Tuesday, September 29, 2009

postheadericon Finding 74 needles in a haystack - the search for camera sites is on

An understatement….. estimating the density of Sunda clouded leopards using camera traps isn't easy!

In addition to the physical challenges of deploying and repeatedly checking the network of cameras the statistical framework to estimate density requires that several (ideally at least 10) different cloudies are photo-trapped on more than one occasion (as many as possible). However, getting a cloudie on camera can be a rare event, requiring as much as 200 camera trap nights* to get a single photograph. Randomly placing your cameras will almost certainly result in zero photo-captures!

Clearly, finding the best locations to situate the cameras is of vital importance, and, ultimately will make the difference between a successful or unsuccessful density survey. A sobering thought! The problem is, finding the perfect camera site is like looking for a needle in a rather large haystack, and finding 74 such sites is a real challenge.

The Bornean Wild Cat Team, after a hard day cutting trails. Left to right: Jo, Sajaril (Itoi),
Remy, and Jasmi (Jas), (Andy -taking photograph).

Fortunately we have nearly 3 years of experience to help guide us, and there are certain tricks that we’ve learned along the way. For starters, when surveying for Sunda clouded leopards in Bornean forest old logging roads are like gold-dust. We’ve found that they can be prime locations to photo-trap cloudies and, in addition, they provide us (and, unfortunately, poachers) with much needed access to the forest interior. We’ve discovered that cloudies tend to favour travel routes that require less effort; natural ridgelines and well worn existing human trails (ideally relatively dry underfoot) can be hot-spots for cloudie activity, although these associations may also be related to hunting success.

Thus, on arrival in Tabin our first priority was to survey the area for existing (i.e. still drivable) and abandoned (i.e, not drivable and swathed in recovering forest) logging roads. We discovered two existing roads, one running north-south along Tabin’s western border and one running west-east for about 10 km through the centre of Tabin. The first road is used extensively by plantation contractors, and so the risk of theft would be too high to deploy cameras here, but it will provide much needed access along the Western border. Despite also posing a theft risk (although to a lesser degree) the second “Core Area” road will almost certainly be a clouded leopard hotspot and will thus be of such intrinsic value, in terms of getting photo-captures, that we will have to run the risk and have decided to place six camera pairs along its ca 10km length.

The search for abandoned logging roads is very much more a hit or miss affair, requiring plenty of guesswork and a lot of luck. We start by looking at a very course-scale topographic map of Tabin and try to guess where the old logging roads were built, typically along flat areas, ridges etc. and we then go out on foot and try to locate them. Frequently our searches come up empty handed, and we simply try to plot a course through the forest to the approximate area that we need a camera. We then search the area for the best location, set up the cameras, and then cut a 200m section of trail, centred on the cameras. We even sweep the trail clear of leaf litter and other debris, in an attempt to “persuade” the cloudies to use the trail.

Four wheeled drive vehicles and ATVs are a must for survey work in Borneo.
New team member, Jas (L) and Sajaril (R).

So how are we getting on? Well, as of the end of September we have already finished setting the first 37 camera sites – which required cutting over 60km of trail! We have been successful in finding several old logging roads, some good (e.g. one we’ve named Jalan Raya bagus = great logging road) and some not so good. We’ve found several very promising ridglines and are very optimistic. Watch this space for news of how we get on!

*A camera trap night is a unit of trapping effort. One camera trap night is equal to one camera operating for a 24hr period; so 10 cameras operating for 10 days equals 100 camera trap nights.