Camera traps for monitoring cat, stoat and hedgehog populations
When stoats ran quickly, camera traps often captured only their hind quarters. Image - Al Glen
Photographs taken with an infrared flash were not easy to identify. Image - Al Glen
Colour photographs taken with a white flash more easily identified pest animals than those taken with an infrared flash. Image - Al Glen
Feral cats, stoats and hedgehogs are controlled in New Zealand to reduce their impacts on native fauna. However, practical, affordable techniques are needed to determine whether control is effective in reducing their abundance. Such techniques will be a critical part of efforts to achieve Predator-Free New Zealand. Camera traps (trail cameras) are useful for monitoring wildlife, but until recently have only been used to monitor relatively large animals. Their use with small animals presents some challenges. First, small animals are less likely to trigger the infrared sensors used by most camera traps. Second, identifying small animals from pictures requires that they are photographed at close range. This means the camera must focus on a small area and fast-moving animals may cross the field of view before a photograph is taken. Finally, small species may be difficult to identify from photographs. Using captive animals, Al Glen and colleagues conducted a series of trials to determine the optimal specifications for a low-cost camera trap to monitor cats, stoats and hedgehogs. The factors tested were:
- Trigger speed (the time taken from when an animal is detected by the sensor until it is photographed – in these trials, 0.2 to 2.1 seconds)
- Type of sensor (infrared vs microwave)
- Type of flash (white vs infrared)
- Type of image recorded (still photograph vs video clip)
Camera traps were set in observation pens, where the behaviour of captive animals was constantly monitored by video cameras to determine how often animals encountered a camera trap, how they behaved, and whether the camera detected the animal and produced an identifiable photograph.
The continuous video footage revealed that stoats frequently ran around the perimeter of the pen at high speed; seemingly as a consequence of their captivity. None of the trigger speeds tested was fast enough to photograph stoats behaving in that way. However, even the slowest trigger speed used (2.1 seconds) successfully photographed 90% of stoats that were not running. Trigger speed had little influence on success rates of photographs for cats and hedgehogs, which moved more slowly than stoats.
Infrared sensors performed consistently better than the microwave sensor, which was often triggered by wind or rain, and occasionally by animals walking behind the camera. In contrast, infrared sensors reliably detected all three species and were rarely triggered by weather.
Each type of flash had advantages and disadvantages. Photographs taken with an infrared flash are black and white, and can appear blurred or grainy. This sometimes made it difficult to identify animals (e.g. if part of them was outside the camera’s field of view; see Photo). Cameras with a white flash took clear, colour photographs, allowing animals to be more easily identified. However, a white flash is conspicuous; it may frighten some animals, causing them to avoid camera traps, and potentially increases the risk of these cameras being stolen or vandalised.
Camera traps that recorded video footage had similar success rates to those that took still photographs, but required much more analysis time and computer memory. A 30-second video clip required 10.4 megabytes of memory, whereas still images required between 250 and 750 kilobytes. Video footage also took much longer to upload and view.
Cats and stoats frequently reacted to camera traps but hedgehogs did so rarely. Three of the six cats tested appeared to be frightened by cameras with a white flash, while one cat also reacted to a camera with an infrared flash. In contrast, stoats frequently showed curiosity towards camera traps, as did one of the six hedgehogs. These behaviours led Al to wonder whether the animals could see the infrared flash, or whether they reacted to some sound made by the cameras. The team therefore tested five commercially available camera-trap models marketed as having an invisible flash. In a fully darkened room, the flash of all cameras was faintly visible to human observers. Using ultrasonic recording equipment, they also found that cameras emitted sound at various frequencies, some of which are inaudible to humans but well within the hearing range of cats.
Camera traps show great promise as an inexpensive method for monitoring the outcomes of pest control. However, most commercially available models are not designed to photograph small animals at very close range. A camera trap for small animals should not emit visible light or audible sound. Field trials are being planned to confirm the effectiveness of camera traps for monitoring cats, stoats and hedgehogs.
This work was funded by the Department of Conservation.