CAMERA-TRAPPING PAGE 1 CAMERA-TRAPPING PAGE 2 Oliver R. Wearn & Paul Glover-Kapfer. 2017. WWF Conservation Technology Series 1(1). WWF-UK, Woking, United Kingdom. The authors are indebted to Carolina Soto Navarro for creating figures in this report, and to all of the people (credited in the figure captions) who kindly agreed to share their images with us. We also thank Jorge Ahumada, Mark Bowler, Paul Meek, Marcus Rowcliffe and Andy Whitworth for reviewing sections of the report. Funding for this report was provided by WWF-UK. WWF is one of the world’s largest and most experienced independent conservation organizations, with over 5 million supporters and a global network active in more than 100 countries. WWF’s mission is to stop the degradation of the planets natural environment and to build a future in which humans live in harmony with nature by conserving the worlds biological diversity, ensuring that the use of renewable natural resources is sustainable, and promoting the reduction of pollution and wasteful consumption. Cover Image: Thermal image of Panthera tigris © Judith Stryker CAMERA-TRAPPING PAGE 3 CAMERA-TRAPPING FAQ What is a camera trap and how is it different to a time-lapse camera? The modern digital camera trap is simply a digital compact camera sensor wired up to a passive infrared sensor which is able to “see” the infrared radiation given out by warm-blooded animals. However, any camera which is triggered by an animal to take pictures can be classed as a camera trap. This could include cameras triggered using any of a whole range of different methods, such as trip-wires, pull-wires, pressure plates, lasers or microwave sensors. A camera which is triggered remotely by a human is not a camera trap, and neither technically is a camera which is programmed to take pictures at set intervals, i.e. a time-lapse camera. The term “remote camera” is sometimes used to include this broader class of cameras, which are triggered in the absence of a human operator (but not necessarily by an animal). Other names for camera traps (mostly used in the hunting market) include game cameras, scouting cameras, or trail cameras. How expensive are camera traps? Off-the-shelf camera traps range in price from ~$50 to more than $1000 and, as ever in electronics, you get what you pay for. Typical mid-range camera traps suitable for robust scientific monitoring cost $300-500, with more expensive units typically having better detection circuitry, increased reliability, and more customisable settings. High-end camera traps can cost $500-1000, and may have video modes with a fast trigger speed (i.e. a short delay between sensing an animal and starting the recording) or be able to send images over mobile phone or wireless networks. Custom DSLR camera traps, which are necessary to take the highest quality images possible, can easily cost a few thousand dollars if bought new, and are a significant time investment to build and test. What animal groups can camera traps be used for? Almost all commercial camera traps sense animals using a passive infrared sensor. This sensor looks for sudden changes in the surface temperature of the environment in front of it, which could indicate the presence of an animal. These camera traps are best suited to anything that has a heat signature (i.e. body size and surface temperature) similar to a deer. This is because technological developments in the commercial camera trap market are still largely driven by the North American hunting market. Luckily, lots of animals have a heat signature sufficiently similar to a deer that the modern camera trap is useful for sampling a wide range of medium- to large-sized mammals and birds. The lower body size limit for camera traps used to be around 1 kg, but most mid- and high-end passive infrared sensors can now detect animals as small as 100 g, provided they are within 2 m. Beyond mammals and birds, passive infrared sensors have also been used to successfully detect reptiles (including small skinks, snakes, varanid lizards, and crocodiles). However, special approaches will usually be necessary in this case, and detections are unlikely to be as reliable as for mammals and birds. Other sensor types, such as active infrared sensors and pressure plates are alternative options when passive infrared sensors fail, albeit at greater cost. In addition, new software-based methods, such as pixel change detection, could expand camera-trapping into the aquatic realm in the near future, to monitor fish and marine mammals. CAMERA-TRAPPING PAGE 4 Can camera traps be used for monitoring animal abundance – will I be able to tell individual animals apart? Camera traps can be used to monitor a host of different ecosystem variables, such as the abundance, diversity and distribution of animals. Abundance monitoring is particularly effective with camera traps and has also been shown to be cost-efficient relative to rival methods (such as line-transects or live-trapping) for longer-term projects. It is often now the top choice of wildlife biologists, even for monitoring the rarest mammal and bird species in a community. A number of approaches can be used to monitor abundance. Most simply, the trapping rate (number of photos per unit of sampling effort) can be used an index of relative abundance, to compare trends across space or time, albeit with some important caveats. More robust methods include mark-recapture modelling (for species which can be individually-identified in images), and random encounter modelling (for species which cannot be individually-identified). In order to be able to indentify individual animals in images, they need to have unique markings, such as the stripes of a tiger or a zebra. You also need to use a camera trap which can take clear photos of the animal markings (in most cases, you will need a camera trap with a white flash, not an infrared flash). The random encounter model was originally formulated specifically with camera- trapping in mind, and requires some additional knowledge of the system, including the movement speed of the focal species and the detection zone characteristics of the camera. This requires additional fieldwork to make these measurements in the field. However, new methods are currently being trialled in order to be able to extract these from the camera trap images themselves. How many camera traps should I buy? Most often, the answer will be “as many as you can possibly get your hands on!”. The number of camera traps you have will determine the amount of data you can collect, and therefore which statistical methods it will be possible to apply. In practice, a better question to ask is “what is the minimum number of camera traps I need to achieve my objects?” This can be estimated approximately using information on your ideal survey design, how long it will take to set up camera traps (which will depend on how accessible your study sites are), and how quickly you need to complete the survey (e.g. to meet certain model assumptions, or due to the availability of resources). If you can afford (or can borrow) your estimated minimum number of camera traps, then your study is more likely to be a wise investment of scarce conservation resources. If you cannot meet the minimum required number of camera traps, then you should stop and seriously consider if you need to re-evaluate your study’s objectives. Even though there may be momentum building in your project, you should strongly avoid the temptation to just proceed regardless. CAMERA-TRAPPING PAGE 5 What camera trap model should I buy? Commercial camera traps are all broadly similar in basic form – being a compact digital camera, triggered by a passive infrared sensor – but they do nonetheless vary in a large number of important ways. Principle among these, are the characteristics of the detection circuitry (how sensitive the passive infrared sensor is, and how quickly the camera can take a picture after sensing an animal), whether they have an infrared or white flash, and whether they are able to send data remotely or not. The camera trap model most suitable for you will depend on what you intend to use it for. For example, if you intend to do random encounter modelling, you will need a camera with a fast trigger speed and, ideally, an infrared flash. If you intend to carry out a mark-recapture survey you will in- stead most likely need a camera with a Xenon white flash. Ultimately, given the vast num- ber of models on the market today, and the rapid pace at which they are updated by manufacturers, it is difficult to make a single recommendation for which camera trap model to buy. A sound knowledge of how camera traps work, and the key ways in which camera traps differ, will guide your purchasing decisions. Can camera traps record video? Not all commercial camera traps can record video, but an increasingly large number do. In most of the camera traps that record video, you simply choose which mode you want to use. Some models also offer a hybrid mode, which takes a single picture and then starts a video straight after. Note that video modes are slower than image modes, and you may therefore miss some animal detections altogether. They should therefore be used with caution in robust scientific monitoring. Video files also take up much more hard disk space than images, and are more difficult and laborious to process. Most camera trap software programs do not currently support video. A compromise between image and video is to use a camera trap capable of shooting in a “near-video” mode, in which rapid sequences of images (< 1 second apart) are taken.