discussion / Wildlife Crime  / 23 May 2017

Snare detection technologies

Snares are a pervasive threat to wildlife around the world – indiscriminately killing hundreds of thousands of animals.

Snares are notoriously hard to locate in the field, and it would be a major advance if a handheld (or UAV-mounted) snare detector could be developed, to better scan landscapes and help Park rangers to more easily find and dismantle them.

WWF's Wildlife Crime Technology Project has been exploring radar as a potential solution, and this has shown some promise. We'd be grateful if field-based practitioners could share their experience with snares in this discussion thread, and if global engineers and tech-minded colleagues on WILDLABS could input on possible solutions.

Thanks.




Hi Rachel, 

Sam Williams ( @SamWilliams ) over on Twitter has underscored the importance of your qustion by sharing some of his work investigating the impact of land reform on the status of leopards in Zimbabwe

Snares are a huge problem. Being able to find them could be a game changer. Any ideas? https://t.co/awXBp5VOXO https://t.co/5FFjshdFih pic.twitter.com/kFdLY8qpyx

— Sam Williams (@_sam_williams_) May 24, 2017

In the discussion of his paper, he writes: 

The extremely high levels of poaching in Savé Valley Conservancy (SVC) were the result of a large human population being settled on private land with large wildlife populations, and were exacerbated by Zimbabwe’s economic crisis and food shortages arising from the FTLRP (Knapp, 2012; Lindsey et al., 2011aMoss, 2007), limiting carnivore abundance in the private LUT. Poaching rates in SVC increased to extremely high levels after the FTLRP began; between August 2001 and June 2009 over 84,000 snares were removed and 4,148 poachers were captured (Lindsey et al., 2011b). The remains of 6,454 poached animals were recovered, including 2 cheetahs, 5 leopards and, 27 wild dogs (Lindsey et al., 2011b). Numerous individuals of prey species were also recovered during this period, such as 2,606 impala (Lindsey et al., 2011b), which would reduce carnivore carrying capacity through removal of the prey base (Hayward, O’Brien & Kerley, 2007). Within the private LUT, rates of poaching per unit area were over 2.5 times higher in the south than the north (Lindsey et al., 2011b), which is probably linked to greater proximity to the resettlement area (Fig. 1). When resettlement occurred the perimeter game fencing was stolen, facilitating access of poachers from the resettlement area to southern SVC and providing abundant material to manufacture snares (Lindsey et al., 2009). While fencing can be an incredibly useful tool for managing wildlife populations (Packer et al., 2013), it is important to use material that cannot be easily used to manufacture snares (such as Veldspan™ or Bonnox™), rather than the steel and barbed wire that was used to construct the fence at SVC (Lindsey et al., 2012).

(image credit: Samual WIlliams via twitter)

This has certainly stirred up some discussion over on Twitter, so in the interest of capturing it here for future searchablity and discovery I will continue to pull them over into this thread. The first discussion centred around the idea of using detection dogs: 

....

You *could* train a dog to not mind sudden restraint; you'd need a Kevlar outfit for her tho.

— jess (@earlyelliott) May 24, 2017

I'm assuming these are foot snares.
Also, whoever sets the snare must have a way to find them

— jess (@earlyelliott) May 24, 2017

Looking into it more, it might get tight enough to cause pain. Easier to teach to sniff out.

— jess (@earlyelliott) May 24, 2017

You could use a dog small enough that she wouldn't set off the traps.

— jess (@earlyelliott) May 24, 2017

 

....

Secondly, this is interesting because it's some out of the box thinking. My thoughts went more high tech, mainly because of the scale

— Stephanie O'Donnell (@Steph_ODonnell) May 24, 2017

I agree that this is a really interesting idea. Not sure if doge have been used in this context before.

— Sam Williams (@_sam_williams_) May 25, 2017

 

...

84000 snares found in one conservancy in less than 10 years? That is a lot of snares to detect

— Stephanie O'Donnell (@Steph_ODonnell) May 24, 2017

Wow that's a lot! Could probably use a metal detector along common paths, then

— jess (@earlyelliott) May 24, 2017

Dogs could play a role on a more local scale - e.g. help foot patrols find snares in areas identified as hotspots by drone-mounted sensors

— Sam Williams (@_sam_williams_) May 25, 2017

Leopards tend to use the same paths, right? Easier to de-snare wildlife corridors.

— jess (@earlyelliott) May 25, 2017

...

Great idea. Do you think it would be possible to train dogs to sniff out wire snares?

— Sam Williams (@_sam_williams_) May 25, 2017

Yes, they are used to find mines, ammunition, guns. Easier in wild w no other metal smells!

— jess (@earlyelliott) May 25, 2017

An interesting choice would be giant rats. Too small to set off snares https://t.co/MdFIvHYnoe

— jess (@earlyelliott) May 25, 2017

...

Rachel, or perhaps one of the members of our Conservation Dogs group, have you had any experience using detection dogs in this manner? If they were indeed effective at detecting snares, this would still be a local, small-scale solution, surely? Given you're looking into radar, are you more interested in technologies that offer a large-scale detection range?  

Wild out-in-left-field ideas:

- A snare usually involves a long wire stretched taut. Could you broadcast a tone and listen for a resonance in the snare wire like two guitar strings tuned to the same note?

- Same idea, but could you detect a long wire as a radio antenna?

Can you post some pictures of what a typical snare looks like, how it's hidden, or what it's composed of? It sounds like a very challenging computer vision problem, but if the material stands out against a natural background (plastic/metal against trees) there may be a thermal or UV signature that you could filter for that would give away traps.

A second thread of conversation involving Nigel Street (@canuck069 ), Nikki Rust (@NikiRust ) and Sam Williams   (@SamWilliams ) over on Twitter is discussing drones or robots:

...

Would a sensitive metal detector fit on a drone and have the drone fly low over game trails?

— N Street  (@Nigel_Street) May 24, 2017

My thoughts exactly. I don't know if metal detector has the range for this, but if so leveraging drones would allow you to scan a vast area

— Sam Williams (@_sam_williams_) May 25, 2017

 

...

Could you program a robot to scour the area searching for metal & cut the metal once it's found?

— Niki Rust (@NikiRust) May 25, 2017

Nice idea! If you could do that you could automate the whole process.

— Sam Williams (@_sam_williams_) May 25, 2017

Pretty sure @NASA have robots that could be programmed to do this

— Niki Rust (@NikiRust) May 25, 2017

Those robots used for bomb disposal would also presumably also be capable of this

— Sam Williams (@_sam_williams_) May 25, 2017

Exactly. We have the tech. Just need someone to reprogram it. Cost implications of course though!

— Niki Rust (@NikiRust) May 25, 2017

Might also need to work on manoeuvrability for snare-cutting robots. Fighting through dense thorn bushes & rugged terrain might be tough

— Sam Williams (@_sam_williams_) May 25, 2017

If only we could have metal-detecting drones that could do aerial scans of the area then swoop down & cut metal they find?!

— Niki Rust (@NikiRust) May 25, 2017

Yes please

— Sam Williams (@_sam_williams_) May 25, 2017