This is an excellent article that explains predatory drift.

Due to a Mastiffs' rather bloody past, and their physical size, you have a responsibility to understand this phenomenon, and take steps to avoid this behaviour. If we don't take responsibility for our dogs, then we will soon find that Mastiffs will be on the list of banned breeds.

Copied with kind permission from Dogs In Canada magazine, and author Jean Donaldson.

Our dog, "Henry", killed a dog at the park a month ago.  We are still reeling from the shock and I can't begin to imagine how the owners of that dog are coping.  We have kept him on a leash and away from dogs ever since.  He is five years old and had gotten along with other dogs until this incident.  He had several playmates and my sister and dogsitter both have dogs with whom he has cohabitated peacefully for weeks at a time.  We called a trainer immediately and she explained that this was predatory behaviour.  This is puzzling to us because since we've had Henry he has never been much of a hunter - he doesn't chase cats, and has an excellent Recall on walks in the woods.  He is fairly large (65 pounds) and the dog he killed was small (11 pounds).  I can see how this would have made it easier for Henry to kill the dog, but not how it would contribute to his wanting to kill this dog as he has interacted peacefully with countless small dogs in his lifetime.  Could it still be hunting?  Will he ever be trustworthy again?

Given his history, I think this is an example of what Dr. Ian Dunbar has coined "predatory drift".  It worth summarizing what predation is before examining predatory drift.

Although often lumped under the banner "aggression", predation is food acquisition rather than agnostic (fisticuffs within a species) or defensive behaviour, although some of the behaviours in the canine predatory sequence - most notably biting - share some topography with aggression.  Predatory behaviour needs careful attention because the results are more often extremely damaging than the results of routine defence and competition.

Some dogs display frank predatory behaviour toward other dogs the same way they might toward squirrels, cats and other critters.  Most such dogs direct regular social behaviour toward dogs on other occasions, or confine their predation to small dogs or running dogs only.  Each dog will have a profile, broad or narrow, of the targeted dogs and contexts that elicit the behaviour.

There are also key divisions among these dogs regarding which parts of the predatory sequence they are predisposed to.  For instance, a dog may be a maniacal chaser but demonstrate great restraint if ever a prey item is caught.  (a Golden Retriever may gleefully run down a squirrel but have no idea what to do with it once it's caught - the squirrel is presumably freaked out but unscathed.)  Others are "finishers" - i.e. they finish the predatory sequence by killing what they catch.  Terriers are the poster children for this phenomenon, although there may be overrepresentation of other breeds, such as Siberian Huskies.  As an interesting side note, this incredible genetic elasticity in the dog's predatory sequence has allowed for the exaggeration through selective breeding of many of our favourite stylized predation behaviours, such as scentwork, pointing, flushing and herding.

Dogs that are known finishers are best managed (kept away from opportunity).  I would elect to manage dogs that are not known finishers if they target small dogs.  This usually manifests when the latter are running or scurrying (retired racer syndrome).  The risk of injury is too high, the behaviour harder than most to modify and the management usually easy to implement.

I believe Henry's unfortunate incident was the results of an equally serious but less well-known phenomenon, predatory drift.  Unlike regular predation, which is motivated as such from the get-go, predatory drift is the kicking in of predatory reflexes in an interaction that begins as a social interaction.  And, unlike predation, which is predictably elicited in a known quantity by a member of the target group, predatory drift can occur among dogs that had never been predatory before and may never be again.  It kicks in because of specific contextual triggers.  The riskiest contexts are:

●Play or a squabble between two dogs extremely different in size, especially if the smaller dog panics, yelps and/or struggles.  The simulation of a prey item is so convincing that the roles in interaction drift from a social scuffle to predator-prey.  The greater the size disparity, the greater the risk, for three reasons.  Firstly, the likelihood of the smaller dog getting inadvertently stepped on or otherwise ouched, even in a normal play session with a reasonably gentle dog, is greater if the dog in question is really tiny.  Secondly, I would speculate that the tinier the dog, the better the simulation of a prey item to the bigger dog.  Finally, the ease with which the larger dog can grab and shake the smaller one goes up as size difference increases.  Grabbing and shaking is often present in predatory-drift incidents.  Most of us have seen dogs grab and shake toys.  Even if non-lethal pressure is exerted, a grab and shake inflicted on a small dog can break its neck.

●Two or more dogs "teaming up" during intense play, or two or more dogs acting together in a chase or squabble context with a dog that begins to panic, yelp and/or struggle.  Dogs have also been known to attack injured dogs and this effect is also facilitated by the attacking unit being two or more dogs as opposed to one.

Because predatory drift can occur in dogs without any particular history, all owners and practitioners should exercise some diligence in the two contexts above - large to small interactions and 'double team' (two-plus one).  While it is quite true that many such interactions are completely benign, predatory drift is common enough that most dog people have either witnessed it or know someone who has.  The risks multiply if both factors are present.  Another obvious factor that increases risk is the involvement of a known predatory dog in the mix, especially a finisher.

If this rings true for Henry and depending on the recommendations of your trainer, it could be that Henry will be given the green light for continued play and socialization with large dogs but kept clear of small dogs from now on.