Dr Peter Apps runs Botswana Predator Conservation’s BioBoundary Project


“We would accomplish many more things if we did not think of them as impossible.” Vince Lombardi

Botswana Predator Conservation’s BioBoundary Project is developing new ways of mitigating human-predator conflict, by using artificial equivalents of natural chemical signals to keep predators away from livestock, or safely inside protected wildlife areas. https://bpctrust.org/project/bio-boundary-project/

Read it quickly enough and that sounds eminently practical and straightforward. Predators use odours, that are detected and decoded by the receivers’ incredibly acute sense of smell, to send messages that modify one another’s behaviour. The odours are mixtures of airborne chemicals, and if we use synthetic chemicals to replicate them, the artificial odours will send the same messages as the natural originals, and have the same effect on the receivers’ behaviour. Those effects can be used to keep livestock safe from predator attacks and protect threatened populations of predators from lethal control measures.

From the challenging…

The strongest hint that using scent to manipulate carnivore behaviour might not be as easy as it sounds, is that it has never been done before; prior to our research only one artificial scent-mark had elicited specific responses from wild predators, and that one had no application to managing predator behaviour.

There are all kinds of reasons why using chemical signals to influence predator behaviour is a significant challenge. A typical predator scent has upwards of 1000 components that test the limits of analytical and synthetic chemistry. To use artificial scents as management tools we have to understand the biology of scent marking as well as the chemistry of scent-marks, and that has its own challenges; once we get out into the wild, and away from mice, rats, hamsters and ferrets in the lab, mammals’ responses to odours are erratically variable to the point of quirkiness, and the details of the messages that the odours are transmitting still need to be deciphered.

The baffling chemical complexity of a predator scent – just one section of a gas chromatography – mass spectrometry trace from African wild dog preputial gland secretion


To the impossible …

“Alice laughed. ‘There’s no use trying,’ she said. ‘One can’t believe impossible things.’ Lewis Carroll

In addition to the wildlife biology and analytical technical challenges, turning predator chemical signals into conservation tools faces another hurdle; a widespread belief that it can’t be done. That belief is based partly on a feeling that nothing as insubstantial as a smell could be a tool to manage the behaviour of something as undeniably substantial as a leopard or hyaena. Smells lack the physical presence of poles and wire mesh, and the high-tech allure of geofences and automated alarm systems, and as humans whose sense of smell has been relegated to a minor role in our social lives, we find it impossible to imagine the significance to other animals of communication by airborne chemicals.

“To believe a business impossible is the way to make it so. How many feasible projects have miscarried through despondency, and been strangled in their birth by a cowardly imagination.” Jeremy Collier

On top of that, many researchers working on mammal chemical communication have convinced themselves that because mammal odours are bewilderingly complicated chemical mixtures, the signals embedded within then must be coded by subtle changes in the concentrations of multiple components. These chemical signals are supposed to be so complex, and so subtle, that they can be replicated only by creating facsimile copies of real odours; a Herculean task in synthetic organic chemistry that would cost about five million dollars per gram. A complete lack of evidence to support it has not stopped this conviction taking hold, and imposing a paralysing pessimism on a field that should be developing for mammals the equivalents of environmentally friendly and commercially viable synthetic insect pheromones.

So, when we set out to use chemical signals to protect predators and livestock, are we attempting the impossible ?

Obviously we do not think so – even though, as Walt Disney said, it would be kind of fun.

Seeing is believing

Our strategy is to focus on the parts of the chemical messages that are critical to the behaviour change we need to create; to keep predators from straying into human-dominated landscapes or getting dangerously close to livestock, we will post simple chemical “no trespassing” signs instead of the complex chemical equivalent of a 10-page property deed. Leaving out the irrelevant chemical fine print, and including in our artificial scent-marks only the few critical compounds that capture the essence of the message, will simplify the chemistry and lower the costs.

“The only way to discover the limits of the possible is to go beyond them into the impossible.” Arthur C. Clarke

For our strategy to work, predators will need to defy the experts’ pessimism by responding to single compounds and simple mixtures. If animals respond to single chemicals as if they were real scent-marks it will show that at least some parts of the signals are coded by simple chemistry, and that scent-based conservation management tools for predators are within reach.

The good news is that responses to single chemicals are exactly what we do see in field experiments funded by the Leopardess Foundation; and we have the hard evidence of camera trap videos to convince the sceptics.

In our very first feasibility study, set up to find out if it was practical to release odours at controlled rates and monitor animals’ reactions with camera traps, we captured videos of African wildcats, genets, civets and slender mongooses scent-marking in response to the smell of a single chemical that occurs in leopard and tomcat urine. That same chemical; 3-mercapto-3-methylbutanol – 3M3MB for short – also induced a female leopard to make a U-turn as she caught its scent, and it was the video of that U-turn which launched our research into repellents as conservation management tools for African predators.

A female leopard grimaces and turns back as she catches the scent of 3M3MB

Now, as we screen more components of predator scents to test their use as repellents, we are recording more responses to single compounds and simple mixtures. The screening setup is an improved version of what we used in the original feasibility study; controlled-release scent dispensers monitored by camera traps that record animal’s responses on 30s video clips.

A camera trap in its steel housing, (right), a repllent dispenser, and a controlled release vial that slowly releases repellent.

How rapidly a potential repellent is released is adjusted so that there is a definite odour to the human nose at the mouth of the vial, but no detectable odour a few centimetres away. We want to repel predators, not people; a human must be able to walk past a scent dispenser and smell nothing. By precisely weighing the vials repeatedly over periods of a few weeks we can calculate that they emit each chemical at a rate of micrograms or milligrams per day. The dispensers are protected by robust housings made from low-cost plumbing connectors welded to metal spikes hammered into the ground, leaving the housing 20-30cm above ground level, roughly nose height for the animals whose reactions we want to record. Each test odour gets a month or two in the field, and then it is replaced by a different one.

A repellent screening station; two camera traps (centre foreground and background) keep watch over two repellent dispensers (left and right)

We are looking for videos of animals being repelled at short range, or for the numbers of videos of a species to drop because animals stay out of the camera traps’ detection zones when they are repelled by the experimental odours. Any videos of animals sniffing intently, licking, biting, rubbing or scent-marking on the dispensers are a bonus confirmation that predators do respond to single compounds and simple mixtures.

In addition to the leopardess being repelled by 3M3MB, we have videos of four other species of carnivores being repelled at short range, by three other chemicals. Everyone knows that “honey badger don’t give a shit” – but we have a video of a honey badger definitely giving one about repellent TC20/21 18; sniffing intently as it approaches the dispenser, stopping dead about 1m away, and then heading off in another direction.

A honey badger repelled by the odour of TC20/21 18

Black-backed jackals – the species that causes the biggest losses among small livestock – are also repelled at short range by TC20/21 18 and by another, chemically similar compound; TC20/21 17. Videos show black-backed jackals stopping and sniffing before they reached the dispensers, sometimes fluffing up the hair on their backs, and then turning away.

A black-backed jackal stops and changes course after smelling TC20/21 18

One of the jackal repellents also repels Cape foxes and bat-eared foxes, and bat-eared foxes also avoid TC20/21 21.

To further confound the conventional view of mammal chemical signalling, predators of seven species responded to six single chemicals as if they were real scent-marks. African wildcats rubbed and urine-sprayed in response to the smell of 3M3MB, TC20/21 18 or TC20/21 21, Cape foxes urinated and anal-gland marked on dispensers loaded with TC20/21, TC20/21 18 or TC20/21 20, slender mongooses rubbed and anal-gland marked on dispensers loaded with  3M3MB or TC20/21 21, and a genet anal-marked in repsonse to 3M3MB.

An African wildcat sprays urine in response to the odour of TC20/21 18

A leopard sprayed urine after it smelled TC20/21 20, brown hyaenas sniffed and mouthed the dispensers in response to the odour of TC20/21 21, and both brown and spotted hyaenas reacted to TC20/21; a spotted hyaena by licking and biting the dispensers, and a brown hyaena by sniffing and then immediately scent-marking with anal-sac paste about 2 m away.

A leopard responds to a single chemical; TC20/21 20, as if it was a chemical message, by spraying urine

A brown hyaena scent-marks immediately after sniffing TC20/21, just as if it was responding to a natural scent-mark

“Everything is theoretically impossible, until it is done.”Robert A. Heinlein

According to the pessimistic orthodoxy, none of this should have happened. Predators shouldn’t be taking any more than cursory notice of single compounds, far less being repelled by them or responding by scent-marking as if the single compound was an actual chemical message.

Only short-range repellents produce reactions that the camera traps can record; with more potent repellents the predators do not approach closely enough for the camera traps to capture them, leading to the records for those species having gaps while the potent repellents are released. The first absence of camera trap captures was of leopards; a camera trapping effort that was expected to yield about 20 leopard videos produced only one, and that was the female that did the U-turn when she caught the scent of the 3M3MB. In the ranching area, spotted hyaena videos dropped from an average of 6.4 per month to zero while TC20/21 17 was being released, and then jumped back up to 7 per month when the TC20/21 17 was replaced with other odours. Caracals were captured at an average rate of 3.7 videos per month, but disappeared when the dispensers were loaded with a mixture of five volatile aldehydes. Versatile and resilient black-backed jackals were the most common captures of all, with an average of 22 videos per month, and then, while TC20/21 21 was being released, they disappeared completely, and re-appeared only five months later.

These are only preliminary results from small scale screening tests, but they are still exciting enough, because anything that can keep the four most serious livestock predators; leopards, spotted hyaenas, black-backed jackals and caracals away from livestock obviously has huge potential for protecting livestock from attacks, and protecting predators from lethal control.

Nonetheless, conclusions would be premature. Could it have been co-incidence, that the hyaenas and caracals just happened to make themselves scarce while particular components of predator odours were being released ? Did this year’s unusually heavy rains and lush growth of vegetation have an effect that previous dry years did not? Do three different odours all repel jackals at long range, or did one or two of them have long-lasting effects ? Were the absences connected to mating or breeding seasons ?, maybe, but at the same times last year when the chemicals were not being released, the video captures continued. Although nobody else has recorded anything similar, jackals persistently avoiding a location after smelling an odour there may not be so surprising. Lab research shows that mammals can remember odours for weeks or months, and also link scent memories to specific locations, and this isn’t the first time that an odour release has kept predators away from scent stations on ranches; in a pilot test, leopards, spotted hyaenas and African wild dogs stayed out of a cattle kraal for 14 weeks while 3M3MB was being released. We can be sure that the repellency and responses are not just because the scents were unfamiliar; some other scents, just as unfamiliar, produced no reaction at all. The specific combinations of species and chemicals that yielded avoidance and other responses mimic the species-specificity of natural scent-marks, and just as with natural scent-marks, not every individual responded in the same way, or responded at all. To confirm the repellents’ effects we plan to repeat the releases of the same compounds at different times of year and in other locations.

None of the small carnivores we videoed are a threat to anything bigger than a lamb or baby goat, but they can do serious damage if they get into a chicken run, and although none of their populations are threatened by lethal control, when farmers resort to gin traps and poison to protect their livestock from common species, more vulnerable predators also fall victim. Black-backed jackals are the most serious killlers of small livestock, closely followed by caracals. Spotted hyaenas are serious livestock predators; even the small clans with four or five members that survive in the livestock areas can easily kill full-grown cattle, and that erodes tolerance for all the other predators. These livestock-killing predators being repelled by simple, low-cost scent-mark chemicals provides real confidence in the prospects for effective human-predator conflict mitigation, and that is why we need to scale up the screening tests with more stations over a bigger area.

Proof positive

What we usually consider as impossible are simply engineering problems… there’s no law of physics preventing them. Michio Kaku

Following the leopardess U-turn video and the successful kraal exclusion tests, we have already moved 3M3MB up the application pipeline to a demonstration project where it keeps kraaled calves safe from leopard attacks.

One of seven calves killed at a kraal before the 3M3MB repellent was introduced

Over a period of 18 months, with the release of 3M3MB alternating with no repellent being present (because of supply problems and COVID-19 movement restrictions) calves were attacked only when no 3M3MB was being released. In addition, when the dispensers were loaded with 3M3MB there was less hyaena activity around the kraal, and spoor tracking showed leopards skirting around 30 to 50 m away.

Peter Apps loads scent dispensers with 3M3MB to protect calves from leopards

A repellent dispenser in its protective housing hanging from a kraal fence.












Even though 3M3MB costs about $25 per gram, it is so potent that it needs to be released at only a few micrograms per hour to keep leopards away. The calves it saved were worth 10 times more than the 3M3MB that was used up. The other potential repellents with positive results in the screening tests cost substantially less than 3M3MB, and they will be affordable to even the poorest of subsistence pastoralists, and the low cost will free repellent-based predator conflict management from the need for continual financial inputs from outside donors. In their final form, repellents will be easy to apply, low tech, and zero maintenance; once we have identified the active ingredients the working repellents can be formulated and distributed as small quantities of repellent in controlled-release sachets or capsules that can be hung on a fence or rubbed on a gate post. Farmers will be able buy them where they buy their livestock feeds and worm remedies.

Building up and rolling out

“It always seems impossible until it’s done.” Nelson Mandela

As the evidence builds up, chemical by chemical and species by species, there is no escaping the conclusion that the paralysing pessimism about the practical application of mammal chemical communication is unjustified; that mammals do respond to single chemicals and simple mixtures, and that, if the necessary resources are invested, we can develop repellents based on chemical signals to provide technically straightforward, economically viable tools to protect both predators and livestock.

The field work is supported by a grant from the Leopardess Foundation https://www.leopardess.org/

If you want to support genuine innovation on the frontiers of conservation science, please contact us via the Contact Button on this page or e-mail Dr Peter Apps at peterapps@wildentrust.org

WP Feedback

Dive straight into the feedback!
Login below and you can start commenting using your own user instantly

Pin It on Pinterest

Share This