DOG AGGRESSION

What is the Basis for Aggression in Dogs?

Domestic dogs are carnivores; their close ancestors were efficient predators. The domestic dog's ancestors also used dominance and aggression to structure their societies. To be acceptable companions and work animals for humans, these instincts--predation, dominance and aggression--need to be very much reduced. In all likelihood, this process started as natural selection on dogs. Dogs that became camp-followers were able to feed on scraps, waste and discards but were probably only tolerated if they weren't a threat to their human associates. Most scientists think the evidence supports the idea that domestic dogs are derived from wolves, but, alternatively, it may be that domestic dogs are simply the evolutionary extension of a canid species that exists now only as the domestic dog (Koler-Matznick 2002). In any case, dogs and wolves can hybridize, indicating considerable genetic similarity between the two species. Wolves are not safe social companions; the amelioration of dominance and aggression in dogs, combined with the ability to the dog to integrate into a human social group using shared communication strategies (Hare et al. 2002) makes the domestic dog an appropriate household member.

As with other aspects of personality in dogs, dominance and aggression has fairly substantial heritability, indicating that selective breeding can enhance or reduce this behavior. It is unclear from studies on dogs whether predatory behavior is related to social dominance, but it, too, can be modified by selective breeding. Tests for aggression in dogs (Netto and Planta 1997; van den Berg et al.2003) allow dog breeders to rate dogs in their likelihood of expressing dominance, and when dogs with low dominance and agrression scores are bred, suitable companion animals result.

Certain breeds have been selected for enhanced dominance and aggression. Pit bulls and Rottweilers currently receive the most public attention in this regard, and pit bulls have been banned in many locations because they are perceived as being dangerous. While advocates of these breeds claim that maltreatment is a more likely underlying cause of the kind of aggression leading to biting incidents (some of which involve human fatalities), in fact we know that personality is fairly unresponsive to environment. Aggressive and dominant personalities likely only remain in check because dogs' owners have established themselves in a position of dominance over the animal, and other people are at risk, particularly when the owner is absent. Groups of dogs running loose can be particularly dangerous; human deaths often result from attacks by two to four dogs of a breed like Pit Bulls or Rottweilers which have been selected for dominance and aggression. A recent incident (November 2005), in which three pit bulls in Aurora, Colorado, mauled a 10-year old boy, illustrates this danger. The boy lost an arm and suffered disfiguring facial injuries (Meyer 2005). Not only is the attack of multiple dogs more difficult to fend off than an attack of a single dog, but there seems to be a catalytic effect among the dogs, fueling the attack.

Two analyses of human dog bite fatalities are available. A CDC report analyzed 199 deaths occuring between 1979 and 1996; the vast majority were attributed to Pit Bulls, Rottweilers, German Shepherds and wolf-dog hybrids (Anon, 1997). Sacks et al. (2000) documented 238 deaths between 1979 and 1998, and came to essentially the same conclusion. Breed-based analyses of bites, rather than fatalities, likely would yield a different picture. Small children are at greatest risk; they are most likely to be targeted by dogs who have been allowed to roam loose, and are also most likely to unknowingly wander into areas guarded or occupied by dangerous dogs. Children may be perceived by the dogs as prey, or as easy targets of dominance. Adult humans in rural areas have also been victims of attacks by groups of loose dogs. As Sacks et al. (2000) point out, it is difficult to correct these data for the relative abundances of animals of each breed, but the predominance of certain breeds in causing human fatalities is clear signal for caution with these breeds.

Although these few breeds are most likely to be involved in fatal bites, all dogs are capable of biting. Again, children are the most likely victims, as they are less likely to recognize an animal's boundaries (such as approaching an animal when it is sleeping, interfering with feeding, grabbing and pulling the dog in a way that causes pain to the animal). Clearly children and dogs must be well-supervised when together, particularly if the child is unfamiliar with dogs or the dog is unfamiliar with children. Understanding the role of dominance and aggression in dog social systems helps owners to develop strategies for integrating a new pet into their household which maintain the safety of the humans and happiness of the dog.

Puppies can be tested for aggression and dominance; they can be tested as well for other personality traits. Prospective dog owners are well-advised to do this, given the strong effect of genetics (as opposed to environment) in the expression of these behaviors in dogs.

Anon. 1997 Dog-Bite-Related Fatalities -- United States, 1995-1996. CDC Weekly Morbidity and Mortality Report 46(21);463-466
Hare B, Brown M, Williamson C, Tomasello M 2002 The domestication of social cognition in dogs SCIENCE 298 (5598): 1634-1636
Koler-Matznick J 2002 The origin of the dog revisited ANTHROZOOS 15 (2): 98-118
Meyer, J. 2005. http://www.denverpost.com/search/ci_3232494
Netto WJ, Planta DJU 1997 Behavioural testing for aggression in the domestic dog APPLIED ANIMAL BEHAVIOUR SCIENCE
52 (3-4): 243-263
Sacks, J. J., L. Sinclair, J. Gilchrist, G. C. Golab R. Lockwood 2000. Breeds of dogs involved in fatal human attacks in the United States between 1979 and 1998. JAVMA, Vol 217: 836-840
van den Berg L, Schilder MBH, Knol BW 2003 Behavior genetics of canine aggression: Behavioral phenotyping of golden retrievers by means of an aggression test BEHAVIOR GENETICS 33 (5): 469-483