This thought-provoking essay left me wondering if it isn't time to stop using the mantra: "There has never been a documented case of a healthy wild wolf, etc. etc." In fact, wolf advocates should perhaps prepare for the possibility that, as wolf populations expand and increasingly interact with human communities, cases of unprovoked wolf aggression will occur. People are killed and injured by bears and mountain lions in North America on a regular basis and not much is made of it. But imagine, if you can, the uproar that will occur if a "healthy wild wolf" ever actually causes a human fatality. Another thought: Many are the wolf advocates who dream of seeing a wolf in the wild, and if that ever happens to you, you'd be doing the wolf a favor by admiring it for a moment -- and then throwing a rock at it. It's called negative conditioning, and probably the best form of wolf advocacy you can perform -- JM
After gray wolves (Canis lupus) were extirpated over a large portion of their North American range during the early 1900s, researchers reviewed the history of wolf-human encounters and concluded that wild, free-ranging wolves posed little or no threat to human safety. However, documented cases of wolf aggression toward people have recently increased, indicating a need for further examination of wolf-human interactions. I reviewed 80 cases of wolf-human encounters and compared behaviors of wild wolves that interacted with people in different contexts in Alaska and Canada. Only 1 case of unprovoked wolf aggression was documented between 1900 and 1969, but 18 cases of unprovoked wolf aggression toward people occurred between 1969 and 2000, including 3 cases of serious injury to children since 1996. Increases in wolf protection, human activities in wolf habitat, and wolf numbers occurred concurrently with increases in unprovoked aggressive encounters. Aggressive behavior was documented in all regions and among all wolf subspecies in Alaska and Canada. Wolves rarely vocalized during unprovoked aggressive encounters, but wolves that were defending dens consistently displayed loud vocalizations. Behavior of rabid wolves was variable and ranged from stubborn, persistent approaches to prolonged attacks. Habituation contributed to unprovoked wolf aggression toward people in 11 cases; nonhabituated wolves in remote areas displayed unprovoked aggression in 7 cases. Where wolves are protected and frequently encounter people, some level of negative conditioning should be applied to prevent habituated and food-conditioned behavior in wolves.
Prior to European settlement, the gray wolf (Canis lupus) was the most widely distributed mammal in North America, but late in the nineteenth century wolf numbers declined as the expanding human population reduced the wolf's natural prey and poisoning campaigns were conducted to eliminate wolves from livestock ranges....
In Alaska and northern Canada, bounties and indiscriminate use of poisons reduced wolf numbers in the early 1900s and then again in the 1950s when government-sponsored wolf control was implemented to reduce wolf predation on big-game prey species and to reduce the spread of rabies. Although wolf populations began a recovery in western Canada in the 1930s, the effectiveness of the 1950s poisoning campaigns reversed population recovery. The use of poisons was largely discontinued by the mid-1960s, and all jurisdictions of Canada and Alaska subsequently classified wolves as furbearers or big-game animals, thereby providing some regulatory protection for remaining populations. During the last 30 years wolves have reoccupied forested areas of southwestern Canada and their numbers have increased in Alaska. Recent estimates indicate 7,000-10,000 wolves in Alaska, an additional 52,000-60,000 across Canada, and approximately 3,200 in the United States south of Canada, primarily in northern Minnesota.
In answer to public concerns about the potential for aggression by wolves toward people, biologists commonly reply, "There's never been a documented case of a healthy wild wolf killing or seriously injuring a person in North America." The statement accurately reflected wolf-human interactions throughout most of the twentieth century, when wolves were virtually extirpated in the 48 states and southern Canada and aggressively controlled in northern Canada and Alaska. Nevertheless, a fear of wolves lingered among the public as both a vestige of European medieval perception and as a consequence of experiences with wolf depredation on livestock. Those negative attitudes impeded reintroduction and natural recolonization of wolves into suitable habitats, but public attitudes have recently changed. In Alaska, wildlife viewing is increasingly important and wolves are among the most desired species for viewing. Interacting with wolves by howling has become a popular activity in several national and provincial parks, and total protection is now afforded wolves in many parks where wolves were previously controlled.
Some evidence exists that before settlement, wolves were less fearful of and occasionally preyed upon Native North Americans. Young (1944) suggested that a transition to a more "timid, wary, and cautious" wolf began in the 1870s, coincident with widespread efforts to extirpate wolves in both Canada and the United States. Whether human exploitation increased the wariness of surviving animals on a large scale is difficult to prove, but McCullough (1982) and Swenson (1999) suggested that hunting of brown bears (Ursus arctos) fostered wariness toward humans, and conversely that protection of bears fostered less wary behavior. A similar response was noted among wolves in remote versus settled areas of Labrador in the 1970s and 1980s. In remote areas wolves appeared naive, occasionally approaching field biologists, but wolves were fearful and wary near settlements where hunters commonly pursued wolves on snowmobiles. Conceivably, such behavioral changes could arise quickly from learned responses or over time from the culling of individuals that possess genetic predispositions for highly inquisitive or aggressive behavior.
Instances of threat behavior by wolves toward people before 1990 were published as examples of unusual behavior that were inconsistent with previous patterns. More recently, wolves inflicted serious injury on 3 children and a young man in separate attacks in Alaska and Canada, and an increase in non-aggressive but fearless behavior was reported among wolves in widely separated national parks and remote industrial sites. Given the changing contexts of wolf-human encounters, in which people are increasingly interested in viewing or interacting with protected wolf populations, it is important to evaluate the potential for wolf behavioral responses that are different than those observed among the highly exploited wolf populations of the past....
....Habituation is a decreasing response to a repeated, nonconsequential stimulus, and although habituation can refer to the loss of any response, the term is most commonly applied to the loss of an animal's fear response to people arising from frequent nonconsequential encounters.
Unlike habituation, conditioning is a form of associative learning that involves development of new behavioral responses to specific stimuli. In the context of wolf-human interactions, food conditioning occurs when wolves associate food with the presence of people. To extinguish a food-conditioned response, managers sometimes apply aversive conditioning, a form of negative conditioning in which a formerly attractive stimulus is paired with a negative reinforcement .... I used the expression "food-conditioned approach" to describe behavior of wolves that were conditioned by food handouts and consequently actively sought out and approached people .... I used the term fearless behavior as a general description for wolves that exhibited a diminished avoidance response in proximity to people. Fearless animals may have been habituated, food-conditioned, diseased, or may have exhibited fearless behavior because they were naive or misidentified people as prey.
Aggression in healthy wolves is generally associated with either agonism or predation. Agonism (i.e. agonistic behavior) arises from a conflict between aggression and fear and includes most aggressive behaviors related to wolf social interactions. Predation involves a series of connected behaviors that begins with orientation toward the prey followed by stalking, chasing, and catching the prey....
Analysis of behavioral patterns and trends
My analysis was based on 80 cases of fearless behavior in wolves compiled by McNay (2002) from published accounts, public health records, and a systematic query of biologists in Alaska and Canada. All cases occurred in Alaska or Canada except for 3 cases reported from Minnesota near the Canadian border. McNay (2002) classified the cases into 7 categories based on wolf behavioral patterns and context. Five categories (agonism, predation, prey testing, self-defense, and rabies) contained encounters in which wolves acted aggressively toward people (i.e., threats, charges, or bites). Two categories (investigative search and investigative approach) contained cases of non-aggressive behavior in which fearless wolves either ignored people or were inquisitive but nonthreatening. Most aggressive encounters resulted from self-defense, defense of conspecifics, or rabies, or were triggered by the presence of a domestic dog (Canis familiaris). I classified those behaviors as provoked aggression and excluded them from my analysis of aggression between habituated and nonhabituated wolves. I classified the remaining cases of aggression as unprovoked; they included agonistic or predatory behaviors directed toward people who were acting passively when approached by the wolf.
I considered individual wolves to be habituated if they repeatedly approached people or repeatedly visited human use areas in the presence of people without displaying a fear response. I considered wolves to be nonhabituated if the encounter occurred in a remote setting where previous contact with humans had not been documented and where there existed little opportunity for previous human contact. Where wolves inflicted bites, I classified the bite as minor if the wolf bit into the person's clothing or inflicted an abrasion, bruise, or single puncture wound in the victim's skin. Bites were classed as severe if they resulted in multiple lacerations or puncture wounds....
Wilson et al. (2000) suggested that the eastern Canadian wolf (C.l. lycaon) and coyote (Canis latrans) evolved in North America from a common ancestor, and Wayne et al. (1995) found that wolves of southeastern Canada and Minnesota were more likely to hybridize with the coyote. Carbyn (1989) documented highly aggressive behavior by coyotes toward children, more aggressive than suggested by the case history for wolves. To evaluate whether aggression among the C..l. lycaon subspecies was higher than among other subspecies, I grouped unprovoked aggressive encounters by wolf subspecies according to Nowak's (1995) classification....
Aggression by wolves toward people was evident in 51 of the 80 cases of wolf-human encounters summarized by McNay (2002). Twelve of these cases involved wolves known or suspected to be infected with rabies, 14 involved wolves that acted aggressively in self-defense or in defense of conspecifics, and 6 involved wolves that were aggressive toward people accompanied by dogs. The remaining 19 cases of aggression were considered to be unprovoked and included threat displays, charges, or bites associated with agonism or predation.
....After 1969 unprovoked aggression was documented for 11 habituated and 7 nonhabituated wolves. The incidence of biting was more common among habituated versus nonhabituated wolves. Bites were inflicted in all 11 cases where habituated wolves displayed unprovoked aggressive behavior but in only 2 of 7 cases where nonhabituated wolves exhibited unprovoked aggression. Bites by nonhabituated wolves were minor, but 4 of the 11 bites inflicted by habituated wolves were severe. All 11 biting incidents by habituated wolves occurred in parks with high levels of human use or industrial sites; all 7 of the nonhabituated aggressive encounters occurred in remote settings.
....Six of the 11 aggressive, habituated wolves were known to be food-conditioned; 5 of those displayed food-conditioned approach behavior. The remaining aggressive, habituated wolves may also have been food-conditioned, but food-conditioning was not specifically documented in the case description. It was unlikely that any of the aggressive, nonhabituated wolves were food-conditioned because all of those cases occurred at remote locations well distanced from human use areas.
Unprovoked aggressive behavior was documented in all taxonomic regions of Alaska and Canada. The 4 cases that resulted in severe bites were distributed among 3 subspecies: 1 in Alaska, 1 in coastal British Columbia, and 2 in Ontario. Human encounters with rabid wolves were primarily reported from northern areas where wolves were sympatric with arctic foxes (Alopex lagopus) ... but 2 encounters with rabid wolves occurred in Ontario. Cases of aggressive behavior by wolves defending dens or rendezvous sites were widely distributed....
Trends in the frequency of aggressive wolf-human encounters
There was only 1 case of unprovoked aggression by a wolf prior to 1969, an apparently predatory attack on a small child ca. 1900 in a remote area of Alaska (Huntington 1993). The lack of documented aggressive encounters between 1900 and 1969 is consistent with previous reviews. At the end of that era, Pimlott (1967) testified to the saefty of wolves using the 60-year, incident-free history of Algonquin Park as an example. Ironically, in recent years more wolf aggression toward people occurred in Algonquin Park than in any other single site in North America, with 5 people bitten, 2 seriously, since 1987. One significant difference between the Algonquin Park that Pimlott knew and the Algonquin Park of today is related to management policies within the park. Beginning about 1909, rangers routinely controlled wolves in Algonquin Park, and from 1944 to 1958 they killed an average of 50 wolves per year within the park; an additional 106 wolves were killed in 1964 and 1965. Wolves have since been protected within Algonquin Park but continue to be harvested in surrounding areas.
The increase in aggressive wolf-human encounters after 1970 occurred as increasing, lightly exploited wolf populations converged with increasing human activity in wolf habitat. The average increase (62%) in visitor use among 3 of Canada's most popular wolf-occupied parks (Algonquin, Banff, and Pacific Rim) between 1975 and 1998 was twice the concurrent increase in Canada's human population .... Until 1996, all of the cases of unprovoked aggression involved low-level threats or minor bites, but between 1996 and 2000, 4 of the 5 cases resulted in serious human injury. Whether those separate incidents were coincidence or whether they indicated an evolving pattern in the level of aggression among habituated wolves was not clear. However, to the extent that wolf populations and human activities in wolf habitat continue to increase, there is the potential for a concurrent increase in the number of aggressive wolf-human encounters.
Behaviors Among nonhabituated fearless wolves
Case histories suggested that wild, nonhabituated wolves posed little threat to human safety. Nonhabituated wolves occasionally acted aggressively, but only 7 cases of unprovoked aggression occurred between 1969 and 2001. Considering that there are approximately 70,000 nonhabituated wolves in Alaska and Canada, one must conclude that the threat of unprovoked aggression from nonhabituated wolves is very small. Rapid, apparently investigative approaches by nonhabituated wolves often culminated in quick retreats once the wolves identified humans as the object of their investigation. Yet in 5 cases of unprovoked aggression, nonhabituated wolves charged or leaped toward people .... When analyzing those incidents, it was tempting to ascribe predatory motivation. For example, in the Ellsmere Island case, Munthe and Hutchison (1978) speculated that the appearance and behavior of the people were sufficiently unfamiliar to inhibit a full predatory attack by the wolves....
Only 2 minor bites were inflicted by healthy, nonhabituated wolves compared to 11 bites recorded among habituated wolves, but the lower frequency of bites among nonhabituated wolves was in part the result of self-defense by the victims. In 3 of the 5 cases where nonhabituated wolves acted aggressively but did not bite, 1 wolf was shot and killed at close range, another was struck with a rifle butt, and a third wolf knocked a young man to the ground but ran when the man fired his rifle into the air. In the 2 cases where [nonhabituated] wolves inflicted minor bites, 1 was killed with a pistol as it was biting the man's leg and the other was knocked unconscious with a heavy object .... In contrast, among the 11 cases where habituated wolves bit people, none of the people defended themselves with more than their voices, hands, or arms.
Nonhabituated wolves acted aggressively in self-defense, but that aggression was often inhibited. Wolves ran or trotted toward people in 6 of 10 cases that describe defense of dens or conspecifics, sometimes coming within a few meters before veering away. In all cases where wolves defended dens or rendezvous sites, aggressive behavior was accompanied by loud barking and howling, but no bites were inflicted. Loud vocalizations were not reported in any other context of aggressive behavior.
Uninhibited aggression was displayed in 5 cases by dominant wolves (4 alpha females, 1 alpha male) that were defending themselves during wolf-capture operations .... Presumably, a dominant animal sufficiently threatened in any context would act aggressively in self-defense ....
Attacks on dogs are among the most commonly reported conflicts between wolves and humans .... When dogs are present they can act as the primary stimulus for wolf aggression, but once a wolf is in a heightened state of excitation, its tendency to act aggressively toward people may be increased. Therefore, in any context where wolves exist, people with dogs may provide an impetus for aggressive behavior in wolves that would otherwise avoid humans.
Nonhabituated wolves sometimes exhibited fearless, investigative behaviors that were not aggressive .... In many of those cases fearless behavior was probably the result of naivete, and witnesses described the wolves as cautious and curious....
Rabies is enzootic in wolves at northern latitudes where wolves are sympatric with arctic foxes. Rabies has also been reported in wolves from southern Ontario, where the arctic fox strain of rabies is enzootic in red foxes (Vulpes vulpes). In those areas rabies must be considered a possible cause of fearless behavior in wolves, especially among nonhabituated wolves that exhibit aggressive or other abnormal behavior. Typically, rabies symptoms include partial paralysis, lack of coordination, or various levels of aggression. Additional symptoms included a glazed stare, a stubborn undeterred approach, frequent shifting of the wolf's aggressive behavior from one stimulus to another, a staggering gait, obvious wounds and injuries, porcupine quills in the mouth or face, lack of a reflex response, and biting the ground or other inanimate objects. During a rabies outbreak among foxes in Cold Bay, Alaska, foxes often darted out and snapped at people walking by. That behavior was more common than were furious, unrelenting attacks. Similar variation in aggressiveness was documented among rabid wolves, which may be related to the progression of the disease in a given animal, with the least aggressive behaviors occurring late in the course of the disease, just before death....
Behaviors among habituated wolves
Habituation without food conditioning appeared to be the operative process for fearless behavior among some wolves in parks, industrial sites, and near remote den sites. Often those wolves developed a pattern of camp robbing (i.e., stealing articles of clothing or camping gear). Camp robbing was a common theme scattered among historical accounts of wolf-human interactions as well as recent coyote-human interactions, suggesting that novelty or "chewability" of those items alone might be enough to encourage campground search behavior without food conditioning. However, food conditioning was known or suspected in 16 cases of habituated behavior from parks and industrial sites. It appeared that crumbs, residual food in pans, or even food odors associated with campsites contributed to a food-conditioned search response in wolves, although conditioned behavior is a complex process resulting from associations with a variety of ambient stimuli present in the learning context, and conditioned behavior can be generalized to new contexts if at least some similar conditioned stimuli exist in the new context. Therefore, when wolves develop a food-conditioned response in one location, they may generalize that behavior to other locations where they recognize a similar array of conditioned stimuli (e.g., human scent, tents, picnic tables, camping gear, or vehicles). That pattern was observed in Algonguin, Banff, and Denali parks where food or novelty conditioned behaviors were first reported from a single location, then subsequently observed in the same individual or group of wolves at other sites, sometimes several kilometers from the original learning context.
....When they receive food directly from people (i.e., handouts), wolves can quickly develop persistent, aggressive approach behavior. During construction of the trans-Alaska pipeline, wolves crouched along roadsides waiting handouts by passing truckers and slept under buildings in a camp where they obtained food. That behavior enhanced the risk of agonistic bites because wolves conditioned to receive a food reward directly from people exaggerated the conditioned response, leaping toward the person and sometimes biting if the reward was withheld .... Those "extinction bursts" can occur as part of the extinction process of conditioned behavior and typically involve an increase in the number and vigor of responses shortly after the reward is withheld.
Wolves occasionally followed, mimicked, or paced alongside people without exhibiting signs of aggression. Witnesses described the behavior as "escorting." .... Observers often described such behavior as playful and concluded the wolf was enjoying its contact with people. However, as habituation progressed, those and other descriptions of playful, cautious, or curious behavior were sometimes replaced by descriptions of wolves that...failed to respond to discouragement of undesirable behaviors. That entire behavioral pattern preceded the 1998 predatory attack on a 19-month-old child in Algonquin Park, and curious, increasingly fearless, but non-aggressive behavior was also reported before a wolf made a predatory attack on a 6-year-old boy near Icy Bay, Alaska in April 2000.
The transition from non-aggressive habituated behavior to strong aggression or predatory behavior can occur quickly, as appears to have been the case at both Icy Bay and Algonquin Park. The change may have been triggered by behavioral stimuli from the potential victim .... Wolves have a keen ability to evaluate prey vulnerability, and therefore children are more likely to be victims of a predatory attack than adults. The 3 cases that clearly contained elements of predation all involved small children. In each case the wolf bit into the child and either carried or tossed the child with a shake of its head. Although few such cases exist in North America, that type of aggressive behavior probably represents the most serious threat by wolves to people, as reported for wolves (C.l. pallipes) in some areas of India.
....Inherent social characteristics of wolves are also important in shaping wolf-human interactions. Agonistic behaviors are the most diverse and numerous in the wolf's behavioral repertoire, and captive wolves sometimes direct agonistic behaviors toward humans. Zimen (1981) concluded that in social interactions a wolf's inhibition to bite depended upon the wolf's expectation that its victim would vigorously defend itself and that agonistic aggression was shaped by the balance of impelling and inhibiting stimuli within the wolf's social environment. It follows that as wolves become progressively habituated by nonconsequential human encounters, biting inhibitions can degrade, and at some point the impelling stimuli exceed the inhibition threshold and the wolf becomes a potential biter. That mechanism may explain why wolves often exhibit non-aggressive habituated behavior for long periods before they bite and why they rarely inflict agonistic bites during their initial encounters with people. Enforcing some threat of reprisal would probably inhibit both agonistic and predatory responses toward people.
Habituation has been proposed to increase the safety of human-bear encounters because defensive charges are less likely when a bear's fear response is reduced....However, others have argued that fearless bears, because of their inherent aggressive nature, are potentially dangerous in any context, and black and grizzly bears were responsible for 63 serious or fatal injuries to people in British Columbia alone between 1970 and 1997. During roughly the same period (i.e., 1969-2001) wolves inflicted only 4 serious injuries continent-wide. Obviously the threat of wolf aggression toward people is small when compared to the potential threat posed by bears.
Sudden encounters are the most common source of serious attacks on people by brown bears. Those attacks arise from a fear impulse that triggers a defensive response. However, among the 27 aggressive wolf encounters I documented that involved healthy, nonhabituated wolves, none appeared to be defensive attacks from wolves that were startled at close range. Even in defense of their young or at kill sites, wolves demonstrated inhibited aggression that ended well short of an actual attack. Therefore, management models that advocate habituation to reduce bear-human encounters are inappropriate for wolves.
Habituation of wolves can lead to agonistic or predatory aggression toward humans with some risk of serious injury or death, particularly if the victim is a small child or infant. Food- or novelty-conditioned behavior can facilitate habituation and should be avoided. Aversive conditioning may be effective if the conditioned behavior is confined to a single site, but opportunistic aversive conditioning will likely be ineffective once a conditioned response is generalized to multiple contexts. Where wolf-human encounters are common, the need for aversive conditioning can be minimized if people consistently apply negative conditioning when they encounter wolves at close range. Shouting, clapping hands, banging pots, discharging noise devices, and throwing objects are examples of an appropriate level of negative conditioning.
When wolves act aggressively, their aggression is best deterred by an aggressive response; people should not run or lie down. When confronting an aggressive wolf, it may be necessary to strike the wolf with thrown or hand-held objects....However, if the wolf has rabies, it will likely be persistent, may show little response to an aggressive defense, and if driven away will often immediately return. Rabid wolves exhibit numerous signs of behavioral and physical abnormality in addition to aggression, and if symptoms suggest rabies, the animal should be killed and submitted for testing.
Injuries to humans inflicted by wolves are almost totally preventable because they are rare and the circumstances under which they occur are often predictable. Negative conditioning at close range as suggested here need not preclude opportunities for viewing wolves. However, where wolf viewing is a management objective, some control of human access and behavior may be necessary to minimize habituation in wolves.
Figure 1: Number of unprovoked aggressive wolf-human encounters recorded prior to 1969 and in 8-year time periods after 1969.
Figure 2: Number of unprovoked aggressive wolf-human encounters recorded for each of the 4 wolf subspecies in Alaska and Canada, 1969-2000.
Table 1: Number of cases in each of the 7 behavioral categories defined by McNay (2002) among 80 wolf-human encounters in Alaska and Canada, 1900-2001.
Table 2: Eighteen cases of unprovoked wolf aggression toward people in North America since 1969.