Wolf/Livestock Aversive Conditioning

July 3, 2000
Ed Bangs 406-449-5225, x 204

Ms. Andrea Lococo
Fund For Animals
P.O. Box 11294
Jackson, WY 83002

Dear Ms. Lococo:

Director Rappaport-Clark asked that I respond to your concerns and questions about the Service's attempts to investigate methods to reduce wolf conflicts with domestic livestock. I hope this letter will clarify what the aversive conditioning program is attempting to accomplish and why we consider this type of research as a potentially important tool for helping to achieve wolf recovery. I have asked Dr. John Shivik, the USDA Wildlife Services researcher in charge of the program to send you a copy of the research protocol. That protocol includes provisions to ensure humane treatment of all the animals involved.

Summary of the Research

This research program is a first step to see if wolves can "unlearn" that livestock can be prey. It is a cooperative effort by the U.S. Fish and Wildlife Service, USDA Wildlife Services, Turner Endangered Species Fund, University of Montana, and Defenders of Wildlife. The majority of support and actual care of the facility and wolves is being undertaken by the Turner Endangered Species Fund, a private organization, at the request of the involved federal agencies. The Sheep Mountain pack was removed from the wild because they repeatedly attacked cattle, primarily on private land. Given the pack's history of attacking cattle, that alternative methods to deter further depredations had already been used with limited success, and the lack of sites to relocate depredating wolves since most areas with native wild prey and without livestock are already occupied by other resident wolf packs, there were few options for the Sheep Mountain pack other than euthanasia.

The proposed study will use common dog training collars to teach the Sheep Mountain wolves that cattle should not be attacked. The collars emit a noise which is immediately followed by a small electric stimulus when the wolf comes to within a couple of feet of a domestic calf. The primary investigator for this project placed a training collar on himself and activated it. He reported that activation of the training collar caused him discomfort but not pain. The collar's electric stimulus is that used to train dogs and is in no stretch of the imagination "shock therapy" that could directly affect brain function or cause any physical injury.

I believe that attacking cattle is behavior that these wolves learned but even then is probably carried out with great caution. I imagine that wolves view cattle as large, formidable, and potentially dangerous prey. I anticipate that any unusual stimulus during initiation of a wolf attack on cattle would be enough to dispute that behavior. The tests will involve the 4 remaining members of the Sheep Mountain pack. They are being held in a large outdoor enclosure on the Flying D Ranch, near Bozeman, Montana. When a wolf approaches within 2 feet or so of a calf during training trials, a transmitter on the calf will activate that wolf's collar. Its collar will de-activate as soon as the wolf moves away from the calf. If the conditioning in the pen appears successful, the wolves will be released back into their old pack territory this fall. They will be fitted with regular radio collars when they are released and will be monitored closely. If they continue to attack cattle despite training they will be euthanized.

Wolf Behavior

Wolves are intelligent social animals that have unique behaviors that could make this type of training potentially useful in a few limited situations. The fact that all dogs directly originated from wolves, and that basic dog behavior is simply "domesticated" wolf behavior means wolves learn (and unlearn) how to identify and attack certain prey. An example of this behavior is in Yellowstone National Park where wolves are commonly associated with bison but only one pack has learned to successfully attack bison. Most wolves, even though they live near livestock, do not recognize them as prey. As an example, roughly 50-90 wolves have lived in northwestern Montana since about 1987. Nearly all are exposed to livestock on a regular if not daily basis. On average wolves killed about 6 cattle and 5 sheep a year, and only about 4 wolves per year were moved or killed to stop attacks on livestock. Livestock depredations by wolves are rare events despite almost unlimited opportunity for wolves to kill livestock. To many people the biggest mystery is; "Why don't wolves attack livestock more often?" Humans know that livestock are vulnerable and would be easy prey, but to a wolf, who is relatively unfamiliar with livestock, an adult cow defending its calf behaves like a potentially dangerous animal, such as a bison. Wolves rarely attack livestock because most were never taught to hunt and kill livestock by their packs. Those wolves that did attack livestock were usually quickly removed by people so other wolves had little opportunity to learn livestock can be prey.

Most wolves were raised by parents that hunted and ate wild native prey and therefore never learned that livestock can be hunted, killed, and eaten. Wolves, especially in packs, are neophobic (afraid of new things). That means that once they learn to act a certain way they are hesitant to try other ways of doing things. Wolves grow up in very tightly controlled family groups and are taught by their older pack mates (usually the alphas or breeding pair), where, when, and what to hunt. Research has indicated that if a pack hunts deer, for instance, they know how to locate and kill vulnerable deer. The pack "tradition" for hunting deer is passed from adults to pups for generations. Wolves may also be slow to learn to hunt other things they are not as familiar with, such as moose, if they haven't been taught to do so by other wolves. The survival benefit of that behavior is obvious. If you have been taught to find and recognize vulnerable deer and how to catch and kill deer by watching other pack members- Why should you try something different and risk getting severely injured or killed? It is common for wolves to be injured or killed while attacking large prey. Wolves need to be careful about what they attempt to kill and this is one reason that wolves tend to most often kill the very young, very old, or sick, weak, etc. However, research in Canada indicated that wolves that learn to attack livestock can quickly learn that livestock can be successfully attacked and those wolves will then increasingly spend more time hunting livestock. Other wolves quickly learn from the first "pioneers" and the behavior becomes more common among more wolves. If wolves are allowed to repeatedly attack livestock, they will pass that behavior on to other wolves and livestock losses caused by wolf attacks will increase.

Why training may work is some situations, especially with wolves

While there is no doubt that dog training with training collars is successful and results in long term learning, there are also several examples of how this basic canid learning behavior has been used to teach wolves and coyotes to not attack livestock. A study reportedly conducted in the old USSR, Republic of Georgia, indicated that reintroduced wolves were trained with dog training collars to avoid livestock before their release. Subsequent reports indicated the released wolves not only avoided livestock but they taught at least the next generation in the pack to avoid livestock. While I personally attended a lecture of this program, I have never seen a peer reviewed paper of the results. While the reports from this study appear promising, we can not simply assume the technique can always be as successful as reported.

Recently in Wisconsin, biologists captured an alpha female wolf after her pack repeatedly came onto a farm and killed livestock. They fitted her with a dog training collar that emitted a noise before an electric stimulus was administered. Her collar was triggered by a short-range transmitter on the ranch that had suffered repeated livestock depredations (reportedly totaling about $10,000). When she approached the farm the collar would activate causing her discomfort, which stopped as soon as she left that immediate area. After her release, data indicated she avoided that ranch, and apparently following her lead, so did the rest of her pack. There is no evidence that basic wolf survival or behavior was modified, other than avoidance of this particular ranch. After collaring and for the subsequent year, no livestock depredations were reported on that ranch. This past spring a calf was killed by her pack on that ranch. Biologists played the noise from the collar over a loudspeaker and just that noise (the training collar was only active a short time after her release because its battery expired) apparently caused her and her pack to leave the area, stopping any further livestock depredations. Again these data have just been collected and have not been tested further or published in a scientific journal. The apparent success of this one instance should be taken with caution and not applied to all instances wolf/livestock, but the results suggest that such "training" may be worthwhile in some circumstances.

Recently a peer-reviewed scientific paper was published in the Wildlife Society Bulletin (Andelt et al. 1999. Coyote predation on domestic sheep deterred with electronic dog-training collar. Wildlife Society Bulletin Vol 27, pages 12-18, the WSB is a peer reviewed publication the professional society of wildlife biologists). Researchers placed dog training collars on penned coyotes and activated the collar when the coyotes tried to attack sheep that were placed in the pen. The coyotes quickly learned to not attack sheep and that behavior stayed with them, without further training for several months. This gives hope that such conditioning may work on wolves, possibly for an extended period of time.

It is clearly evident to all of us involved in this project that this type of intensive management is of limited value in most situations. This type of research is humane and I believe better than the alternative (lethal control)in this instance. However, in areas where wolves are common, the expense and effort of this type of intensive management could not be justified. Lethal control is easy to carry out, cheaper, more effective, and when restricted to just the few problem animals will not significantly affect overall wolf population growth. The side-benefit of this type research is we could learn something that may lead to another form of non-lethal predator control technique that could be more generally applied. Most importantly this type of research may lead to other techniques of particular benefit in management of very small populations of highly endangered wildlife. I believe we owe it to our wildlife resources to be open minded and to try new ways to resolve conflicts with an ever expanding human population. The Service will continue to investigate a host of other non-lethal wolf and livestock management strategies to reduce the level and severity of wolf/livestock conflict and the number of wolves that are killed.

History of the Sheep Mountain Pack

The Sheep Mountain pack formed just north of Gardiner, Montana in 1997. A female (#16) bred with the alpha male of the Chief Joseph pack. The first year the Sheep Mountain pack (the lone female #16 and her pups) was actually named Chief Joseph 2. #16 was struck and severely injured by a vehicle that summer but she eventually recovered. Fortunately at least some of her pups were apparently adopted by the father's Chief Joe pack. In 1998 female #16 gave birth to another litter of pups and several other dispersing wolves joined the pack. The pack was then named Sheep Mountain. The pack has lived in the same territory since 1997, which includes a mix of private and public land, numerous elk and mule deer, and at certain times of the year many widely dispersed cattle. During the fall and winter, the Sheep Mountain pack's territory contains one of the highest densities of elk in the Greater Yellowstone area. Livestock are grazed primarily on private land but during the summer many cattle are grazed on adjacent USDA Forest Service allotments.

In spring 1999, this pack, which had not attacked cattle before, established a rendezvous site (a place where pups are kept) on a section of private land that was grazed by cattle in the summer. The livestock producer held his livestock off of that pasture, for nearly 2 weeks, while he worked with Defenders of Wildlife to try to find alternative pasture. Unfortunately none was found and the cattle were put on the pasture they had grazed historically. Defenders did pay to have a rider visit the cattle daily on this and a neighboring Forest Service allotment for several weeks. In addition, the Service increased aerial monitoring of the wolves. After the pups were weaned, Service biologists went to the rendezvous site and set up a light siren device. They also physically approached the wolves to try to cause them to move their pups and the high level of adult activity associated with raising pups, from that immediate area. The wolves did move the pups about a mile away, to USDA Forest Service land that was not being used by livestock. However, late that summer the wolves killed a calf, and then others. On both occasions telemetry monitoring indicated the alpha female was not involved, but several other pack members and the alpha male were at both depredation sites. As a result 3 subadults and the adult male from the Sheep Mountain pack were killed. Shortly thereafter, pack members killed another calf on a USDA Forest Service allotment and 2 more subadult pack members were killed. No further depredations were documented although a few local livestock producers reported higher numbers of missing livestock than normal during fall roundup. The pack of 7 remaining members continued to use their normal territory but did not appear to continue hunting livestock.

However, in early 2000, Sheep Mountain wolves attacked 2 yearling cattle on private land. Both were severely wounded and one died within days. Because the wolves had not fed on the cattle (gotten a food reward) and because the natural migration of elk into the area was beginning, no control was conducted. This spring the alpha female did not produce a litter of pups and the pack continued to move widely throughout their territory. In May the Sheep Mountain pack killed a yearling steer on private land. The Service made the decision to remove the entire pack and try to condition them to stop killing cattle. Five pack members were darted but unfortunately one accidently died during capture. Four pack members, including the alpha female #16, were put in a large enclosure on a private ranch near Bozeman, Montana. The Turner Endangered Species Fund offered its wildlife veterinarian and a biologist to assist with animal care needs, and the Flying D Ranch, owned by Mr. Ted Turner, offered a secure place and funding to hold the wolves. The Turner Endangered Species Fund uses its resources to look for opportunities to assist in the recovery of endangered species and increase wildlife diversity, particularly on private lands. The remaining radio-collared Sheep Mountain wolf was monitored but the lone remaining uncollared pack member apparently dispersed and was not seen again. After 2 unsuccessful attempts to catch the lone remaining collared pack member it was killed so that it would not be able to re-teach the conditioned wolves that livestock can be prey.

If the training appears successful, the Sheep Mountain wolves will be released back in their home territory this fall. In the fall large numbers of elk disperse into the Sheep Mountain pack territory, providing an abundant supply of wild prey. By fall livestock are removed from public land allotments and calves (the type of cattle most vulnerable to wolves) have been shipped to market. In winter cattle are concentrated on private land and if the pack started to hunt livestock again, that behavior could be detected quickly. Most importantly, by releasing the alpha female in fall, she should have ample time to attract another mate and breed during the February wolf breeding season.

Why the Sheep Mountain pack was selected for study.

In the case of the Sheep Mountain pack, this type of management may be uniquely beneficial. The Sheep Mountain territory will always be highly attractive to wolves because it has one of the highest concentrations of elk in the Yellowstone area during the winter and spring. Dispersing wolves will continually try to colonize that area if it doesn't already have a resident wolf pack defending that space. Because of the large amount of private land, this area will always have some livestock. If this entire pack was permanently removed, other wolves would soon disperse into the area and try to form a new pack. If some innovative way is not found to prevent or teach wolves in that territory that livestock should not be regularly hunted, then we may get into a chronic problem of having wolves move in, kill livestock, and then be removed. That pattern could repeat itself again and again. I hope that this type of research may help reduce the chances of a cycle where livestock and wolves are in constant conflict resulting in the death of many livestock and wolves and a highly frustrated public.

If this pack could be taught the "tradition" of killing just wild prey and not attacking livestock, any new wolves joining the pack might also adopt that hunting behavior. A Sheep Mountain wolf pack could then occupy the area with minimal conflict with livestock and local people. At this point in time that is wishful thinking but I believe this latest effort is certainly worth exploring. I believe it is better than simply removing wolves time after time or expecting that local residents will give up their livelihood and lifestyle. In addition to this research the Service will continue to work with local livestock producers to resolve potential and actual wolf/livestock conflicts through a wide variety of techniques that have already been used, such as guard animals, extra herders/riders, good animal health practices and livestock carcass disposal, fencing, scare devices, additional monitoring including providing telemetry receivers to ranchers to monitor wolves near their livestock, harassment of wolves, maintaining abundant sources of wild prey, relocation of problem wolves, and lethal control of chronic problem wolves.

Thank you for your interest in wolves and the Service's wolf recovery program in Montana, Idaho, and Wyoming. For further information about the overall program please see http://mountain-prairie.fws.gov/wolf/annualrpt99/. Any other ideas you may have to resolve these types of complex natural resource issues would also be appreciated. I hope you will share this letter with members of your organization so that they can make informed decisions about their feelings about this type of program. Please contact me if you have additional questions or concerns.

Sincerely,

Ed Bangs

Wolf Recovery Coordinator

Comment on this topic here

webmaster?subject=Wolf%20Livestock%20Aversive%20Conditioning%20Page">