Behavior medicine: Working for food

Q: Should treats be avoided during training? Some trainers say they are bribes and that dogs should work to please. Is this correct?

A: Used correctly, food is an excellent primary reinforcer — not a bribe. When sufficiently hungry, any dog should find food motivating. In fact, learning principles apply to all animals. If the consequences of a behavior are good, the behavior is likely to increase (reinforcement).

However, it takes repetition, consistency and timing to learn what consequences are associated with what behaviors. Behaviors that are reinforced immediately, consistently and predictably will be learned more quickly, while intermittent reinforcers are best for maintaining behaviors.

Food preferences vary between dogs, but novel foods are often most enticing, and hunger will further increase motivation. The most motivating food should be saved exclusively for (contingent upon) new learning. Lower-value food can be used intermittently to maintain existing behaviors. Food can be used to lure a dog into the desired behavior (e.g. sit, down, come). Once the behavior can be achieved, the food should be hidden and a hand signal or verbal command used to cue the behavior.

If the rewards are consistent and predictable, the dog controls the acquisition of treats by exhibiting the very behaviors that the owners wish to train. Favored rewards can be used to prevent and manage fear and anxiety by associating (or pairing) the food with exposure to potentially threatening stimuli (counter-conditioning), such as veterinary visits.

The primary concerns about treats are obesity and nutrition. If the treats are healthy and no more than 10 percent to 20 percent of the daily calories, they should have minimal impact on nutrition. To prevent obesity, all calories used in training must be included when determining total daily intake. Using multiple tiny morsels of meat, cheese, vegetables or low-calorie dog treats allows for a maximum number of training opportunities with a minimum number of calories.

Because food acquisition (e.g., hunting and scavenging) is an inherent part of each animal's daily regimen, it can be counterproductive to provide food from feeding dishes.

Working for food is a concept that will give the dog maximal opportunity for mental enrichment and physical activity while expending calories to obtain food. Pets can work for food through reinforcement training (discussed above) and by using food search, food manipulation, toys and food-stuffed toys as an alternative to feeding out of a bowl.

Similarly, there are toys where the chew products are attached to the toy to make chewing more challenging and time-consuming. Furthermore, these food and chew toys provide an active way to keep a dog "entertained" when the owner is not available to supervise or interact with the pet. Also, they can be used to train dogs to enjoy their time on their mats, in their crates or confinement area. These toys could also be provided when guests arrive to keep the pet occupied.

At the end of each day when training is completed and food toys have been emptied, any remaining calories can be given as an evening meal. This should keep your dog food-motivated throughout the day and provide a small meal before the last daily walk and bed time.

Giving food from a bowl means multiple lost training opportunities, and provides little in the way of enrichment or stimulation, especially if there are still more than 23 hours left in the day after eating. Keep in mind: rewards that are of highest value, contingent upon a specific behavior, consistent and properly timed will lead to fastest learning. However, if the dog gets rewards inconsistently or unpredictably, it wastes valuable training opportunities, may slow new learning, make the pet anxious and conflicted, or even reinforce undesirable behaviors.

If you have a topic or question you would like answered in this column, please send an e-mail to northtorontovets@rogers.com.

The authors practice at North Toronto Animal Clinic, Thornhill, Ont., Canada.