One of the most misread behaviours

A dog is lying with a bone or treat, and someone walks past; the dog freezes, and a low growl starts up. The usual reading of that moment is that the dog is being aggressive, or worse, that he is ‘challenging his owner for status’, and that the correct response is to take the bone away and show the dog who is boss.

The science says that almost every part of that reading is wrong, and acting on it is one of the more reliable ways to turn a manageable behaviour into a bite. Ultimately, the single most important decision an owner makes when they meet resource guarding behaviour is whether to reach for confrontation or move away from it. This article looks at the science based way to approach this behaviour.

What resource guarding is

Resource guarding describes the behaviours a dog uses to keep control of something he values, in the presence of a person or another animal. The resource is usually food, a chew or a toy, but it can be a sleeping spot, or even a person.
The behaviours run along a spectrum, from the barely noticeable to the dangerous: eating faster as a person approaches, turning the body away, hovering over the object, a hard stare, a freeze, a lifted lip, a growl, a snap, and at the far end, a bite.

It is worth being precise about the language, because the field itself has argued over it. A 2018 study surveying canine behaviour experts noted that the terms resource guarding and possessive aggression are often used interchangeably, and that this inconsistency muddles communication between owners and clinicians and can affect how well the behaviour is treated (Jacobs et al., Frontiers in Veterinary Science, 2018). The useful distinction underneath the terminology is one of degree. Keeping hold of something valuable is normal. The question is where on the spectrum a given dog sits, and how much of what looks alarming is actually just communication.

Whether it is normal

The behaviour is normal in the sense that it is common; it appears in dogs with no history of abuse or deprivation, and the underlying instinct to hold onto a valuable thing is one that would serve any animal well in a competitive environment. Guarding food and objects is adaptive when resources are genuinely scarce and contested.

However, researchers have pointed out that a companion dog on a reliable feeding schedule does not actually need to guard, because unlike a free-ranging dog he is not competing to stay alive (a point made by Bradshaw and colleagues, and discussed in Jacobs’s work on the behaviour). This means that resource guarding in a well-fed pet is better understood not as a rational response to true scarcity, but as an emotional response, most often rooted in anxiety about loss, running on instincts built for a world the dog doesn’t live in.

That distinction matters enormously for what we as owners do next, because it tells us that the dog guarding his bowl is not making a bid for control of the household. Rather, he is worried, and we cannot punish an animal out of being worried. We can only make the animal more so.

Why the growl is good

The most counterintuitive thing in this whole topic is that the growl is not the problem. The growl is the warning that lets us avoid the problem.

A dog that growls over his food is telling us, clearly and non-violently, that he is uncomfortable and would like us to give him space. This is a functioning early-warning system. The dog that has learned growling gets him punished does not stop feeling uncomfortable, but he might stop warning. He then may skip the growl and go to the bite, because the safe middle step has been trained out. This is why experienced behaviourists are close to unanimous on one point: do not punish the growl.

This is also why the folk cure of regularly taking a dog’s food away to teach him who is in charge is precisely backwards. From the dog’s point of view, when we do this, we confirm his biggest fear: people approaching the bowl really do take good things away. When we do this, we have made the resource more worth defending, not less. In one clinical description, an owner who reached in to take an item was met with a growl, persisted, and was bitten. The dog did not escalate out of nowhere. He escalated because his warning was ignored.

What the evidence says about punishment

The case against confrontational methods is not a matter of training-world fashion. It sits on some of the more solid evidence in companion animal behaviour.

The most quoted study surveyed 140 owners of dogs referred to a veterinary behaviour service, asking what techniques they had tried and how the dog had responded (Herron et al., Applied Animal Behaviour Science, 2009). The confrontational methods produced aggression at strikingly high rates. Hitting or kicking the dog for undesirable behaviour drew an aggressive response in 43% of the dogs it was used on. Growling at the dog, 41%. Physically forcing an item from the dog’s mouth, 39%. The alpha roll, 31%. Grabbing the dog by the jowls and shaking, 26%. By contrast, non-aversive approaches such as rewards and additional exercise produced very little aggression.

That is directly relevant to resource guarding, because forcing an item from a dog’s mouth is exactly the intervention a guarding dog most often meets, and it is one of the most reliable ways of getting bitten in the whole list.

What the evidence supports

• Resource guarding is common and appears in dogs with no history of abuse or deprivation.
• A growl is a warning. Punishing it removes the warning, not the feeling, and raises the risk of a bite.
• Confrontational methods, including forcing items from the mouth, provoke aggression in a large share of dogs.
• Reward-based methods are at least as effective and carry far less risk.
• A reliable, trained ‘drop it’ on cue is associated with a lower risk of biting.

A wider review pulled together seventeen studies on training methods and concluded that aversive methods, meaning positive punishment and negative reinforcement, can jeopardise both the physical and mental welfare of dogs, and that there is no evidence they work better than reward-based methods, with some evidence the opposite is true (Ziv, Journal of Veterinary Behavior, 2017). In fairness, this was a focused literature review rather than a formal systematic review, and the individual studies it draws on have their own limitations, but the direction of the evidence is consistent across surveys, observational work and interventions, and it points one way.

On the specific question of what helps, a study of risk factors for guarding found that dogs taught to reliably drop an item on cue were significantly less likely to bite in guarding situations (Jacobs, 2018). The thing that reduces the danger is not dominance, but a trained, rewarded, voluntary release.

What to do instead

The evidence-based approach has two halves that work together: manage the environment so the behaviour is not triggered, and change the underlying emotion so the trigger stops mattering as much. Crucially, neither half involves a confrontation.

Manage first. If a dog guards a particular chew, the immediate answer is often simply not to hand out that chew, or to give it in a separate room behind a closed door where nobody has to walk past. This is not a failure to train. It is preventing rehearsal, and every guarding incident that does not happen is one that does not get more practised and more entrenched.

Then change the association. The two techniques with the best support are counterconditioning and teaching a trade. Counterconditioning means turning a person’s approach into a reliable predictor of something good: rather than the person near the bowl meaning the food is about to be taken, it comes to mean an even better piece of food is about to be added. Over many repetitions, at a distance the dog is comfortable with to begin with, the dog’s feeling about the approach changes. The trade is the same logic made into a cue. The dog learns that giving something up produces something better, so ‘drop it’ becomes a good deal rather than a threat.

The mechanism running underneath both is worth stating, because it is the whole reason the approach works. We are not suppressing the behaviour, but rather removing the reason for it. A dog that no longer believes people take his things has nothing left to guard against.

When the guarding is between dogs

So far this has mostly described a dog guarding from a person. In practice, a great deal of resource guarding in South African homes is dog-on-dog, and it deserves treating in its own right rather than as a footnote, because both the risk and the management are different.

The research bears out how common it is. The same group that clarified the terminology also looked specifically at guarding directed at other dogs, in a survey of thousands of owners, and found that dogs living in multi-dog households were significantly more likely to show guarding aggression, avoidance and rapid eating than dogs living alone (Jacobs et al., Applied Animal Behaviour Science, 2018). Put plainly, adding a second dog raises the odds of guarding, because now there is a genuine competitor in the room rather than a person who was never really after the bowl in the first place.

This is the important difference from the human version. When a dog guards from a person, the fix rests on teaching him that the person is not a threat to his resources, which is usually true. Between two dogs, the competition is real. The other dog might actually take the bone. Therefore, the emphasis shifts away from changing a false belief and towards removing the competition, so that neither dog has anything to contest.

In a multi-dog home the management is the intervention, and it is not a temporary measure while training happens. It is often the permanent arrangement, and there is nothing second-rate about that.

One thing worth saying: the old advice to let dogs ‘sort it out between themselves’ is not sound, and it is not kind. Dogs do not reliably resolve resource conflict on their own, and every fight that is allowed to happen teaches both dogs that the next confrontation is worth escalating. Left unmanaged, a low-level grumble over a bone can harden into two dogs who genuinely cannot be fed in the same room, and relationships between housemate dogs can break down in ways that are very hard to repair.

Managing guarding between dogs

• Feed dogs completely separately, in different rooms or crates or behind a baby gate, every meal, without exception. This is the single most effective step and the one most often skipped.
• Pick up bowls once each dog has finished and walked away, so there is nothing left to linger over or contest.
• Do not leave high-value chews, bones or favourite toys lying about. Bring them out for separate, supervised sessions and put them away afterwards.
• Watch the flashpoints that are not food: a doorway, the sofa, a favourite bed, or the owner’s attention when everyone comes home at once. Guarding of space and of people is easy to miss.
• Make sure there is plenty to go round. Several beds, several water bowls, enough toys, so that no single resource becomes the thing worth fighting over.

If a fight breaks out

Even in a well-managed home, guarding can occasionally tip into a fight, so it is worth knowing in advance how to separate two dogs without adding a bitten person to the casualty list. The instinct is to grab collars and reach in with both hands. That instinct is how owners lose fingers, because dogs in a fight are reacting rather than thinking, and will redirect a bite onto whatever is closest, including a familiar hand, without recognising it.

The safer sequence, drawn from the guidance veterinary and welfare bodies give, is this. First, try to interrupt without touching either dog at all: a sudden loud noise, or a barrier such as a large piece of board, a bin lid or a chair pushed between them, or water from a hose or an upended bucket. Startling the dogs apart is far safer than pulling them apart.

If they must be physically separated and there are two people, the method with the widest support is the ‘wheelbarrow’. Each person takes one dog, grasps the tops of his back legs where they meet the body, lifts the back end off the ground like a wheelbarrow, and walks backwards, turning in a curve rather than a straight line so the lifted dog cannot swing round to bite. Grasp high, at the thigh rather than below the knee, to avoid injuring the joint. The dogs should then be taken out of sight of one another and given time to settle before anyone checks them over for injuries.

Doing this alone is genuinely dangerous, and there is no fully safe one-person technique, which is the strongest possible argument for prevention. If guarding in a household has already produced a fight, that is the point to bring in a qualified behaviourist rather than to keep managing it alone, and any dog involved should be checked by a vet, since puncture wounds often look trivial and are not.

When to call a professional

Mild guarding of a favourite toy that resolves when the dog is left alone with it is the sort of thing an owner can often manage with the approach above, plus patience. However, some situations need more than an article can begin to offer.

A dog that leaves his food to go after a person or another dog is doing something beyond ordinary guarding and should be seen by a vet or a qualified behaviourist, partly because a sudden change in this behaviour can have a medical cause, from pain to gastrointestinal disease. Similarly, guarding in a home with young children is not a situation to experiment with, because the cost of getting it wrong is too high. And any dog that has already bitten, or that guards so intensely the household is organising itself around avoiding it, needs professional help rather than a home programme.

Because South Africa’s behaviour profession is not tightly regulated, choosing a behaviourist well matters. The single most useful filter is method. A behaviourist who talks about dominance, being the pack leader, or correcting the dog is working from a framework the evidence has moved past, and in this particular behaviour that framework is not just outdated but actively risky. Ask directly what happens when the dog growls, and if the answer involves punishing it, keep looking.

The verdict

Resource guarding is one of the clearest cases in dog behaviour where the instinctive human response and the evidence-based one point in opposite directions. The instinct is to meet a growl with force and assert control. The evidence says that force is what converts a warning into a bite, and that the dog was never after control in the first place.

The dog with the bone is not staging a global takeover. He is telling us, in the only language he has, that he is afraid of losing something he values. Whether that fear is of a person or of the other dog on the sofa, the good news is the same: this is one of the more solvable problems in behaviour, and the solution is neither complicated nor confrontational. It is to remove the thing the dog is worried about.

References

• Herron ME, Shofer FS, Reisner IR. Survey of the use and outcome of confrontational and non-confrontational training methods in client-owned dogs showing undesired behaviors. Applied Animal Behaviour Science, 2009; 117: 47 to 54. • Ziv G. The effects of using aversive training methods in dogs, a review. Journal of Veterinary Behavior, 2017; 19: 50 to 60. • Jacobs JA, Coe JB, Widowski TM, Pearl DL, Niel L. Defining and clarifying the terms canine possessive aggression and resource guarding: a study of expert opinion. Frontiers in Veterinary Science, 2018; 5: 115. • Jacobs JA, Coe JB, Pearl DL, Widowski TM, Niel L. Factors associated with canine resource guarding behaviour in the presence of dogs: a cross-sectional survey of dog owners. Applied Animal Behaviour Science, 2018. • Wheelbarrow and fight-separation guidance consolidated from veterinary and welfare bodies (for example, AKC and Preventive Vet).