The importance of scent trails

Authors: Grant Ryan, James Ross, Elaine Murphy and Merel Jansen

As highlighted in earlier blog posts, the Cacophony Thermal camera is much more sensitive for the detection of predators than the next best tool. It is able to see 3 to 50 times more activity, allows us to make observations that have never been possible before.

As part of our trials around possum detection efficacy with OSPRI we have had some of our thermal cameras and also standard trail cameras, pointing at chew card, a widely used way of detecting possums. The table below shows the results from the first 11 days of this experiment. Both our thermal camera and trail camera were positioned at a close distance (~1.5m) to the chew card. We expected similar detection rates on both cameras but even in this favourable scenario, the trail camera missed about half the predator interactions.

Experimental setup


DateTimePredatorHeat CameraTrail cameraChew card
18/04/201904:39:29rat1 1
18/04/201905:32:57rat1 1
19/04/201900:26:54rat1 1
19/04/201905:34:46rat1 1
19/04/201922:43:29rat1 1
21/04/201901:59:54rat1 1
21/04/201904:50:37rat1 1
21/04/201923:10:59rat1 1
22/04/201923:51:02rat1 1
23/04/201900:15:06possum1 1
26/04/201907:00:09rabbit1 1

When observing the footage generated by the Cacophony Project thermal camera during this experiment, it became obvious that once a few predators had found the chew card, many others do as well too. It appears that the first animals create a scent trail towards the chew card which are then followed by others.

This behaviour is clearly highlighted when we look at the thermal camera video footage in detail. The video below is sped up 10 times to illustrate what is happening more quickly. Predators mostly walk straight past the chew card until one discovers it - from then on most animals are drawn towards the chew card.

The trail camera missed most of the visits that did not involve the chew card so without the additional information from the thermal camera one might assume that chew cards and trail cameras detect all predators. It will be interesting to repeat the experiment with the thermal camera set back further to see how many other predators walk by before the chew card is discovered.

The trail camera was set to take photos rather than videos so it also missed some of the animal behaviour. It is very obvious from the video that the predators are following trails of previous animals but this is not so clear from a set of still photos.

It is widely known that animals follow scent trails. There have been many studies of social facilitation of feeding in rats. Norway rats follow trails left by other rats to find food (Galef & Buckley 1996). Naive rats will choose an unfamiliar food after observing another rat feeding on that food (Posadas- Andrews & Roper 1983; Galef et al. 1984; Strupp & Levitsky 1984; Galef & Whiskin 2000, 2001). Mice also learn food preferences from observing other mice feeding (Valsecchi et al. 1996). The new thing that has been observed here is how different predators follow each other's scent trails. This is particularly promising finding for the multi-species, multi-catch trap that The Cacophony Project is developing.

An additional benefit of cameras compared to a chew card is that they don't get eaten and taken away by the animals being monitored! Cameras provide a better idea of what is actually happening compared to a chew card that has been nibbled on by multiple animals but you would have no idea how many visits or often even what type of animals visited. 

Possums interacting with a chew card

Scent trails are known to be useful for catching predators. For example, the catch rate of possums in a trap is higher on a second night if one has been caught on the previous night and that this difference goes away if it rains (anecdote from talking to ZIP).

Similarly, Good Nature's setup guidelines recommend deploying chew cards around a location and then installing traps where the chew card that has been eaten the most. This is a useful method in an urban environment but probably not so scalable in natural bush.

Based on our experience of watching tens of thousands of videos of predators walking past traps with food lures, we suspect that the same thing happens for traps as for chew cards. The default behaviour for most traps is that the animal walks straight past it but once one animal goes into the trap then others are likely to go there too.