Years ago, I took care of a gooey-sweet adolescent Border collie, (Tilly, I’ll call her) who flattened her ears and folded like a bird’s wing every time you said her name. She was responsive and polite, and the other dogs seemed to like her as much as I did. It was especially rainy when she visited, so I appreciated that she never objected to endless paw wiping and toweling off, not to mention body checks for ticks and dental inspections. One morning I saw that she had grabbed something from the leaf litter in the woods, the kind of “something” you figure would be better off melding its way into the soil rather than ending up in the stomach of even the hardiest of dogs. I couldn’t tell what it was, but it looked well on its way to rotting itself into organic mush. Probably not the best snack for a dog to eat. I didn’t think twice about reaching toward her mouth to extract her woodland treasure, given how deferential Tilly was to both me and the other dogs. At least, not until I saw her body go stiff and her eyes go hard as the quietest of growls floated into the misty, spring air.
Uh oh. That’s the posture that behaviorists, trainers and owners of resource guarding dogs know well, (or learn fast), and it immediately sends the primitive part of your brain into Alert Mode. I always picture some version of a submarine’s warning signal blaring: UH ooooGA! UH ooooga! as the captain yells DIVE! DIVE! to the first mate. It’s a relatively common posture in the world of dogs, and it’s message is clear: “This is Mine. Attempts To Take It Away Will Be Met With Force.” Common although it might be, what do we know about its origins, and how should we handle it when it happens?
DEFINING THE TERM First, I should be clear about what I mean by “resource guarding” (RG, also known as “possessive aggression”). I define “resource guarding” as behavior that discourages another to take, or get too close to, an object or valued area in a dog’s possession. Usually this refers to food, treasured toys or sleeping areas, but I’d argue that some dogs guard their humans as if they were the best bone in the house. RG can range from a quiet head turn to a deafening growl, forward charge or an actual bite.
[Note: If you are primarily interested in how to treat or prevent RG, skip to the bottom of the post. I'm beginning with a discussion of more theoretical interest.]
(Someone asked a related, and great question: Should we include “territorial aggression” into the category of “resource guarding”? Hummmm. On the one hand I’d say No, in part because of my dislike of the term “terr’l aggression,” since so often it is used to describe agonistic displays from dogs who are not actually aggressive but are afraid of strangers. Given that neophobia is a very different motivation than a desire to possess something, much of what is called territorial aggression may have little to do with possessiveness. On the other hand, I’ve worked with several dogs who showed absolutely no sign of fear when I approached the house, but signaled what I interpreted as “You might want to rethink coming any closer to my den.” Thus, I’ll use my standard answer to all good but complicated questions: “It depends.”) By the way, Lee Niel and Jacquelyn Jacobs, in the Department of Population Medicine at the Ontario Veterinary College, University of Guelph, are currently doing research on “possession aggression/resource guarding,” which included asking a number of behaviorists their definition and what term they most often use. Stay tuned, I look forward to seeing their results when they come out.
WHEN IS IT A PROBLEM? Between Dogs: It is perfectly reasonable for one dog to signal another that his chew bone is HIS chew bone, thank you very much. Appropriate signals are head turns, stares and, depending on a host of other factors, a quiet growl. Appropriate responses are immediate withdrawals or strategic (and often brilliant) attempts by an item-less dog to worm her way into the others good graces. (Not to mention the famous distraction technique of Einsteinian dogs: BARK BARK BARK BARK!!! says the dog who wants the chew bone, vigorously vocalizing at the front window. Dog with bone drops it on the way to join in the barking, while Einstein Dog circles back and gets the bone.) Every owner has to decide what is acceptable in their own household; my criteria are quiet warnings like head turns or stares are acceptable, anything else is discouraged.
Between a dog and a person: This again is very much up to the owner, but I’ll go on record as saying that, in general, I expect every one of my dogs to let me take anything away from them without protest. Caveats: First, I am very careful not to abuse that right. I work hard to train my dogs to drop things on cue so that I am not taking something out of their mouth by force. Second, there are exceptions: Tootsie grabbed a dropped metal twist tie and ran off to swallow it a few days after she came to the farm. You’d better believe I swooped in like a falcon and took it out of her mouth. On the other hand, before Tulip my sheep-guarding Great Pyrenees died, she would occasionally find the body of a small mammal or bird in the woods or pasture. I made an executive decision that if she was in charge of protecting my flock from coyotes and stray dogs, she could be in charge of any treasures she could find in the woods or pastures. Not so in the house however. The two of us seemed to come to that agreement easily and with clarity. (But I had to teach her to be polite if she had a treasure and was approached by another dog: Stay tuned for the blog on how I did it some time later in spring or early summer.)
CAUSES OF RESOURCE GUARDING: That’s easy to answer: We don’t know. Seriously, we really, really don’t know. Does growing up in a large litter and having to fight for food make a difference? Could there be a genetic predisposition to resource guarding? Katie Martz and I could find nothing in our searches, so I emailed a list of Certified Applied Animal Behaviorists to ask if there is any research on genetic or environmental factors related to RG, and and there simply doesn’t appear to be anything out there on this specific topic. However, PhD behavior-geneticist Steven Zawistoski and PhD psychologist John C. Wright reminded me of some of literature that might relate in some way. Remember the early behavioral genetics studies done at the Jackson Laboratory by Scott and Fuller? Steve and John directed me to some of the early papers that might relate. Pawlowski & Scott (1956) did some of the early work on priority access to a valued item (which is the correct definition of “dominance,” by the way) among 4 breeds (Basenjis, Beagles, Wire-Haired Fox Terriers and American Cocker Spaniels) and summarized their results: “It is concluded that these differences are the result of genetic inheritance, which probably acts through physiological mechanisms which affect the threshold of stimulation.’ Of course, dominating access to a bone is not exactly the same as guarding it, but it includes it, because some dogs maintained ownership of the bone by doing what we define as RG.
More recently, Liinamo et al (2007), looked at genetic variation in “aggression-related traits in Golden Retrievers in the Netherlands, asking if owners saw “aggressive” behavior in a variety of contexts. Those related to RG were family members either approaching or removing a dog’s food, or removing a dog’s toy. The context of approaching or removing food had high “heritability” factors (.94 and .95) which does not mean that the behavior is “mostly genetic,” but means that there is a large amount of genetic variation related to the trait, and thus one could begin a selection process of selecting for or against a particular trait. (I always have to stop and take a breath when interpreting the term “heritability,” because a trait like “herding” in Border collies would show a low heritability, it being pervasive in BCs, and thus showing low genetic variability. Make sense? (Steve Z explained to me that he considers this term the genetic equivalent of the term “positive reinforcement,” because it means the opposite of what one might think.)
I would argue, based on the little research we have and my own experiences with hundreds of RG cases (1,000′s?), that there is a genetic component to the behavior. I’ve worked with litters of 11 dogs in which the biggest and strongest (and first to get to the nipple) pup became the RG dog very early in life. On the other hand, there is a great deal of research on a variety of species that reminds us that experience plays a significant role in “winning” and “losing” competitions. (See Hsu & Wolf 1999 for example.) One early win makes subsequent wins more likely, and vice versa. I suspect that this is one of those complicated behaviors that has both a genetic and an experiential component, and that the resultant behavior is some kind of interaction between nature and nurture. But again, we really don’t know. Anyone looking for a PhD topic?)
TREATMENT FOR INTERSPECIFIC GUARDING: I’m going to talk here about resource guarding between dogs and people. Treating it between two dogs uses the same basic principles, but requires enough alterations in technique to deserve its own article. That said, the most effective technique for stopping a dog from guarding resources from human intervention is to change your dog’s internal response to anothers attempt to possess their “treasure.” That is why you are best off using Desensitizing and Classical Conditioning to teach your dog to love it when you approach and reach toward an object. In other words, in this case you are not training your dog to respond to a cue, but conditioning an internal response to someone approaching something that they cherish.
Before going any further, stop here an contact a behaviorist or progressive trainer who understands how to use classical conditioning if your dog has ever put you at risk of being seriously injured. You’d call an electrician if you thought your wiring was unsafe in your house, wouldn’t you? Meanwhile, or if your dog is threatening but not dangerous, follow the steps outlined below.
STEP ONE: Be an armchair ethologist by thoughtfully and specifically writing down what objects your dog guards, what your dog does to cause you to say she is guarding, and how close you need to be to see any sign of guarding. Here’s an example:
Objects: Chew bone, stuffed Kong, favorite stuffed toy in the shape of a deranged dinosaur.
Behavior & Distance: My dog first stops chewing or eating, and stands motionless if I get within 4-5 feet of her while she is chewing on her Kong. If I move to within 2-3 feet, her body tenses and her mouth closes. If I walk right up to her and reach toward the object, she will growl.
STEP TWO: Find something your dog likes even better than what she guards. Usually it will be some form of meat, but every dog is different. Be sure to experiment–every trainer or behaviorist has seen X,000 numbers of people who swear their dog “doesn’t care about food” until we get out our super stash of cooked chicken or freeze-dried liver and get their dog turning somersaults for it. Food is ideal because you can have it on hand and chop it up into pieces that allow you to create lots of reinforcement.
STEP THREE: Stocked with lots of treats, set up a situation in which your dog would guard. In the example above, give your dog a stuffed Kong, leave the room and re-enter with a handful of cooked chicken. Stop WELL BEFORE you would predict a reaction (any reaction) from your dog. In the example above, that would be at about 7-8 feet away. Toss a piece of chicken so that it lands right beside your dog’s mouth. (If you are like me, and flunked softball in school, just toss another one if you miss.). Wait for your dog to eat it up, and toss another piece. Repeat once or twice, then leave the room. If your dog leaves the Kong and comes over to you for more, look up at the ceiling and ignore her. You want her to learn that food only comes out of the sky if she is eating and you are standing nearby.
STEP FOUR: After a few sessions of this, start where you began in the last session, but don’t toss any food until you walk forward one step closer, no more. Toss chicken and withdraw one step. Walk forward one forward again, toss a treat and then WALK AWAY. You want your dog to think “NO! Don’t walk away!!” If, however, your dog reacts by stiffening, make a mental note to start farther back or to only approach in half steps. You can either stop there, or leave the room and re-enter it, repeating Step Four one or two times.
STEP FIVE: Gradually, ever so gradually, decrease the distance between you and your dog. Walk to within 5 feet in one session, then 4 in the next. Go back to just 5 feet for 2 sessions, then go to 4 and possibly 3 IF the dog is responding well. “Responding well” means that your dog is switching from “Oh No! She’s going to take my bone away” to “Goody! Here she comes! Whenever I have a chew bone and she comes close to it I get something better! How cool is that????” That means your dog’s body is loose and not stiff. She does not start chewing frantically as you approach. Her mouth is open and she looks as if she is happily anticipating your approach.
What if she leaves the bone and come to me? Well, good girl Fidette, that means you’ve stopped guarding the bone in search of something better. Again, simply ignore her and wait for her to return to her bone. It might take awhile for some dogs, but if you look away (this part is important) she will eventually give up and go back to her Kong or dinner bowl.
STEP SIX: Once you can approach your dog and stand right beside her, begin skipping the food toss until you are a few strides away, and start classically conditioning a reach toward the object. Keep in mind that you are working on re-wiring her brain so that she forms a new association between your actions and how she feels about them. Walking toward her is a different action than reaching toward her, so you need to think of it as a different category. (Understanding the distinction between each action you make is perhaps the most important aspect of being able to use classical conditioning to turn around a behavior, and it is not something we do naturally without training ourselves to be expert observers and thoughtful analysts of behavior.) First, bend toward the food or toy, drop a treat and then straighten up. Do this several times, or as often as necessary for your dog to remain relaxed. Remember: your dog drives the system here, not an idea you have in your head for how long this should take. Gradually move your arm and hand closer and closer to the food or object, eventually taking it away and giving your dog something wonderful in return. I once convinced a head-strong and very RG’y dog to give me the dead bird she had in her mouth, and when she did, I gave it back to her. The people watching were appalled, but that’s what she wanted more than anything in the world, and she trusted me ever after.
STEP SEVEN: Keep it up. Forever. Not every day, or even every week, but at least every month or so you should remind your dog why it is in his or her best interests to let you take anything away.
PREVENTION: That’s easy–just follow the step above, but you don’t have to go as slowly as you would if you were trying to turn around an established behavior. Willie and Tootsie both love it when I pick up their bowls, because it means they are getting something even better. Neither have ever even suggested a modicum of RG’g, which is exactly why I continue to remind them how fun it is to let me take things away from them!
OPERANT CONDITIONING?: One last comment–there is a role for operant conditioning here, which is to teach dogs to “Leave It” or “Drop It” (those are different in the mind of a dog I suspect: in one case the dog is focused on something, in another he or she has it in his or her mouth, and possession is the law in canid society.)
MEANWHILE back on the farm: Spring has sprung! Although today it is cool and rainy, it’s been absolutely glorious for a few days. (My sympathies to those in Minnesota and northern Wisconsin who got over a foot of snow yesterday. Not fair, not fair at all!) The lambs are finally settling in, all 17 of them. We have one orphan (Ralphie, rejected by his mother) and 4 other lambs who need supplemental feeding. Feedings occur 4 times a day (down from 8, whew!) and take about a half hour total to get the milk ready and the lambs fed. The ewes are getting a mix of fresh, spring grass, alfalfa hay and a corn/oat mix and so should be making lots of protein-rich milk for their babies. (Would someone tell momma Buttercup that bawling at us at 120 decibels every time she hears our voice will NOT result in any more food? Not to mention that there are sheep starving in China….)
The lambs are now old enough that Willie can help me move the flock around, and we’ve gotten up the portable, electric fences so that we can do controlled grazing all summer. We are re-seeding one third of the pasture, it suffering from the drought last year, and will have to keep the sheep off of that for two months or so once it is seeded. Meanwhile, the daffodils are sun-shining their blooms all over the yard, and the crocus are starting to fade. Here are some now, along with the first bee I saw the spring: