The wolf populations of Europe are growing and recolonising former territories. This recolonisation is accompanied by an increase of livestock predation. In some areas, extensive grazing practices is threatened due to the technical, financial, and/or psychological cost induced. This, in turn, threatens wolf conservation as it is used as the main arguments against the predator. At the European level, the wolf is a protected species and extensive livestock farming is valued. This raise a heated conflict of interest between extensive farming and the species conservation goals. In many cases, the preventive measures currently available to farmers – shepherding, fences, livestock guarding dogs, neophobic deterrents - are not sufficient to reduce predation (for example, predators may find their way around the preventive measures in place despite being well implemented, some landscapes/pastures are impossible to be fenced or guarded by dogs) or they create other issues (for example, conflict of landscape usage, local nitrate pollution at night enclosures, competition for funding, etc). Besides, lethal methods are prohibited or strictly limited. Where the latter is true, they raise emotional debate and their effectiveness within the legal framework boundaries is unclear. For those reasons, sustainable non-lethal methods are still needed and would greatly improve the implementation of the cohabitation policy presently advocated in Europe. Early research on preventive measures and animal behaviour theories suggest that it would be possible to condition an aversion (i.e. to “teach” a negative experience, a rejection) in predators toward preying on livestock. However, tests on wild wolves are rare and no scalable solution have been found yet. This project will be dedicated to the testing of two methods based on those aversion conditioning principles. We hypothesize they can lead to sustainable methods that could be effectively implemented on a large scale to protect sheep from wolves. We will collect both fundamental behavioural knowledge on wolf reaction to aversive agents and practical data on livestock protection.
We aim to test two new non-lethal methods to prevent sheep predation by wolves.
One method consists in a lightweight cover made of fabric stripes impregnated with a concentrated chili pepper extract and hanging around the sheep: a "defence harness". The harness should not impact the sheep, be cheap and easy to install to become a viable solution on a large scale and durable in time. The fabric has a strong colour contrast so the harness becomes a clear visual stimulus warning the predator.
The second method is rooted in the research on conditioned food aversion (CFA): the mental association of a food item with toxicity/sickness symptoms (naturally or artificially triggered) leading to the recognition and avoidance of toxic food. Baits will be treated with a harmless odour stimulus, new to the wolves (e.g. vanilla food flavouring), and a non-detectable vomiting/toxic agent, not impacting wolf survival but rapidly triggering negative digestive symptoms. Tobajas and his colleagues have identified several good candidate emetic compounds but they all raise ethical and technical issues. To date, thiram, a fungicide, is the best candidate, both in terms of safety and effectivity. However, its use would require derogations for research purposes as it has been recently banned from the European market: thiram primary use in agriculture is suspected to cause environmental pollution and there is a lack of documentation on its (environmental) toxicity. Thus, thiram would work to validate the method principles but would not work on the long term for largescale/manager implementation. Alternatives must be found. The baits will be strategically placed around attacked sheep flocks for wolves to find them. Two “delivery” bait systems are being considered to limit access to the emetic compound only to wolves: encapsulating the emetic compound in hollow bones, that only wolves would be able to crack open, or placing the baits under a heavy mass that only strong mammals like wolves would be able to push open. If wolves indeed learn to predict and avoid the sickness symptoms thanks to the odour stimulus, they should avoid baits treated with the odour stimulus alone. If so, neighbouring sheep will be equipped with the same odour stimulus (the “conditioned stimulus”) to see if the learned aversion for the odour-treated bait can be transferred to the sheep. The odour stimulus could simply be placed on the sheep ear-tag or collar but we expect it will need to be sprayed all over the wool to be effective.
Conditioning refers to a “learning by association” process, where the presentation of two simultaneous or sequential stimulus are paired in the mind of an animal so that the first stimulus can be used to predict the second. If a neutral stimulus, a bell sound, for example, is always present just before a painful (i.e. aversive) stimulus, an electric shock for example, most vertebrates will eventually associate the two stimulus so that the neutral stimulus will provoke anticipation of the negative stimulus: this is a conditioned aversion (reaction) where the neutral stimulus becomes the conditioned stimulus. The number and frequency of encounters of the paired stimuli, the intensity of the aversive stimulus, the coherence between the neutral and aversive stimulus, the saliency and novelty of the neutral stimulus are all factors that influence the conditioning process. The two methods we intent to test are based on such principles and have been conceived with all those factors in mind. The idea is to increase the cost (for the defence harness method) or the perceived cost (for the conditioned food aversion) of predation of sheep for the wolf, and, doing so, decrease significantly, if not eliminate, predation.
For the defence harness, the aim is to conceive a sort of armour that can be permanently worn by the sheep and provide a harmful (aversive) experience to a preying wolf that would bite in it. We hypothesize that the harness, a very salient visual and olfactive stimulus (the chili pepper extract has a pungent smell) will be readily associated with the pain and subsequently avoided. Only if and after the learning occurred will the equipped sheep be protected. One can compare it to aposematism and defence mechanism use by prey species such as wasps or the porcupine. Previous research on this topic suggests it is impossible, if not counterproductive, to try to hide the origin of the aversive stimulus so that the sheep itself appears to be the danger instead of the harness. A first trial carried out in February 2022 allowed us to refine the harness design and suggests that harnesses do not affect sheep welfare. Further tests are planned from autumn 2022 to autumn 2023 to monitor sheep welfare, shepherd dog’s reaction to the harness and ultimately, wild wolves’ reaction.
The second approach intends to conditioned a food, or more precisely, an odour, aversion. Food aversion conditioning is a powerful learning mechanism. It is instinctive to many species and prevent animals from food intoxication. For this reason, it is a promising and well recognised approach. However, no practical implementation strategy have ever been found and many trails have failed due to the lack of an undetectable aversive agent (i.e. tasteless and odourless). By placing an odour stimulus on baits and an undetectable emetic compound, the wolf should, after 1 to 5 consumption – “learning experiences” –, use the odour stimulus to distinguish between a safe bait and a “toxic” bait causing digestive symptoms. If the wolf is capable to generalise this learned association to different food baits treated with the same odour stimulus, the odour stimulus could be sprayed on sheep to protect them. This approach allows the separation of the conditioning (learning) phase from the protection phase. This is particularly helpful for large scale application where the alternative to few strategic baits would be the simultaneous treatment of thousands of sheep with an emetic compound; an unrealistic solution. The odour stimulus should also accelerate the learning process because it adds a stimulus unknown to the wolf that triggers its attention; in scientific terms, it avoids “latent inhibition”. Everything is to be learned: how to avoid baiting non-target species? How cost-effectively can wolves be baited? What bait works best? Can wolf learn from other pack members? How fast learning and forgetting occurs in wild wolves? Would wolf generalise what they have learned with a bait to a live prey? Would food aversion discourage predation behaviours or only consumption behaviour?
CONDITIONS DE SUCCÈS
- Field constrains and the many factors influencing predation rate render robust scientific evaluation of the tested methods difficult and costly.
- Only few works are available to inform the protocol. We are still at the infancy of understanding wolf ecology and wolf aversion conditioning.
- It is difficult to find support as the society is divided on how to deal with wolves and with livestock predation. Little to no support to the project have been provided by stakeholders in France after a year of trial. At this stage (mid 2022), neighbouring countries seems to show the same constrains. Finding cooperative farmers is difficult so far.
- Evaluating the proposed methods within a scientific protocol raises great ethical concerns and the wolf is a strictly protected species. Many steps of the project thus require ethical committee and local authority approbations that are difficult to obtain. This is greatly delaying the project implementation.
- An authorised, effective, safe, and acceptable emetic compound is still to be found for the conditioned food aversion method.