide other detection options for situations where PIR sensors may be relatively
unreliable. Historically, camera traps used passive infrared sensors (ie the camera trap was
triggered when an infrared beam was broken by the animal). Other researchers have used
treadle plates or trip wires to trigger camera traps (Glen and Dickman 2003).
4.9. Trigger speed
Trigger speed is an important function for many wildlife surveys. Table 4 summarises trials on
21 models of cameras, each with n=5 units, to determine their trigger speeds.
Table 4. The average detection times from first detection to first image of 21 camera trap models (data
courtesy of TrailcamPro).
Model
Average Time
Reconyx HC500
0.197 s
Reconyx HC600
0.203 s
Leupold RCX-1
0.937 s
Leupold RCX-2
0.963 s
Spypoint IR-8
1.133 s
Bushnell Trophy Cam Black Flash
1.300 s
Bushnell Trophy Cam
1.344 s
Wildview Extreme 5
1.377 s
Scoutguard SG580M
1.449 s
Primos Truth Cam 35
1.557 s
Uway NightXplorer NX50
1.567 s
Moultrie M-80
1.581 s
Moultrie M-100
1.648 s
Stealth Cam Archer's Choice
1.760 s
Scoutguard SG565
1.858 s
Stealth Cam Unit
2.165 s
Bushnell Trophy Cam Black Flash XLT
2.438 s
Stealth Cam Sniper Pro
2.607 s
Moultrie D55 Incandescent
2.674 s
Moultrie D55 IR
2.681 s
Stealth Cam Rogue IR
4.206 s
An introduction to camera trapping for wildlife surveys in Australia 23
A ‘fast’ trigger speed minimises the time between detection and image capture, thereby
increasing the probability of a target being recorded. ‘Slow’ trigger speed can result in
images being taken without the target in them. Fast trigger speeds may be unnecessary if
your target will be within the field of view for some time (eg at a feeding station) and you
only require presence-absence information. If, however, the target is likely to be within the
field of view for a brief period, faster trigger speeds will likely increase your probability of
detection.
4.10. Secure Digital (SD) and Secure Digital High Capacity
(SDHC) cards
At the time of publication, Secure Digital (SD) and Secure Digital High Capacity (SDHC) flash
memory cards ranged in storage capacity up to 64GB. In addition to capacity, card ‘speed’ is
important as it relates to how quickly data can be written to the card from a source, such as
a camera trap. This is particularly important for camera trap models that have multiple image
and/or high-quality video function. Historically, card speed was described as a ‘Class’ with an
‘x rating’, but the new measurement unit is called the ‘Speed Class Rating’. Camera traps
require a fast speed class. Class 2 is suitable for most camera trapping uses, and there is
currently no speed-related benefit in those cards higher than a class 4. Some manufacturers
make specific recommendations regarding cards so it is important to consult the manual prior
to purchase. Camera manufacturer Pixcontroller have been using Eye-cards with Wi-Fi
capabilities that enable images to be sent to a home computer, iPhone, iPad or Android
device.
4.11. Batteries and other power sources
Like other vital accessories, batteries should be selected after co nsulting the camera trap
manual. Some recent camera traps use a gel cell battery, but the three most common battery
types used are Lithium, Nickel-Metal Hydride (NiMH) and Alkaline. The relative life or
available power of these types of batteries varies (Figure 9), and their performance will also
be affected by weather extremes.
Lithium
Lithium batteries are recommended for many camera traps because of their sustained
capacity and high-power output (Figures 9 and 10). They are also unaffected by extreme cold
weather. Their power resilience compared to other battery types is unsurpassed with power
being delivered to the camera trap until <20% of power remains (Figure 9).
NiMH (Rechargeable)
Nickel-Metal Hydride (NiMH), or rechargeable batteries, have many advantages over alkaline
and lithium batteries in that they are a multiple-use battery, and depending on the brand,
they hold their charge for a long time (Figures 9 and 10). The initial cost of NiMH may be
relatively high, but it has the advantage of multiple uses and minimises numbers of spent
batteries going into landfills. The authors’ experience is that performance can vary
24 Invasive Animals CRC
considerably between brands (Figure 9). In Australia, the Eneloop batteries are good quality
and have been recommended by experts.
Alkaline
Alkaline batteries are readily available, and consequently, their use in camera traps is
widespread. They tend to be cheaper per battery than the other types above, but tend to
discharge quicker than NiMH and Lithium. They also suffer in extreme cold weather, losing up
to half their capacity in sub-zero conditions (R. Howe, personal communication, 2011).
Despite these apparent disadvantages, many people consider them to be a convenient option
that is ideal for short-term deployments.
Figure 9. The power discharge of Lithium, Alkaline and two brands (Ansmann and Tenergy) of NiMH
batteries mAH = milliamps/hour. (Data courtesy of TrailcamPro).
An introduction to camera trapping for wildlife surveys in Australia 25
Figure 10. Battery life of three types of AA batteries (n=8) in series in a Reconyx camera trap
(a: alkaline, b: lithium and c: NiMH batteries) (data courtesy of Trailcampro).
a. Alkaline Batteries
BattBatteriesBatteries
b. Lithium Batteries
c. NiMH Batteries
26 Invasive Animals CRC
4.12. External Batteries
Some users choose to connect their camera traps to an external battery and/or a solar panel
for extended use. However, only some camera traps afford this option, and it may only be
useful to you if battery life, rather than memory, is limiting. That is, if your memory card fills
up with images or video long before your batteries expire, the additional operating time is
not being utilised effectively. Furthermore, where theft or vandalism is an issue, external
batteries and/or solar panels may increase the likelihood of your camera trap being detected.
4.13. Camera care and storage
Most camera traps are reasonably robust, but you should not forget that inside they contain
electronic components. Moisture, therefore, can be a significant problem. If rain or moisture
humidity breach camera housings, then camera trap performance can be seriously affected
and be rendered temporarily or permanently useless. At sites with high humidity, however,
excluding moisture from the unit is almost impossible. Many researchers consequently use
desiccant in their camera traps. Desiccant comes in many different forms. Some desiccant
comes in single-use packets, whereas other types can be dried in a microwave or standard
oven for re-use. A problem faced by many researchers has been trying to find a suitable
location to place the desiccant, particularly in newer and smaller camera traps (Figure 11).
Figure 11. Finding a place to deploy moisture desiccant can be challenging. The best location in a
Reconyx camera where a canister can be used is in the corner of the housing (image: Paul Meek).
An introduction to camera trapping for wildlife surveys in Australia 27
Camera traps can also be damaged in transit, whether to or from the site of deployment.
Protecting the lens and external sensors is particularly important to maximise the longevity of
the equipment. Users come up with various solutions for this, but many buy commercial
storage cases with foam inserts to minimise the effects of dust, moisture and impacts.
Storage cases for transporting camera traps can be purchased to suit the number of cameras
in your kit. Dust and water proof seals are a priority, and foam inserts can be fitted to
provide impact suppression during transportation. Many brands are available in Australia. See
http://www.rei.com/product/634288/rubbermaid-action-packer-24-gallon and
http://pelican.com/case_category_single_lid.php?CaseGroup=Trunk&LidType=%.
28 Invasive Animals CRC
5. Camera settings for wildlife surveys
Camera trap models vary in their features and functions, which you must take into account to
suit the intended use. You can quickly exclude many camera traps in your decision process
based on differences between functionality and intended use. Nonetheless, it is unlikely to be
viable for users to buy many models to test for themselves, so it is advisable to discuss the
settings and various advantages and disadvantages with colleagues and/or researchers to help
fine-tune the settings and image data.
You can program the setting on your camera trap by either of the following ways:
manually programming through the settings menu on the device
installing software specific to the camera trap and programming through the SD card.
The latter is not always an option and will depend on the type of camera trap you have.
5.1. Time, date stamp and temperature recording options
Digital images are stored with time and date data (called EXIF files). Nonetheless, it is
typically up to you to ensure that they have correctly set these initial values at the time of
deployment. Using the time and date stamp data in association with photos is integral to
many analyses (see Section 7 for further advice). Be aware that daylight savings and moving
between time zones can have impacts on your data if you do not account for them. Some
camera traps will also record moon phase and/or temperature data with each image. The
temperatures recorded by many models, however, do not represent ambient temperature.
Meek et al (in press) found that camera traps deployed together at the same location
recorded substantially different temperatures. A camera in a shrub protected from the sun
recorded a temperature of 26 degrees Celsius, but one in direct sunlight recorded 30 degrees
Celsius.
5.2. Sensitivity
With many cameras, you can control how the camera trap responds to stimuli by changing the
sensitivity. The Reconyx Hyperfire 600, for instance, can be extremely sensitive and may
capture many images of moving vegetation with the ‘high sensitivity’, rather than ‘medium’
or ‘medium/low’, setting. Conversely, setting the sensitivity to high is important to maximise
the sensitivity of the PIR’s heat signature differential when ambient temperature approaches
the body temperature of animals. In the Reconyx PC850, adjusting the sensitivity to high and
shutter speed to fast will reduce flash illumination, which is ideal to survey a nocturnal small
mammal. In the Pixcontroller, sensitivity settings can be programmed to reduce false
positives by manual adjustment. Check the manual to see if sensitivity settings are available
and trial them under various conditions to understand how it affects your camera trap’s
performance relative to your needs.
An introduction to camera trapping for wildlife surveys in Australia 29
5.3. Trigger speed and delays
Trigger speed refers to the time between detection and capture of image or video. Some
camera traps have user selectable trigger speeds, but it is common for it to be fixed. The
nature of your camera trapping needs will govern the trigger speed you require. Situations
involving relatively fast moving targets tend to require faster trigger speeds, such as where
animals are moving quickly along a trail. When a target is being attracted to, and then
encouraged to stay within the detection zone, such as at a feeding station, slower trigger
speeds may be sufficient.
Equally dependent on your survey requirements is your need for the use of a delay period.
Where target species’ activity patterns are predictable, a delay period may be useful to
activate the camera only during periods of interest, thereby preserving battery life and,
potentially increasing independence of photographs.
5.4. Number of photos
A single photo can be sufficient to establish presence and identity and obtain demographic
information for a target animal. Nonetheless, taking several successive photos increases the
chance of obtaining the required information. This can be particularly useful for recording
animal behaviour, and as raised above, bursts of consecutive photos can sometimes be used
to simulate video.
The number of photos taken per event will depend on the model of camera trap. Some
camera traps allow for bursts of sequential photos whenever the subject is within the
detection zone, whereas others will only take one photo at a time with unavoidable delays
between successive triggers. When multiple triggers are likely (whether due to target animals
or not), setting a camera to multiple photos can quickly deplete the memory and/or
batteries. Research to quantify the relative advantages of different strategies, with respect
to number of images per trigger, is underway (Meek et al unpublished data).
5.5. Flash setting
Flash intensity of some camera traps can be adjusted for through overriding settings, such as
sensitivity and/or various night modes, but generally, manufacturers have not provided
opportunity to readily change intensity of illumination to suit the distance between the
device and the target. It is sometimes possible to add additional flash units or to decrease
flash intensity by covering some of the flash shield with opaque electrical tape, for instance.
Camera traps have rarely, if ever, been designed specifically to illuminate close subjects.
This is problematic and often results in white-out of close-by animals. The Pixcontroller
DigitalEye uses a Sony camera, which automatically sets the exposure settings instantaneously
and rarely overexposes the subjects at 1-2 m. You may lose battery life if you change a
sensitivity setting to heighten illumination intensity.
30 Invasive Animals CRC
5.6. Recovery time
Related closely to number of photos, the recovery time of camera traps is important when
using camera traps to survey wildlife. Recovery time is essentially the lag between successive
triggers (ie how soon the camera is ready to be triggered again by activity within its detection
zone after taking an image, or burst of images). Recovery time (eg instantaneous, within a
few seconds, or after nearly a minute or more) will have significant impacts on surveys that
require more or less continuous images. For example, if you are trying to capture images of a
family group, say moving in single-file along a trail, or studying a behavioural pattern or
activity, then it is easy to have a rapid recovery time to maximise the chances of
photographing each individual. Conversely, if your focus is presence-absence data, say at a
waterhole, a substantial delay between triggers may be acceptable.
An introduction to camera trapping for wildlife surveys in Australia 31
6. Field deployment of camera traps
How you deploy camera traps depends on the objective of the study, the camera trap model
and the nature of the local environment. It is not the intention of this document to provide
specific details on the diverse range of possible survey designs or their particular analysis
requirements. A number of authors have already attempted to provide insight and instruction
in these areas (eg Kays and Slauson 2008; Karanth et al 2011; O'Brien 2011; O'Connell and
Bailey 2011; Rowcliffe et al 2011; Rowcliffe et al 2012).
Cameras may be:
permanently located
returned to the same sites repeatedly
reallocated within a site for successive surveys
positioned temporarily for one-off investigations.
Some studies and locations may even require daily or nightly removal and replacement of
cameras to avoid theft or vandalism. This approach was adopted at Mutton-bird Island (Coffs
Harbour, NSW) where the risk of theft was deemed to be extraordinarily higher than
elsewhere due to proximity to a town and high human visitation (Zewe et al submitted).
Camera traps are often attached to trees and or posts. Using trees may be quicker, but posts
can allow precise, repeatable placement and further reduces the risk of damage to trees in
conservation areas.
As a general rule, if you want to be able to maximise the usefulness of the data you are
collecting, try to deploy cameras in the same way each time you survey your target species.
Some camera traps’ performance is a function of their height above the target and their
angle of incidence relative to the targets direction of travel (Ballard et al unpublished data).
Deployment consistency can avoid the issue of differential detection probabilities that will
significantly affect some survey designs.
Reading the manual thoroughly is essential to understand the functionality of your camera
and optimising survey outcomes. Before deploying your camera traps, ensure that you have
enough batteries and memory cards and that you have settled on a means of placing/affixing
the camera traps. For seamless data collection, it is ideal to have two memory cards and
twice as many batteries as necessary per camera trap. Set and check the settings on each of
your camera traps prior to deployment, including the date and time stamp. Although it may
sound unnecessary, it is worth taking GPS locations and making notes about the specific
deployment details for each camera trap (eg the type of tree or proximity to a local feature)
to help you recover the equipment and data. Recovery can be easy if you have only one or
two cameras or if cameras are spaced at regular intervals, but if you have many cameras, say
50 or more, that have been deployed randomly or haphazardly, the GPS locations will be
invaluable.
32 Invasive Animals CRC
6.1. Photographic principles
When designing camera trap surveys, remember that the fundamental tool being used is a
camera. As such, the basic principles of photography also apply. The camera should be stable
and positioned to account for its focal range. It should not face directly into the sun. Unlike a
hand-held white flash SLR camera where the user can compensate for the conditions, camera
traps rarely allow such flexibility. Consequently, you need to consider lighting throughout the
24-hour period, keeping in mind that shade can occur both during sunlight and moonlight.
Shading can affect shutter speed; in low light the shutter speed may be slow, leading to
blurring.
6.2. Camera trap height
The most suitable height to set camera traps will be determined by the target species, the
objective of the investigation and the camera’s functionality. Nelson and Scroggie (2009)
made recommendations about camera trap height for a range of species. As a rule of thumb ,
the height of the camera should be similar to the core mass of the animals you are attempting
to detect. In the case of small mammals, setting cameras <50 cm above ground level is the
standard. In other species the height may be up to 2 m. Specific recommendations on height
are in Section 8. Based on the protocols recommended by other organisations, camera traps
are usually placed in the height range of 20-50 cm, but this has been decided by trial and
error, not experimental data. Meek et al (unpublished data) compared camera trap data from
two height classes in detecting animals during carnivore surveys. One camera trap was set
about 90 cm above ground level and the other at 300 cm. There was a difference of
approximately 40% in both events recorded and detections of species between the low- and
high-set camera traps, and low sets were more successful.
6.3. Camera trap direction
In the southern hemisphere, facing camera traps to the south, south-east or south-west will
reduce the likelihood of the camera traps facing into the sun (Figure 12). Depending on the
sensitivity setting and the camera trap being used, false triggers can occur in the morning as
the sun rises and starts to warm sunspots and vegetation. This can also occur where the sun is
shining directly on the face of the camera. Avoiding this saves battery life.
An introduction to camera trapping for wildlife surveys in Australia 33
Figure 12. Observed pathway of the sun in the southern hemisphere (courtesy of Museum Victoria
copyright).
6.4. Centralising the detection zone
Ensuring that the camera PIR sensor is optimally positioned to detect targets can be time-
consuming. Many camera models have a walktest function, or similar, that allows you to
check the placement for purpose. To use the walktest, position the camera, switch it on and
select the walktest function. This will engage the PIR sensor to detect a passing heat
signature, but rather than taking an image, an LED illuminates to signify detection. When
available, we strongly recommend using this function during deployment of camera traps,
even if viewers (below) are used to perfect the position of the camera trap.
Viewers can be a useful tool for improving the precision of camera trap placement (Figure
13). These can be a cheap card reader with a view screen or a laptop computer. In either
case, you can review images from the camera trap, in situ, to refine the deployment prior to
the survey. These tools help alleviate some of the problems of camera placement and
detection zones. If you chose to buy a cheap camera, just take your SD card with images from
your camera and see if the camera will read your images. There are several brands available
on the internet: http://cuddeback.com/scouting_camera_products/cuddeview.html
http://www.spypoint.com/EN/trail-cameras/accessories/viewer.html
http://www.moultriefeeders.com/productdetail.aspx?id=mfh-vwr-11.
34 Invasive Animals CRC
Figure 13. The Cuddeviewer (or similar device) allows you to view images taken by cameras that use SD
cards and compact flash cards (image: Paul Meek).
Some camera traps, such as Leopold and Scoutguard, have a cable-connected programming
device that can also function as a hand-held reader, with live-viewing or picture-reading
capacity. These, too, can be used to assist in camera trap placement.
When you use trees to attach camera traps, the size and angle of the tree stem can affect
how the cameras are set. The use of wedges, or sticks, to achieve a preferred angle is often
necessary (Figure 14). Some camera traps, such as Moultrie, have a light beam that can be
used to aim the camera at an optimum site in the landscape. Alternatively, a laser pointer
can be bought from a stationary store and used in combination with the camera to work out a
rough estimation of where the camera trap is pointing. Nonetheless, some trial and error may
be necessary to determine the relationship between the detection zone and where the laser
is pointing.
An introduction to camera trapping for wildlife surveys in Australia 35
Figure 14. Attachment of cameras using wedges and sticks to aim the camera directly at the focal point
(image: Paul Meek).
6.5. Attachment to poles, trees or tripods
A plethora of devices are available for fixing camera traps to trees, posts or other surfaces.
The features of the camera (eg tripod mounts), location and objective of the study will
determine what you use. Many users mount camera traps to trees with straps (Figure 15).
Python locks and their equivalents are excellent for limiting theft, but even these cables can
be removed (Ballard and Fleming 2011) if the thieves are determined, or worse still prepared.
Most popular camera traps have commercially available theft-proof boxes and are widely
available.
Camera traps, such as Reconyx, Leopold, some Scoutguards and Uway, do have standard
tripod mounts that can be used where this is convenient and suitable (eg when risk of theft or
interference is low to nil). Otherwise, there are numerous camera trap attachments to help
set them in the field. Faunatech has a range of products (Figure 16). Moultrie has their Deluxe
Tree Mount. Reconyx has five designs. A range of options are available (Figures 17and 18) and
can be reviewed on trail camera websites. One of the cheaper brackets is the Outdoor
Camera Mounting Bracket that retails for about AUS $20 (Figure 18), although this has a
weight restriction (so check your camera trap weight).
36 Invasive Animals CRC
Figure 15. Trial of three camera trap models using tree mounting for small mammal investigations
(image: Paul Meek).
Figure 16. The Faunatech Rockpod is a sturdy and adaptable tripod-type device (image: Ross Meggs).
An introduction to camera trapping for wildlife surveys in Australia 37
Figure 17. Ezi-Aim screw (top) camera trap mounting device for trees and a tripod-type mounting
bracket design (bottom) by KORA (image: Paul Meek).
Figure 18. Outdoor camera mounting bracket and Reconyx steel-post fitting for attaching camera traps
to pickets and posts (image: Paul Meek).
38 Invasive Animals CRC
The use of Steel-strap bracing (available in rolls that can be cut to measure) is also cheap and
convenient. Additionally, it allows for accurate placement of the camera’s detection zone to
maximise coverage (Ballard and Meek unpublished data, Figure 19). The value of the pliable
steel strapping is one end can be secured into a tree, post, or a steel peg to create a solid
base and then the strapping can be twisted to direct the camera trap exactly where it is
required. With other camera traps, such as Scoutguards, the strap or bracket can be attached
to the housing using gaffer tape.
Where the placement of a camera is precisely predetermined by survey design, trees may not
be available in the right location to place cameras in areas such as beaches and deserts. In
this case, a steel post may be required. In other situations where the ground may be too
impenetrable to drive pegs, tripods may be necessary.
The number of cameras being deployed will also affect how camera traps are placed,
meaning that using special fittings may be too expensive. In this case the cost-effective
solution shown in Figure 19 may be a cheap and effective option. Security will also influence
what placement method is used. Despite all camera traps being sold with nylon straps of
various forms, these offer no deterrent to thieves. Nonetheless, there is no shortage of
options available. The main constraints are cost and the type of camera trap brackets
available for the model you are using.
Figure 19. This generic bracket can be attached to camera traps that have tripod fittings and screwed
into trees or fastened with bolts to posts. Similarly, lengths of steel strap bracing, cut to length , can be
used as a more flexible substitute. In either case, a short piece of threaded rod is used with wing nuts to
secure the device (image: Paul Meek).
An introduction to camera trapping for wildlife surveys in Australia 39
6.6. Spatial distribution of camera traps
Designing appropriate surveys can be complex and it is always advisable to consult a
competent biometrician to ensure that the data you obtain are relevant to your question and
can be analysed. How you design your camera trap arrays will be influenced by:
the purpose of your study
the species of interest
local environmental variables
the type and number of camera traps you have.
Deciding on the spatial distribution of camera traps (eg a linear transect vs a grid or some
other allocation) can be difficult. Kays et al (2009) recommend that where the objectives are
to document entire animal communities, a randomised design is imperative.
When determining the distance between camera traps, researchers have typically taken into
consideration the size of the home range of the target species. For some species, inter-
camera distances of as little as 25 m may be sufficient to record independent data (Kays et al
2009), but for others, hundreds of metres may be necessary.
6.7. Deployment time
A general rule for the duration of deployment of camera traps is the longer you leave camera
traps deployed, the better the dataset(Kays et al 2009). A common time frame for camera
trap deployment is two to four weeks (Kays et al 2009), although this can vary depending on
the species and habitat. In some cases, camera traps are set for less than two weeks (Kays et
al 2009; Meek 2010). Paull et al (2011) suggest that a minimum deployment time for camera
trapping studies on Australian small- and medium-sized mammals is 14 nights.
Nelson et al (2009) used camera traps to survey small mammals in Australia and detected
Smoky Mouse (Pseudomys fumeus) for the first time on the 18th night, more or less over a
period of 10 nights. The optimal deployment time, or asymptote, however, will depend on
your target species, meaning further research is required for each species. For instance, an
asymptote for long-nosed potoroo (Potorous tridactylus) is about 12 days, but there is no
reliable asymptote for long-nosed bandicoots (Perameles nasuta) (A. Claridge, personal
communication, 2010).
Table 5 is a summary of current camera trap asymptotes from the literature.
40 Invasive Animals CRC
Table 5. The number of nights required for several Australian species to reach an asymptote in
detection (H = horizontal placement; V = vertical placement).
Species
Author
Asymptote (survey
nights)
Long-nosed Potoroo
Claridge (personal
communication, 2010)
12
Southern Brown Bandicoot
De Bondi et al 2010
5
Southern Brown Bandicoot (H)
Southern Brown Bandicoot (V)
Long-nosed Potoroo (H)
Long-nosed Potoroo (V)
Smith and Coulson 2012
Smith and Coulson 2012
Smith and Coulson 2012
Smith and Coulson 2012
30
15
97
17
Smokey Mouse
Nelson et al 2010
10
6.8. Weather recording
Supplementary data, such as those describing local habitat and environmental conditions, are
often collected to aid in the interpretation of wildlife survey results and are useful for
camera trap surveys. For instance, data on ambient temperature at camera trap sites can be
particularly useful as most camera traps used for wildlife surveys are heat and motion
sensitive. Such data can be used to scrutinise likely temperature differentials between
targets and ambient conditions, providing insight into detection probabilities (see Section 4).
You can also collect local weather data throughout the survey period using field-based
weather stations or data loggers, such as i-buttons. These data can also be compared with
EXIF data stored by camera traps.
6.9. Active survey designs
Active surveys use a lure to attract target animals into the detection zone of the camera trap.
They rely on changing the behaviour of the target animal to increase detection probability.
Baits/Attractants
As in live trapping exercises, choice of bait for active camera trapping will depend on the
target species (see for example, Paull et al 2011).
It is often necessary to use some form of permeable container or cage to maintain many food
lures at the site of deployment. Tea infusers have been widely used for this purpose in live-
trapping exercises for small- and medium-sized mammals, which can be adapted for use in
camera trap surveys (Figure 20). The addition of a wire cage or cutlery draining rack (Nelson
2009) will prevent medium-sized mammals, such as possums, from removing the bait. PVC
vent cowls (Figure 21), for instance, can be used (Paull et al 2011) and are easier to carry in
the field (A. Claridge, personal communication, 2010).
Carcases may also be used in predator/scavenger surveys, particularly when trying to
establish presence or absence, or a minimum known-to-be-alive value, for a target species in
An introduction to camera trapping for wildlife surveys in Australia 41
a particular area (Figure 22). Such techniques hold promise for enumerating mammalian
carnivores but can be compromised by frequent visitations by abundant, non-target
scavengers, such as corvids and varanids (Ballard et al unpublished data).
Figure 20. An example of a method for maintaining food lures used in northern NSW. A tea infuser
containing bait is wired behind a cutlery drainer and suspended on a steel picket. The ruler can be fixed
to the picket to assist with image scaling. By suspending the tea strainer inside the cutlery drainer the
setup not only excludes small- and medium-sized mammals but ants too (image: Paul Meek).
42 Invasive Animals CRC
Figure 21. PVC Vent Cowl used for active surveys of medium-sized mammals on the south-east coast of
Australia (image: Andrew Claridge).
Figure 22. A wild dog photographed at the carcase of a dead horse on private land in north-east NSW
(image: Guy Ballard and Sam Doak).
An introduction to camera trapping for wildlife surveys in Australia 43
Camera trap placement
The target species, habitat and camera type (particularly the type of PIR sensor and
detection zone) will dictate the optimum placement for camera traps. In studies where
obtaining flank or neck shots are needed for use in identification, camera traps need to face
perpendicular to the path of the animal, or be set in such a way to photograph the relevant
body part (Figure 23). It is yet inconclusive whether horizontal or vertical is a better
placement position for most species, but in most cases, the cameras are set horizontal to the
ground. De Bondi et al (2010) and Smith and Coulson (2012) concluded that vertical
placement was better to detect Southern Brown Bandicoot (Isoodon obesulus) and Long-nosed
Potoroo. These authors recommend that camera trap users conduct pilot studies to determine
optimal placement for their species of interest.
Figure 23. Pine Marten (Martes martes) bait delivery system to improve the opportunity for capturing
images of the distinctive gula pattern used in image recognition software (Copyright: Erwin van
Maanen).
6.10. Passive survey designs
Unlike active surveys, passive surveys use no bait or attractants to lure the target into the
camera trap’s detection zone when it is critical to the analysis that animal behaviour is not
influenced. In Australia passive surveys are often used for carnivore studies to analyse indices
of activity or abundance.
44 Invasive Animals CRC
Camera trap placement
Most users deploy camera traps at a height equivalent to the core of the target species’ body.
In practice, this is often from 20-90 cm above the ground for animals ranging in size from
quolls (Dasyurus spp) up to feral pigs (Sus scrofa).
6.11. Animal responses to camera traps
Widespread concern exists about negative responses of wildlife to white and infrared flashes,
and there is some evidence to support this (Schipper 2007; Newbold and King 2009). Small
mammals and possums in Australia did not avoid sites where white flash camera sites were
used (Meel et al unpublished data). In contrast, Prasad and Sukumar (2010) observed a
reduction in fruit consumption by frugivorous ungulates at camera trap monitoring sites in
southern India. This problem was solved by locating the cameras above the eye line (ie
overhead). Studies in Canada found 40% of wolves (Canis lupus) showed an adverse response
to infrared cameras (Gibeau and McTavish 2009). Similar trials are being carried out in
Australia (Meek et al unpublished data) to determine whether mammalian carnivores respond
to various types of infrared camera traps.
A few camera trap models can have separate PIR sensors fitted so that they can be placed in
more subtle locations, but this is not commonplace. The critical issue is how the animal
behaves in response to the flash, whether this affects the data you are collecting and
whether the long-term behaviour of the animal changes in response to the initial short-term
behavioural response. Importantly, if the purpose of the investigation is merely to assess the
presence and absence of an animal, then the startling effect will likely not matter. If,
however, the survey relies on repeated visits to the site by the animal, changes in behaviour
that reduce this probability may introduce serious bias.
6.12. Camera trap emissions: sounds and sights
Animals are often photographed looking at the camera trap. The reasons for animals’
response to camera traps are unknown. Meek et al (unpublished data) hypothesised that
infrared light or other light emissions might be visible to some species and these or audible
outputs from the camera trap may be the cause. Tests to determine the audio output of IR
camera traps produced evidence that all camera traps emitted audible noise in the range 12.5
Hz 20 KHz (Meek et al unpublished data). These noises are well within the detectable range
for feral cats, for example, so their detection by a cat would then be dependent on loudness
and distance of the cat from the camera. Similarly, measurements of the infrared spectrum
of a range of commonly used camera traps suggest that target animals in Australia, such as
wild dogs, red foxes and feral cats, are likely to see camera traps when they are triggered
(Meek et al unpublished data).
6.13. Camera trap security
Theft and vandalism are recognised as limiting factors in the use of camera traps worldwide
(Kays and Slouson 2008). They are particularly relevant issues for surveys where cameras need
An introduction to camera trapping for wildlife surveys in Australia 45
to be set in high visitation or high visibility areas for humans (eg beside a track). There are a
range of security options available when placing camera traps, such as:
deterring people from accessing an area
choosing sites with low human presence
camouflaging and securing the devices with locks and other devices.
Python cables and their equivalents can be used to fix camera traps to trees or pickets,
although an organised thief can cut these cables. Some cameras have purpose-built security
boxes (Figure 24) that are thief proof, although they are cumbersome and time-consuming to
erect. They can be purchased from all the Australian camera trap dealers, and specific
information can be obtained from http://www.camlockbox.com. Many units have
camouflaged surfaces, but this is typically based on vegetation from the northern
hemisphere. Consequently, users sometimes glue foliage and/or barking to the outside of the
cameras, or even recessing cameras into trees.
Some recent models have a camera lens separate to the processing hardware which can be
buried underground (eg Bullet cam), or can be located up to 20 m away using a wireless
system (eg Pixcontroller). As a last resort, some users place signs on or near the camera traps
describing their purpose and asking they be left intact so that data are not stolen (Meek
2012).
Figure 24. A security casing for the Reconyx Hypefire (photo courtesy of Reconyx).
46 Invasive Animals CRC
At long-term study sites, there may be little option other than installing permanent security
housings (Meek et al unpublished data, Figure 25). These structures are obvious, but regular
users of the tracks may become accustomed to their presence and unsure when cameras are
inserted. These security posts use standard security boxes welded on steel posts with some
modifications. The security box is faced slightly downwards at the front (10 degrees) with the
standard front lock having been removed. A modified flat steel key has been manufactured to
go through to the rear of the box, and a lock shield was constructed at the rear to reduce bolt
cutters access to the locking mechanism.
Alternatively, a covert camera trap can be set up at an entry point to the survey area to
monitor people activity. If cameras are destroyed or stolen, the registration numbers of all
vehicles entering the area will be captured. Using the new Xtern Farmcam, Buckeye Orion
and Scoutguard SG580M, you can set the wireless motion sensor away from the covert
camera, and where applicable, send images and detection messages to a mobile phone from
the camera trap using a modem. In Australia a licence is required for these devices, and at
present only one camera trap is legally approved for this use, the Buckeye Cam Orion (R.
Meggs, personal communication, 2012).
Figure 25. Schematic diagram of the camera trap security post. a) permanent set up b) removable
ground level plate set up (Meek et al in press).
An introduction to camera trapping for wildlife surveys in Australia 47
GPS tracking systems can also be used to locate stolen cameras (Bancroft 2010). At least one
camera trap, the Pixcontroller Raptor, has a GPS telecommunications system designed to
detect when a camera is moving and sends an SMS of the location to the user. As discussed in
Section 2, the use of camera traps for gathering photos of people is subject to ‘Privacy’
legislation (Appendix 1). The use of camera traps for covert activities in this publication
should not be interpreted as endorsement or a recommendation by the authors.
The use of dummy units placed in obvious locations can also be a useful option, distracting
thieves from noticing another camera adjacent to the dummy.
Another option is to close access to the study site or sections of it for the duration of the
monitoring, although this may only exclude some people from an area. Alternatively, you can
choose times of year when activity by people is less likely to occur. For instance, in
Switzerland they set traps in winter because fewer people are in the forest to encounter
traps.
48 Invasive Animals CRC
7. Data management and analysis
Storing camera trap images is an enormous management issue, but there has been no data-
storage standard adopted in Australia. Many users adopt a simple folder system, but for long-
term studies, where large volumes of images are being stored, implementing an appropriate
database system, or similar, is vital.
The steps in camera trap data management are:
(a) Planning
(b) Collecting
(c) Data cleansing
(d) Coding
(e) Storing
(f) Backing up
(g) Accessing
(h) Analysing
(i) Reporting
(a) Planning
Prior to a survey it is essential to consider your survey design and analysis so that you gather
the right data and can analyse the results. For instance, your data coding scheme determines
how you store your data.
(b) Collecting
In addition to camera trap images and/or video, information about the survey, site and
deployment should be recorded and maintained. A generic datasheet for this purpose is
provided in Appendix 3. An electronic version and corresponding Microsoft Access database
can be found at www.feral.org.au.
The use of a chalk board or white board to record the site details can later be used to
correlate photos with a location (Figure 26).
An introduction to camera trapping for wildlife surveys in Australia 49
Figure 26. Dry-erase or chalk boards allow a photo record to be taken at each site with specific site
location details to ensure that images correspond to the correct site (image: Paul Meek).
(c) Data cleansing
When the survey has been completed and it is time to start analysing images, first you need
to remove images taken during set-up and retrieval. It is important to record the total
number of images taken, but when it comes to analysis, these images are superfluous and
should be removed. You can use the site datasheet provided in Appendix 3 to record this
information and, in turn, enter it into the camera trap database.
The images chosen to be stored will depend on the investigation and the objectives or
hypotheses being tested. These choices will not be discussed in this manual and will require
expert advice to ensure accurate assumptions and decisions without compromising the
investigation.
(d) Coding
Coding allows you to quickly identify and sort camera trap images, either manually or using
software. One coding strategy is to record location, time, date, site and camera code for
each image, as per the following:
50 Invasive Animals CRC
(a) Transect set
Use CT_Ca_T1_C1 where
CT= Camera Trap survey type
Ca= Code for the location, in this case Carrai
T1= Transect number, in this case Transect 1
C1= Camera site, in this case camera Site 1
(b) Grid Set
Use CT_Ca_G1_Tr4 where
CT= Camera Trap survey type
Ca= Code for the location, in this case Carrai
G1= Grid number, in this case Grid 1
Tr4= Camera site, in this case Trap 4
(c) Point set
Use CT_Ca_S1_C1 where
CT= Camera Trap survey type
Ca= Code for the location, in this case Carrai
S1= Site code, in this case Site 1
C1= Camera site, in this case Camera site 1
Although these conventions may be appropriate in many situations, there will always be
exceptions to the rule. As long as the core data are noted, the format is not as critical.
(e),(f) & (g) Storing, backing-up and accessing camera trap data
The importance of setting up a back-up system to save image data cannot be overemphasised.
Numerous tools can be used to review data. Ideally, a dual-screen computer system makes
data analysis and data entry easier. Programs, such as DeskTeam, MapView and Camera base
1.5.1, provide viewing and coding options on screen. If you have numerous videos to watch,
devices such as Xtreamer or VLC Media can be used for processing videos on televisions.
A number of Digital Asset Management (DAM) systems are available, specifically designed for
managing images. Some are freeware and others are licensed. Programs include IDImager Pro,
Imatch, Photomechanic, Geosetter, Auto Photo Organiser 2.4.739, Lightroom and ACDSee Pro
2. A useful review of packages is also available online.
An introduction to camera trapping for wildlife surveys in Australia 51
The most common practice for data storage, however, is designing your own Microsoft Excel
or Access database so that it is specific to your needs.
To cater for the specific needs of non-standard camera trap research investigations (eg
occupancy or mark recapture), data can be entered into one of the databases descried below
and then extracted into an Excel spreadsheet to carry out further coding of data. Use of this
software can save an extraordinary amount of time entering metadata, although there are
still some time needed to set up the databases before uploading images and processing the
results. None of these programs are a panacea for problems of image management and
analysis, but they do have various positive attributes and constraints.
MapView
MapView is developed by Reconyx for their professional series (PC) of camera traps and is
only available when you purchase a PC model. The program has some useful functionality for
basic data storage and mapping. The software accesses your data, extracts the metadata
(EXIF), including photo quality data (eg saturation), and sends it to a database. MapView also
has a fast viewing function that allows you to view images at a customised speed or manually.
It uses Google Maps or a map of your uploading to allow camera trap sites to be marked and
then relates future data to those points. Mapview also allows tagging and basic coding of your
data. The data can also be exported in a CSV file and entered into your database of choice.
BuckView
This baseline Reconyx software program is available for use with the non-professional series
cameras and provides a basic storage system for images. It has similar functionality as
MapView with geo-coding and Google map access as well as image viewing options.
Scouting Assistant
The Chasing Game website has produced a basic software package for managing images,
which also has the ability to stitch together images to create near-video clips with the options
to assign music and text onto a video clip.
Plot Stalker
Moultrie has developed a software program for their plot-watching and time-lapse cameras
called Plot Stalker. This program allows a series of images to be stitched together to create a
video of the plot or feeder.
Cuddevision
The camera trap company Cuddeback provides a free software program called Cuddevision,
which is designed for basic manipulation of image data. This program, however, is more
suited to hunting data than scientific storage of images.
52 Invasive Animals CRC
DeskTeam
TEAM Network is focused on presence/absence surveys of large carnivores in tropical
rainforests around the world. For consistency in data collection between all of their projects
spanning 17 countries, they have developed their own internal database called DeskTeam
(TEAM Network 2011). This JAVA-based program allows SD card data to be uploaded and the
EXIF data are automatically extracted and stored with the images. You can then access these
data along with some identification field and code the data according to the world taxonomy
list of mammal and bird species. It will also allow you to code groups of photos so that a
series of images from the same event of the same species can be automatically coded.
Uploading data from multiple sites over the internet is complicated, and the data upload can
be slow. TEAM Network is now looking at using BitTorrent sites to upload the data. This
database is only available to partner organisations at this time.
Wildlife Picture Index
The Wildlife Picture Index (WPI) is a program developed by the Wildlife Conservation Society
(O'Brien et al 2010) in collaboration with many international conservation and research
organisations to measure changes in biodiversity in tropical rainforest hotspots around the
world. The program uses the state variable of occupancy of mammals and birds of 1 kg and
above as the measure of health. It is based on camera trapping programs across many nations,
and surveys are carried out annually so that early warnings of a declining WPI can be detected
before the communities suffer irreparable decline. Although O’Brien et al (2010) did not
clearly indicate how the camera trap data was provided, the strong links with TEAM Network
suggests that the data are captured, stored and coded using the TEAM Network software.
Camera Base 1.5.1
This software program was developed by Mathias Tobler specifically for camera trapping data
storage and analysis. It is based on his long-term experience in using camera traps for
scientific research. The current version is Camera Base 1.5.1 although there have been some
minor changes to enable access to data on networks. The manual is only version 1.4 (2010),
but this is adequate for the operation of the program. The program requires Microsoft Access
although there is a Runtime program to allow it to run without Microsoft Access.
Camera Base 1.5.1 allows camera trap data to be uploaded directly into the program where it
automatically renames the data and extracts the metadata from each image and saves it into
a file with the images. A valuable function of this program is the ability to enter data from
two cameras set at the same location so that it can compare and calibrate detection between
cameras.
The program is still being fine-tuned and modified to suit a range of needs. When attempting
to share access, organisations encounter problems as each database is individual to each
computer. Scripts are available from Mathias to resolve this problem. Similarly, Microsoft
Access can cache and jam if moving between stored images too quickly - it is important to
install the ImageRegistryFix.reg file to avoid this fault.
An introduction to camera trapping for wildlife surveys in Australia 53
The program will allow you to analyse the data to obtain outputs, such as image summary
data and reports, activity analysis, capture-recapture analysis as well as exporting data into
formats for further analysis in Mark, Presence, Estimates and other programs.
Photospread
Photospread is a program developed in the USA for organising photos, primarily camera trap
photos (Kandel et al 2008). This software allows you to load and manage photos and tag
information as well as reorganising the images or sets of images according to your needs. The
software also allows manipulation and coding of your data but does require some fundamental
understanding of programming and basic scripts in Microsoft Excel. A demonstration is found
at http://www.youtube.com/watch?v=rf7rA-roBlM.
Jim Sanderson’s program
This unnamed software has been developed by Jim Sanderson to code, store, manage and
analyse camera trap images. This Disk Operating System (DOS) program uses ReNAMER to
manage file names. Using this software requires some fundamental knowledge of DOS and
programming to set it up properly. The program is open source and can be downloaded from
http://www.smallcats.org/CTA-executables.html.
(h) Analysing data
Analysis of camera trap data depends on the hypothesis (ie the survey design and nature of
the data collected). There is a wealth of papers describing the statistical methods and
analysis options for camera trap data (Rowcliffe et al 2008; Kays et al 2009; O'Brien 2011;
O'Connell and Bailey 2011). We reaffirm the need for camera trap users to consult with a
biometrician to make appropriate decisions in this regard.
Understanding how camera traps work is critical to sound scientific investigations and the
interpretation of the data. The following factors can all affect the data you collect:
detection zones
how each model works
trigger speed
delay periods
the number of images per trigger.
The area of detection can be determined using basic trigonometry if you know the focal
length and chord:
½C x FL, where C = chord and FL = focal length.
In camera trapping this can only occur where the chord is flat and does not extend past a
point, often encountered if the camera is facing down to the ground or is against a wall or a
tree. In situations where the detection area extends to the periphery of the sensor and where