Interdog aggression: What are the warning signs?

Editor's note In the last of a two-part series,Dr. Overall outlines her recommendations for an interdog attack that ledto the death of a dog. The first part of the series was published in theFebruary issue, p. 8S.

People who show and train dogs may know an awful lot more about dog behaviorthan does the average person; however, such exposure also may engender aspecific form of "blinders".

While dogs may occasionally have agonistic interactions, repeated behavioras exhibited by the Corgi to the Husky is not normal. It's problematic andabnormal. That it happened repeatedly signals a problem, not a contest aboutwho is going to be "alpha". The "alpha" concept is anoutdated one with almost no data to support it. Even if there were an absoluterank, rank is not taken - it's conferred by others.

People who breed, train or show dogs often tolerate, and perhaps evenaccidentally encourage, truly aggressive behavior between dogs because theythink it is normal when it is not.

Of the three clients I have had whose dogs have killed one of their group,all had similar pedigrees and histories as this client.

It's not normal

Here are some points to consider about this case.

* Dogs that disagree with each other usually bark, growl or snap,and almost never make contact. When they do, there is usually a nick onan ear or the top of the head or shoulders.

If the issue sparking the dispute is not going to be an ongoing problem,the dogs then solicit one another for play or grooming and they go slowlyso that they are understood. This is not the same as mere tolerance of eachother's presence.

* Hanging or biting on the neck are not normal dog to dog behaviors.They are behaviors that are associated with predatory events. For confirmationof this, review the location of the wounds on the Corgi.

* Fences are not sufficient to separate dogs that have a disputeif they can still see each other or otherwise continue to interact. Allfences do in such cases is to raise the stakes without penalty. The dogscan and do increase their threats to a level that would not otherwise occurif there were costs in terms of physical responses from the other dog.

Accordingly, in such situations, fences make inappropriately aggressivedogs worse, and better at being inappropriately aggressive, quickly.

* A herding dog may nip a running dog or a person, but the otherdog will signal it to stop. If not, its behavior is inappropriate and out-of-context.If it cannot do so, it is abnormal behavior. Herding behaviors can be redirectedto more appropriate foci such as toys, games or stock.

Behavioral signs

There are no truly "submissive" or "dominant/alpha"dogs and by putting these labels on dogs we blind ourselves to all of theinteresting information that the dogs are communicating with its species-typicalpostures.

Dogs that roll on their back are signaling that they are withdrawingfrom active, solicitous interaction.

If their limbs and tail are flaccid and their neck is fully exposed,they may invite/tolerate more passive interaction (e.g. sniffing, petting)from others.

If they tuck their tail and put their paws over their chest and groin,they do not wish to interact, period.

A normal dog recognizes this and withdraws, not because the first dog"submits" to them, but because they are capable of respondingappropriately to the signals. An abnormal dog recognizes an opportunityand moves in for the kill.

Client after client has reiterated exactly this scenario to me...thatthey thought their dogs had worked it out and one was "submitting"and that's when the serious "emotional" and physical injury occurred.

In this case, the Corgi exhibited inappropriate behavior without a resultingfatality time and again. To allow dogs to continue in such distressing socialsituations is to make the aggressor more confident, no matter how abnormalthey are, and to make the normal, deferential dog a victim.

Unfortunately, if you buy into the myths of "alpha" and submission,you would read the signals incorrectly and further reinforce these relativeaggressor/victim roles.

As the client alluded to, the higher-ranking dog seems to engender peace.High ranking dogs are those whose behaviors are appropriate given a numberof contexts. They feel the need to challenge no one because they do nothave to do so.

This is exactly what the hound mix had always done, but when pushed,he protected his companion from the dog behaving in an abnormal manner.Even perfectly normal dogs can be pushed too far.

Social responses

Normal social responses are the outcome of a complicated signaling danceinvolving deferential behaviors that help map routes where all needs canbest be met. This assumes normal behavior.

Note that in our case the outcome was not affected by age or size orstrength of the animals involved. The females involved in the disputes wereall evenly matched in size and age, and the male was very old. The onlytruism that we can rescue here is that, when damage is done, bigger, heavier,more heavily muscled dogs can do more damage.

· Activity levels must be considered. The client noted that theother dogs had begun to be concerned about the Corgi's behaviors and avoidedher. Listen to the dogs; they do a better job of interpreting abnormal dogbehavior than we do. After all, they speak fluent dog.

If the Husky and the Bull Terrier were beginning to be anxious aboutthe Corgi's behaviors, as it seems, would this anxiety be made worse orbetter if everyone is a ball of unexercised energy? Worse.

Would the anxiety be made worse or better by being turned out in an exuberantgroup? Worse.

Would the dogs have a harder or easier time reading each other's signalsin the dark? Harder.

Would signal translation difficulties make anxiety worse or better? Worse.

Taken logically, all clients and veterinarians can ask themselves thesetypes of questions and better assess the risks in any multi-dog household.

Prevention

What happened here is tragic but does not have to be the death knellfor the other dogs. They were not the problem...and the problem is sadlygone. Could this tragedy have been prevented? Maybe, but it's a lot of work.

Treating interdog aggression involves acknowledging and rejecting themyths discussed above.

More often than not, the aggression is not because of a determined upstartwho is challenging the status of some older or more "dominant"dog. As I have written before, it's because the abnormal dog will not toleratethe status, signaling and behavior changes that the normal dog undergoesas part of social maturity and wants that dog gone.

It's the aggressor who is inappropriate. This is a true case where thevictim has always been blamed. An article by Patricia McConnell in a recentAPDT newsletter also notes this: that the dog that people usually rewardso that they can "keep" their status is usually aggressive out-of-contextand without provocation, and that reinforcement of these behaviors onlymakes the situation worse, not better.

Caught early enough, the aggressor can be treated with behavior modificationdesigned to teach him or her to allow the other dog to exist.

Behavior modification can be as complex as desensitizing and counter-conditioningthe dogs to each other using treats and good head collars that can closethe dogs' mouths (e.g. Gentle Leader).

Or it can be as simple as letting the victim have free range during theday and locking the aggressor in a spare or lesser "status" room.

Completely separate

Aggression is based in anxiety and anti-anxiety medication has huge applicabilityhere for both the victim and the aggressor. The victim must be protected- at all costs - from the aggressor. The victim needs to have a sense ofcontrol and safety. This can be provided in part by putting a bell on theaggressor's collar so that the victim always knows where the aggressor is(Bear Bells: Silverfoot; Box 2090 Squamish, BC V0N 3G0, Canada; silverft@mountain-inter.net).Predictability lessens aggression. Dogs should be supervised and when thisisn't possible, kept completely separate.

If their relationship does not improve, what are the choices? You canplace one of the dogs. If you place the aggressor, it is safest if she orhe goes to an only-one-dog home. Note that in the case discussed here theproblem became more, not less, global with time and experience.

It is usually easier to place the victim, which no one wants to do. Ifthe clients decide to keep both dogs, and treatment with behavior modificationand medication has failed, those dogs will have to live completely separate.

This means separate rooms for sleeping, rotations on walks and in theyard, and separation by at least double gates - the canine equivalent ofa no-fly zone - for feeding and daily activities. The dogs cannot be permittedto threaten or otherwise endanger each other.

This situation has a zero-tolerance zone for mistakes, so doors musthave a working lock so that children cannot accidentally unlatch it. (Highlyplaced hooks and eyes on both sides of the door are ideal for this). Isthis hard? Of course, but if you are highly motivated you can do it in away that is safe and humane for all parties.

Finally, the client had two additional questions.

If the dog tastes blood, does this mean they will kill again? No, notunless the context of the taste was hunting to feed themselves.

Why were the females involved - are females more common in this formof interdog aggression? It's an interesting, important, and presently incompletelyanswered question. Females are over-represented in this form of aggressionin my patient population, regardless of hormonal status, which suggest thatthese data should be further examined.