Resource guarding can be a serious problem for some dogs. (Resource guarding is when a dog displays aggressive behavior when another dog or a person tries to take something away from the dog, such as a toy or bone.) This spring, several of my friends at UNT used Constructional Aggression Treatment (CAT) to successfully treat Rocky, a Labrador retriever who would display aggressive behavior (snapping, growling, biting) if another dog tried to take one of his toys or even approach him when he was chewing on a toy. The owner could not leave toys out when there were other dogs present and always had to be extremely cautious when Rocky was around other dogs.
What is Constructional Aggression Treatment (CAT)?
I’ve written a bit in the past about using CAT with fearful horses. Fearful behaviors and aggressive behaviors are often very similar–the animal behaves in some way to increase the distance between his self and the unwanted animal or object. So, an aggressive dog lunges or bites to increase the distance between it and something else, just as a fearful animal runs away to increase the distance between it and something else.
With Constructional Aggression Treatment, we give the animal what he wants (distance from something) whenever the animal displays any signs of friendly, relaxed, or calm behavior. We start small, accepting any small signs of better behavior and later on build up more and more friendly behavior.
Treating Resource Guarding using CAT
So, for Rocky, the trainer and another dog would approach Rocky while he was chewing on a tennis ball. As soon as Rocky showed any signs of more relaxed or friendly behavior, the trainer and other dog would leave, which is exactly what Rocky wanted. Before beginning the training, Rocky could use aggression (snapping, biting, etc.) to get the other dog to go away. Now, only friendly or relaxed behaviors would make the other dog go away.
Not only did Rocky figure out the new conditions, but he started trying to interact with the other dog and actually started to play with the other dog and share his toys!! This is what CAT calls “switch over,” where the animal switches from wanting something to go away to wanting interaction.
With a bit more training, Rocky’s new behavior quickly generalized to other dogs as well. Last I heard, Rocky is living with a new dog, who just joined the household. He will pick up a toy, take it to the new dog and ask him to play. Now that’s successful training, if you ask me! Check out the video below, which shows clips from the training process.
On Youtube: Constructional Aggression Treatment (CAT) for Resource Guarding
**Side note: Please be extremely careful when working with aggressive or fearful animals! Work at a level where the animal does not feel stressed or frustrated. Also, take precautions to insure that people and animals do not get injured. For instance, with an aggressive dog, keep the dog on a leash. Or, with a very fearful or aggressive horse, start by working with a fence between you and the horse. **
There is so much Wrong with the handling of this decoy dog, it makes me not able to watch the entire video. The fact that he is DRAGGED unwillingly out of the reactive dog’s face and allowed to practice completely inappropriate body cues towards the reactive dog is incredibly upsetting that this was considered proper “behavior modification protocols”. The least they could do is to treat the decoy dog with respect and positive methods to get him to disengage from the reactive dog and not let the decoy dog practice behaviors that are behaviorally detrimental to positive dog social interactions. I am incredibly disappointed that this was what graduate students studying dog behavior considered correct methods. Very disappointed.
Hi Katrin,
Thank you for commenting and sharing your opinion.
This is the first time this procedure has been used for resource guarding and the undergrads who did the project did a great job modifying the CAT procedure so that it could be used successfully to treat resource guarding. The final results were fantastic–a previously aggressive dog who now solicits interactions and shares his toys.
As with learning anything new, it’s always a learning process to try and figure out the best procedures. In the first several sessions, the little dog was pulled away to end the trials. I totally agree with you that this could have been set up better!
However, the guys weren’t happy with having to do this and modified their procedure in later sessions so that the decoy dog would be cued to leave willingly. Examples of this can be seen in later parts of the video. For example, at about 2:45, the black lab is cued to leave and leaves willingly with the trainer. Throughout the project the guys were continually trying to modify their procedure to make it safe, enjoyable and effective for all the dogs involved.
Since you are a dog trainer, I would be interested to hear what you think of the end behavior of Rocky (the dog who was the subject).
~Mary
I agree that the CAT did work very well for changing Rocky’s end behavior. CAT has so many applications, it is good to seeing it used in this area with success. I was rather taken aback that Rocky was not on lead with some form of back up restraint with the decoy dog so close to him in the early stages, as if something had gone wrong the decoy could have been hurt. I also probably would have tried for a more open space as opposed to a living room full of furniture as again had something gone wrong either dog could have easily been hurt and difficult to separate. I did like how the testers encouraged Rocky to share, by reinforcing him with getting the ball back when he did try to initiate play and sharing with the decoy dog. I think the basics of the behavior mod were good, there were just issues with the decoys that should have been addressed and ironed out long Before the decoy was ever actually brought into the session.
And I would not have been comfortable starting a behavior mod with a resource guarding dog with another dog who had as inappropriate body language as the 1st decoy had. When Rocky exploded on that dog in the “pre-behavior” video, that decoy had been incredibly inappropriate in his “play” style towards Rocky. And when that same dog was used as the initial decoy he was completely setting Rocky up for failure staring at him and straining at the end of a tight lead. Not my choice for a 1st decoy at all. I prefer to use a very neutral, relaxed dog as an initial decoy in a behavior mod protocol with a reactive dog so that the reactive dog is not set up for failure being more likely to react due to the decoy’s inappropriate signals. Those types of decoys come into the picture at a much later date once the reactive dog has a positive foundation of success with neutral dogs.
The owners of the decoy dog were against doing any training, that dog lives in a house with 4 college sorority girls that think the dog is perfect and did not even take our offers to do free obedience work. A neutral dog is ideal when working CAT almost always. However in this instance dogs could walk by Rocky without seeing displays of resource guarding. The neutral dog will stand there as Rocky would lay there. When we got the first decoy he displayed the behaviors that would evoke the resource guarding because he would focus in on the ball. The living room was the place that was used solely because that is where the resource guarding was happening, in peoples homes.
In no way are we approving dragging a dog on a leash
Thanks Katrin for the additional comments and suggestions. I always value your opinion.
And thank you Cbptraining for explaining a bit more about why this decoy dog was picked and about some of the additional considerations that were taken into account during training.
~Mary
Well, interesting post,
thanks!