Feral cats are nocturnal, elusive creatures and can cover a lot of ground. They can range up to 6 kilometres, making monitoring difficult. So how do you go about detecting them?
Camera traps are a useful, non-invasive way to determine whether you’ve got feral cats in your reserve. But most monitoring budgets don’t allow a comprehensive dotting of cameras across the landscape. Researchers Margaret Nichols, James Ross, Alistair S. Glen and Adrian M. Paterson from Lincoln University and Landcare Research recently compared systematic placement of cameras (in a grid) with strategically placed cameras to see whether choosing your placement spot increased detection rates and if so, what the best spots to choose were. The results of their investigation have been published in the research journal ‘Animals’.
“Carnivores often occur at low densities, have cryptic behaviour and require sophisticated monitoring and statistical modelling techniques to circumvent issues of low detection. As a result, camera traps have become an increasingly popular tool for providing population estimates for a variety of carnivore species, including feral cats (Felis catus). However, there is often wide variation in the deployment of camera traps within a landscape (i.e., number of sampling units used at each site and their distance from each other). For example, camera traps may be placed in a variety of ways, such as with a horizontal or vertical orientation; baited or unbaited; non-biased or biased allocation across a landscape, as in systematic grids/transects; or deliberately placed near likely target species ‘hot spots’, such as trails, roads, and water features.”
This study took place at two sites in Hawke’s Bay.
“Site 1, Toronui Station is a 1600-ha pastoral property with a mix of open farmland and native forest. Site 2 lies within a 26,000-ha portion of the Cape to City ecological restoration area. The habitat is similar to that in Site 1, with a mixture of native forest, open farmland, and some semi-urban habitat. Different camera trap models were used at the two sites, as these sites were originally intended for use in separate studies.”
Depending on resource availability and density, cats can have variable home-ranges that may overlap.
“Feral cat home ranges in Hawke’s Bay farmland are estimated at approximately 1.9 square kilometres (males) and 0.9 square kilometres (females), with a density of 3–6 per kilometre. However, in other habitats around New Zealand such as steep forest terrain in the southern North Island, feral cats were found to have linear home ranges of up to 6.34 km (males), 3.83 km (females), and smaller home ranges for females with kittens (0.84–2.0 km). Cats may prefer a variety of habitats, but most often those that include water sources and a mix of forest cover (both exotic and native).”
The researchers aimed to compare detection probabilities and mean abundance estimates for cameras using the two different deployment strategies (systematic vs strategic) at the two similar pastoral sites. Cameras at Site 1 were deployed at pre-determined GPS points on a 500-m grid, and at Site 2 they were strategically placed with a bias towards forest and forest margin habitat. Some cameras were also deployed in open farmland habitat and mixed scrub.
“In June 2014, we placed 40 camera trap in Site 1 for 21 days. Cameras were placed systematically on pre-determined grid points, across a 7 square kilometre grid with approximately 500 metre spacing between individual cameras. Due to the unbiased deployment method used in Site 1, several cameras were deployed in different habitat types, such as forest (both exotic and indigenous), forest margin (any edge between a forest and another habitat), mixed scrub (scrub, rocky areas, or a combination of the above), and open farmland (exposed farmland/paddocks).”
Stock damage to cameras and a large number of non-target images were issues encountered at Site 1.
“In November 2015, we placed 60 camera traps in Site 2 for 21 days. Cameras were placed spatially independent of one another (≥ 2-km) according to literature on mean cat home range size. Cameras were deployed with a bias towards the forest and forest margin habitat (assumed to be high cat activity areas), however, as well as in mixed scrub and open farmland habitats. Paddocks with large numbers of livestock, including red deer, sheep and cattle were avoided, to protect the cameras from damage and to reduce the number of non-target images, that we had previously experienced when deploying cameras at Site 1.”
Camera traps at both sites were programmed to capture images in bursts of three with the minimum possible delay between triggers for each camera type.
“Each camera’s field of view was positioned horizontally, parallel with the ground (10 cm from ground to the base of the camera set on brackets screwed into trees or wooden stakes, and facing south to reduce false triggers from moving light). If necessary, vegetation was cleared from the camera’s field of view, to reduce false triggers from vegetation moving. A perforated vial containing ferret odor (towels impregnated with the scent of a male ferret) was placed 1.5 m in front of all cameras as a scent lure and secured with a tent peg to avoid removal by animals.”
Four cameras were damaged by livestock or user error.
“At Site 1, 39 cameras remained operative, for a total of 819 trap nights. At Site 2, 57 cameras remained operative for a total of 1197 trap nights. Camera traps at Site 1 captured a total of 61,416 images, including 2687 non-target wild species (birds, ship rats, mice, hare, possum, hedgehogs, mustelids, pigs, and goats), 36,143 false triggers, and 19,338 livestock. Camera traps at Site 2 captured a total of 87,709 images, including 20,657 of non-target wild species, 56,664 false triggers, and 7271 of livestock.”
Sites 1 and 2 had similar detection probabilities for feral cats despite variable numbers of camera traps and different deployment strategies (8.6 % Site 1, and 8.3 % Site 2).
“Estimates for cat abundance at both sites were highest in forest margin and forest habitats for cameras in both deployment arrays. Our study supports the concept that camera traps placed in the ecotones of different habitats may maximize detections of cryptic target species. Other studies also suggest that cats prefer a combination of forest cover and open habitats for hunting.
Based on their results, the researchers recommend that for future studies and operations, particularly where cats are an apex predator, that cameras be placed strategically in the margins between forest and other habitat types whenever possible to increase detections. Further research is also needed on ways to increase detections, including ways to use multiple cameras.
“Deploying cameras in pairs or in clusters as per could increase detections. While single cameras can be strategically placed in areas that increase their chances of detection, multiple cameras combined with a biased placement may further increase numbers of detections. Further research into different lures such as sound lures, etc., and other social lures may also increase detections of feral cats if they are present.”
This research is published in the journal Animals.
