Notice bibliographique
Résumé
In Israel, human–wildlife conflicts are a challenging issue as animals and people increasingly share space, as is shown here. An ibex clambers up on the roof of a car to reach tree leaves (left). Children approach wild ibex (right). Safety concerns have motivated research to see if aversive conditioning is a viable solution to human–wildlife conflict. Photographs: Oded Berger-Tal. When local celebrity Rutti came to town, some residents were thrilled. Fervent fans created a Rutti Facebook page. But others feared her, wanting her evicted. Then circumstances grew more complicated when Rutti became a mother. In the city of Modi'in, Israel, people remain divided over what to do about this wild hyena that moved into town and began helping herself to food scraps. Not typically dangerous to humans, hyenas are nevertheless “large carnivores that look menacing,” says Oded Berger-Tal, adding that “Disney movies have given them a bad name.” As human populations increasingly encroach on wild ones, this city in Israel is one of many places experiencing conflict with animals. So Berger-Tal, who studies conservation behavior at Israel's Ben-Gurion University of the Negev, initiated a research project to see if hyenas could be taught to avoid human food. Berger-Tal and other scientists and wildlife managers are testing a technique called aversive conditioning to address the thorny challenge of human–wildlife conflict. Aversive conditioning (AC) refers to training an animal to pair a particular behavior, cue, or location with some form of punishment or negative stimulus. If training is successful, animals learn to associate one cue as a consequence of the other. AC has been used on hyenas, lions, bears, coyotes, elk, dingoes, and quolls, among others. In theory, it could substantially reduce the need for removal of “problem” animals. In practice, however, the tool has had mixed success. AC can work, but does not always, so researchers and field practitioners are experimenting and synthesizing their knowledge to find out why. Behavioral ecologist Lysanne Snijders, at Wageningen University, in the Netherlands, leads a project to map evidence on AC. Outlining her team's systematic map protocol in the journal Environmental Evidence, Snijders and coauthors explain that animal conditioning—which includes aversive and reward-based training—can be a valuable tool in the human–wildlife conflict mitigation toolbox. AC draws from classical and operant conditioning, the former referring to how an animal can learn to associate a neutral cue such as a bell or a whistle with the appearance of a second important cue, such as food or pain, as made famous by Pavlov and his dogs. Operant conditioning, in contrast, is when an animal voluntarily changes its own behavior in response to cues, wherein a particular behavior results in a specific outcome. For example, in the Urban Coyote Intervention Program in Edmonton, Alberta, Canada, when a coyote too closely approaches a human volunteer, the animal gets pelted with a weighted tennis ball and learns, researchers hope, to avoid human interactions. A coyote walks across a snowy path in suburban Edmonton, Alberta, where the Urban Coyote Intervention Program is trying to restore coyote wariness of humans. Photograph: Dale Brochu. Coyotes are well known for their ability to adapt to urban settings. But in Edmonton, over the past decade, animals seem to be getting bolder, sometimes approaching people or attacking their pets, explains University of Alberta graduate student Gabrielle Lajeunesse. Coyotes are midlevel carnivores, scavenging and eating small prey such as rats, mice, rabbits, and grasshoppers as well as deer and other large prey. They rarely harm humans, but when they lose their fear because of foraging on garbage or because of humans intentionally feeding them, they can become a menace. A secondary concern in Edmonton is the high local prevalence in coyotes of the liver tapeworm Echinococcus multilocularis, zoonotically transmissible to humans and potentially fatal. To see if coyote reticence of humans can be restored, Lajeunesse recently finished her first field season implementing an AC experiment in Edmonton residential areas. Neighborhoods were designated as control or treatment areas and resident volunteers were trained to measure coyote “overt reaction distance”—that is, the point at which an animal first moved away when someone was approaching. If that distance is 40 meters or above, “we consider the animal to be wary enough and do not treat it,” says Lajeunesse. But if the coyote does not move away from humans when it is less than 40 meters away, trained volunteers run toward it, while shouting and throwing sand-weighted ribbon-trailing tennis balls in the coyote's direction. Lajeunesse notes that animals will keep coming back to neighborhoods if people do not manage attractants. “Food conditioning is a known precursor to conflict in any animal,” says her graduate advisor Colleen Cassady-St. Clair. Lajeunesse adds that it is important to avoid feeding wildlife directly as well as indirectly with unsecured compost, piles of spilled bird seed, and bags of garbage near homes or litter in parks. Will her tennis ball–toting army succeed in making coyotes more wary of people? The results remain to be seen. Another canid that is causing conflict is the Australian dingo. Researcher Rob Appleby, based in Brisbane, is exploring nonlethal options for managing this species. Appleby is a codirector at Wild Spy, a company that specializes in wildlife conservation research and management technology. He has been engaged in work on K'gari (colonial name, Fraser Island) in Queensland, a popular destination for tourists, where wild dingoes have become emboldened. In 2001, a 9-year-old boy was fatally attacked on the island, prompting a public outcry. Dingo attacks on humans are rare but continue to occur periodically, likely because of human feeding. After the boy's death, there was a large dingo cull. Appleby became interested in alternatives to lethal removal. That led him to learn of American scientists using shock collars for coyotes. “Most of the AC literature is a little bit hit and miss when it comes to success rate,” he says. However, as a pilot project, he put shock collars on a small sample of dingoes, demonstrating proof of concept—the remote-controlled collars caused an immediate cessation of concerning behavior toward humans. He sees such methods, which undoubtedly have animal welfare concerns, along a spectrum of options, with the last resort being lethal control. “Unfortunately, lethal control still continues to be the mainstay for managers in various parts of the world,” he says, but there are many other options, with AC among them. Two volunteers in Edmonton launch weighted tennis balls toward fake coyotes as part of their training to participate in the Urban Coyote Intervention Program. Photograph: Gabrielle Lajeunesse. Reducing conflict is also the goal of wildlife managers in areas of heavy human visitation, such as Rocky Mountain parks. Jasper National Park is one of the busiest parks in Canada, drawing some 2.5 million visitors annually. The town of Jasper—with its hotels, restaurants, golf courses, and residential areas—abuts the rugged wilderness, creating the potential for wildlife conflict, explains Steve Malcolm, wildlife management specialist with Parks Canada. Normally, daily management of animals is done using hazing—removing an animal from immediate conflict using deterrents but without follow-up action. However, if an animal becomes particularly problematic, conflict management staff may work around the clock, focusing on behavior management. If a problem bear shows up at two in the morning, “then I get up at two in the morning,” says Malcolm. Within the park, officials use a series of tools to gain the compliance of grizzly and black bears, cougars, and elk. AC tools range along a continuum of severity from the human voice saying “Hey, bear,” to noisemakers, paint balls, bean bags, or rubber bullets. “Once animals are exhibiting what we want, we lay off the tools. Eventually, all you have to do is show up, and they'll start responding,” says Malcolm, explaining that their time and labor-intensive training program allows humans and animals to coexist. Other carnivore AC work aimed at minimizing human–bear conflict included Lori Homstol's work in Whistler, British Columbia, a resort town with a high density of black bears with little wariness around people. Taking an experimental approach while working with conservation officers, Homstol, during graduate research with St. Clair, would follow a bear closely for 5 days, tracking situations of potential conflict—such as walking through a residential area in daylight. Bears designated as “conflict animals” were assigned to a control group or one of two treatment groups: AC with or without sound. For the sound group, Homstol would blow a whistle 1 or 2 seconds before a conservation officer hit the bear with a rubber bullet or with slingshotted marbles. The marbles were likely not painful “but had a psychological power,” she explains, because the bear experiences human “touch” from a distance. Homstol measured bear wariness before and after treatment to see if behavior changed, measuring overt reaction and displacement distances—the distances at which the bear noticed them and left the area as they approached. In theory, once bears were trained with AC, hotel security guards could blow whistles to deter the animals from underground parking or garbage, then call the conservation officer. The bears in the control group, as expected, habituated to people. In contrast, AC-treated bears did not habituate, had a longer overt reaction and displacement distance, and usually moved away when humans approached. Homstol did not find a difference between the sound group and the regular group. Homstol remembers one adult male in particular. “He had learned that if he approached golf carts, people would scatter, and he could help himself to the beers on the golf cart.” She and her research assistant spent 2 days from dawn to dusk aversively conditioning him. They rode in a golf cart, and when the bear approached closely, they used loud noisemakers to deter him. After 2 days, he stopped approaching golf carts. Although she is no longer in Whistler, Homstol's bear-teaching methods are still used there. In Zimbabwe, lions are frequently in conflict with livestock herders, so as Lisanne Petracca, at the University of Washington, explains, an AC initiative called the Hwange Long Shields Community Guardian program stepped in. Prides in this region tend to be radio collared, so workers in the program get real-time fixes on where the lions are. If lions approach an area where conflict or harm to livestock is likely, the lions are chased by the guardians. “They use plastic horns called vuvuzelas… and the goal is to get the lions spooked so they cross back into forestry land and into Hwange National Park,” she says. Petracca's research revealed that the consistency of chasing is important and that younger lions may be more amenable to AC. For older, chronic stock killers, vuvuzela chasing was ineffective: The fear of the chase does not outweigh the benefit of a good meal. Black bears are often attracted to garbage and can become dangerously habituated to human food sources, leading to conflict and danger. Aversive conditioning is one way that humans can reverse this behavior so that bears and humans can live harmoniously as neighbors. Photograph: Yellowstone National Park. Breaking a food habit is also what motivated Vincent Janik, director of the Scottish Oceans Institute at the University of St Andrews, to investigate acoustics as seal deterrents. Acoustic deterrent devices often used at fish farms create extremely loud noises that not only discourage seals but also prevent nearby cetaceans from accessing their natural habitat. “If you don't hurt the animal with the sounds you are producing, they are likely to habituate,” says Janik. But over time, physical pain from loud undersea noises can erode marine mammal hearing, making the acoustic deterrent less effective and also undermining a cetacean's ability to feed and communicate using echolocation. Janik set out to find a quieter, more annoying noise—the seal equivalent of “fingernails across the blackboard,” says Janik, “the kind of sound that makes you want to leave the room.” Janik experimented with sounds to see which elicited a startle reflex in wild-captured gray seals in the North Sea. Janik's team then worked to pair the startling noise with a nonstartling warning sound. Because marine mammals have hearing sensitivities in different sound ranges, “we can play a low frequency sound that startles a seal but doesn't startle a dolphin,” says Janik. He is now in discussions with fish farm regulatory agencies in the United Kingdom, and investigating whether startle noises can deter seals gorging on runs of declining salmon at Ballard Locks, near Seattle, Washington. Keeping marine mammals away from food is also what got Zac Schakner interested in AC. In the Gulf of Alaska, longline fishers for sablefish have experienced increasing sperm whale depredation. Sperm whales cue into the sound of fishing vessels pulling up gear, using it like a dinner bell for fresh fish. For mitigation measures, Schakner could not experiment on endangered sperm whales but embarked on a study to see if he could use AC to keep California sea lions away from bait docks—floating cages where fishers store their bait herring or sardines. Sea lions were having an “all you can eat buffet,” says Schakner, then a graduate student at University of California, Los Angeles (UCLA), now a fisheries scientist with the National Oceanic and Atmospheric Administration. So he tried to condition them with an intense burst of sound cued by a neutral tone. It failed. Sea lions left only temporarily. California sea lions are also feasting on endangered salmon heading up to their runs on the Columbia River. Despite hazing with rubber bullets and loud shell crackers, the efforts failed. “Sea lions are too smart,” adds Schakner's UCLA collaborator Dan Blumstein, underlining a tricky conflict between one recently recovered species and another endangered one. Calling the situation a lose–lose conflict, he says, “It is a perfect example of the ridiculousness of modern conservation,” underlining that nonlethal conditioning is necessary but does not always work. One specialized type of AC that attempts to break animals of problematic food habits is called conditioned taste aversion (CTA), explains Maggie Watson, at Charles Sturt University, in Australia. During her graduate work at the University of Queensland, she investigated whether CTA could be used to teach sugar gliders, an introduced small possum, to avoid eating the eggs of endangered Australian parrots. Over one breeding season at two sites, Watson took quail eggs treated with the chemical thiabendazole and conducted a series of tests. Fish-loving seals and sea lions are a menace for fishers who use bait docks to store their live fish bait. These marine mammals have learned this is an all-you-can eat buffet. So aversive conditioning is being trialled. Photograph: Thomas Götz, University of St. Andrews, United Kingdom. CTA is like intentionally giving an individual food poisoning, Watson explains. “If you've ever been food poisoned, you have this uneasiness about [again eating] whatever it was.” In CTA, researchers often use an emetic or other vomit-inducing drug to teach animals to avoid specific foods. In this case, CTA worked. Sugar gliders subjected to CTA learned to avoid eating parrot eggs. The advantage of CTA is that, in principle, the consequences are immediately so severe that, via it has CTA is says Watson, but for it adds to the conservation St. wildlife to aversive of St. frequency of aversive conditioning wariness in habituated Janik management of carnivore with a startle on a fish The of hazing lions as a conflict mitigation for carnivore management. Snijders Appleby St. taste aversion as a tool for human–wildlife in Another use of CTA the of researchers and wildlife managers is endangered to avoid the As the of has across the about the of a as they to eat this on work by at the University of at the University of to see if the could be taught to made adding an emetic to but not in learned to avoid eating the after being conditioned with But the challenge came with in a area of whether via their could learn the in the the wild CTA had a low success other researchers some success with this As for to the conflicts created by hyenas in Berger-Tal used CTA He tried to condition hyenas by them human food in feeding in urban areas. The were in in To up but eating in theory, CTA work, says Berger-Tal, in many it in the CTA and AC are says Berger-Tal, but not there AC, as to human–wildlife conflict, is still in a but researchers are to knowledge so St. that one advantage of AC over other methods for human–wildlife conflicts is that it can sometimes be used potential and animals of from and back to the could teach researchers and practitioners a St. is the of between conditioned and stimulus. research that can learn to associate sound with pain and taste with but they cross she explains. So animals may not be to associate a food from garbage with the punishment of a rubber use of this technique when problem bears are and a in The of continues to across Australia. such as this male have as they to eat this Photograph: the other of animals are more likely to learn to avoid a conditioned if their is an a low punishment that may over The of The of negative consequences be In other punishment not be in St. team made this when elk. tried to chase them off an endangered by them daily with using the and Not the after conditioning In Israel, there is public about hyenas into town and to eating So aversive conditioning to this behavior has been Photograph: Oded Berger-Tal. In the consequences be and modern deterrents to on but can only in specific making it for wildlife to over if punishment is only also be for to the one with conflict. for good behavior is the mainstay of modern training and That it is challenging to individual When St. and Rob in Rocky Mountain they that and to AC. They of individual can effective labor-intensive AC, Watson, like animals to which has and consequences as well as the immediate and conflicts AC is an important tool in the conservation toolbox. humans to around wildlife is another of conflict. more to be by animals and humans, AC is to be an area of research and management for to is a and based in Canada, where she about the natural on
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,001 | 0,004 |
| Méta-épidémiologie (sens strict) | 0,000 | 0,000 |
| Méta-épidémiologie (sens large) | 0,000 | 0,000 |
| Bibliométrie | 0,000 | 0,000 |
| Études des sciences et des technologies | 0,001 | 0,002 |
| Communication savante | 0,001 | 0,002 |
| Science ouverte | 0,001 | 0,002 |
| Intégrité de la recherche | 0,001 | 0,002 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,042 | 0,003 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».