Discovering Dark Koalas

You might have read about dark matter, and more recently about dark DNA. Basically it’s stuff that we know is there, but it’s very hard to find. We can’t detect it easily, but it plays an important role. Based on those characteristics it is clear that we have discovered dark koalas in the Blue Mountains.

We’ve been a bit quiet on the news front in the past few months due to spending a huge amount of time out in the field, chasing evidence of these dark koalas. We just completed some research trying to work out what the best methods were for detecting these elusive beasties. One of the key things we need to do is to map out where koalas occur in the Blue Mountains region, to work out which habitats they like and which tree species they prefer. Three years into the project, this has proven to be much trickier than I’d hoped.

In others areas of Australia koalas can be surveyed relatively easily by walking around and looking up into the tree canopy. These are called “direct count” survey methods, where you are looking for the actual animal. They can be carried out during the day, or at night using a spotlight. These methods are commonly used and they are fabulous for engaging volunteers as people get to see koalas as a reward for getting a sore neck from looking up all day.

Based on all our efforts trying to find koalas to put collars on them and then tracking them, it became clear that direct counts might not be the ideal survey method to use in the Blue Mountains. For example we can track down one of our collared koalas by radio signal, narrow its location down to one tree, and still spend 20 minutes or more trying to see it. These things just dissolve into the canopy.

The other type of surveys, known as “indirect” survey methods, involve looking for signs of the animal rather than the animal itself. Scat counts are one such method and they are a great tool as you don’t have to be in the same place at the same time as the koala to find evidence it has been there. Since scats accumulate in the landscape this method has the advantage of sampling time as well as space, a multidimensional approach more worthy of dark koalas.

Scat counts are known to be particularly useful for areas where koalas are sparse, or in low density. If a koala is moving over 170 hectares then you might have low odds of seeing it, but better odds of finding some poo from a week or two before.

dark koala in canopy
A dark koala looks a bit like this, except it would be hiding in the clumps of leaves and impossible to detect. Imagine this photo without the koala in it, and that is exactly what dark koalas look like.


scat in leaf litter sml
A fresh scat in the leaf-litter, evidence of dark koalas. Look for the small greenish blob near the base of the tree

However, we’ve got a pretty good idea that one of our study sites has at least a medium density of koalas, their homes ranges are small and there are plenty around. So in theory direct counts should work in sites like these.

We tested out these different survey methods to see which ones worked the best in different habitats. The results were surprising, I had underestimated the power of dark koalas. We tested the methods across three different vegetation types where we knew koalas were present, based on data from our collared animals. Each vegetation type was slightly different; from open woodland with smaller trees and more open canopy, through to tall forest with thick canopy. The visibility of koalas, or detection probability, was therefore different in each vegetation type.

Direct count methods failed to confirm the presence of koalas in two out of the three vegetation types. And we tried both daytime surveys and spotlighting at night. The scat counts on the other hand confirmed the presence of koalas in all three vegetation types.

At the sites where the direct counts did work, they only detected one or two koalas from 30 surveys, not nearly enough to estimate koala numbers or density. So, now we know that if only direct count methods are used in areas where habitats are complex there is a very good chance of failing to find koalas when they are present, even if they’re in reasonable numbers. That’s really important as you can imagine, particularly if you’re assessing land for development to see if this threatened species is present or not.

Detection dogs confirmed their status as the best secret weapon to detect dark koalas, by finding more scats and in a fraction of the time it took people to find them. But if you are detection dogless, then having mere humans out doing scat counts is still worthwhile.

Moving on from this result we need to focus on honing our scatology skills and collect some more information on how long scats remain in the landscape at these sites, and how much poop your average dark koala produces, to see if we can come up with clever models to estimate koala density from scat counts. Then we need to survey the rest of the Blue Mountains region using scat counts, which could take some time.

We had a dedicated group of people helping with this research, particularly Adrian Sujaraj who is completing his University of Sydney honours research project with us and has shown amazing dedication by painlessly raking the leaf-litter for koala poop across 90 sites. He is doing a lot of the hard yards field-testing the survey methods. We’ve also had a fine team of undergraduate university placement students and volunteers helping out and gaining critical skills in poop and vegetation identification, you never know when you might need them. A big thanks to all!

This research was made possible by funding from the NSW Office of Environment and Heritage under their Saving our Species program, and ongoing core funding from San Diego Zoo Global.

Executive Director, Science for Wildlife