Using Research-based Protocols with Shelter Dogs

Written by Kimm Hunt

Peer reviewed

Summary: Shelter personnel and consultants are always looking for new tools that to introduce dogs into a new home, help resolve inter-dog aggressive behavior within a household, or form shelter playgroups. This article describes how a single evidence-based protocol can be used to do all three. The training plan is a modified protocol researchers and shelter staff in Brazil developed that is designed to identify dogs that can be group-housed together.

Background

In the late 2000s, some states in Brazil passed a law prohibiting euthanasia of healthy dogs and cats in animal shelters.1 Shelters suddenly needed to accommodate long-term residents because they could no longer euthanize for kennel space. This created a need to house dogs in pairs and groups to maximize shelter capacity. Group housing, when possible, improves the welfare of long-term and permanent residents. This is something that the European Union made into law via a parliamentary animal welfare mandate that social animals be housed in stable groupings.2
Since shelter populations change daily, co-housed dogs typically need frequent regrouping. Researchers and animal control staff in Brazil developed a protocol to test the suitability of dog pairs for co-housing. It uses enrichment, counter-conditioning, and systematic desensitization to increase the likelihood a particular pair can be kenneled together. The protocol takes 30 minutes from first introductions to group kenneling.

Forming playgroups

I am always on the lookout for research-based protocols to adapt and use in my work with animals. I was working with a dog breed rescue that needed a fast, reliable way to form small playgroups that required minimal supervision. The organization had a small kennel, about a dozen dogs, two staff members, and, amazingly, a fenced acre. The staff wanted to allow each dog to spend a minimum of 30 minutes outdoors every morning and afternoon while they worked in the kennel. Most of the day there was only one staff member present, with a one- or two-hour overlap in the early afternoon. They did not have enough regular volunteers to rely on trained helpers to manage twice-daily playgroups.

I remembered reading a study a few years prior about selecting shelter dogs for group housing. Fortunately, it was open-access and I could read the entire article. It’s not a double-blinded, placebo-controlled, large-scale experimental research study. The authors didn’t collect data prior to using the protocol to determine its effectiveness compared to the previous methods shelter staff used—casual behavioral observations and trial-and-error groupings. But the staff involved in the research felt that the new process was faster and more reliable than their previous methods, resulting in fewer fights and failed groupings. This was good enough for me to give it a try.

Basic protocol for forming shelter playgroups and dog introductions

Adapted from Santos et al.4

Equipment needed:

Head collars (Gentle Leader/Halti or head collar fashioned from rope leashes—see Santos et al. for instructions), flat collars, or harnesses and leashes
11 field markers (cones, weighted cardboard boxes or frisbees, buckets, flags, etc.)
Two approximately 50 square-ft testing areas
Two handlers; an additional helper to set up the field markers and film is ideal
Mid-value reinforcers

Phase 1 (5-10 minutes)

Set up a triangle that is approx. 30 x 30 x 42 feet (12 x 12 x 15 paces). Place a field marker at each point of the triangle. Keep dogs at least 30 feet apart when Handler 1 enters the triangle with Dog 1.

Handler 1 walks up to Point A of the triangle and pauses, allowing the dog to sniff/mark the cone. Reinforce any calm behaviors using R+ relevant to Dog 1.
Handler 1 walks to Point B in the triangle and pauses, allowing Dog 1 to sniff/mark the cone. Reinforce any calm behaviors.
When Handler 1 reaches Point B, Handler 2 walks Dog 2 up to Point A and allows him to sniff/mark the cone. Reinforce any calm behaviors.
As Handler 1 walks to Point C, Handler 2 walks to Point B. Both allow their dogs to sniff/mark. Reinforce any calm behaviors.
Handler 1 then walks about 15 feet away from the triangle/Point C to allow Dog 2 to approach Point C, letting the dog mark/sniff when they get there. Reinforce any calm behaviors.
Hander 2 walks back to Point B and lets dog sniff/mark. Reinforce any calm behaviors.
When Handler 2 reaches Point B, Handler 1 walks back to Point C and lets dog sniff/mark. Reinforce any calm behavior.
As Handler 2 walks back to Point A, Handler 1 walks back to point B. Allow dogs to sniff/mark. Reinforce any calm behaviors.
Handlers 1 and 2 walk dogs to second testing area, keeping dogs at least 15 feet apart.

Phase 2 (10-15 minutes)

Set up a circle approximately 30 feet in diameter, and place field markers at 12, 3, 6 and 9 o’clock. Inside the circle, mark 4 points that are 6 feet apart.

From Santos, et al., 2013.4

This phase has five steps. Repeat each step three times, moving on only when there are no signs of aggression.

At each step, reinforce any calm behavior. If one or both dogs show aggression during three successive attempts at a step, stop the process and repeat the next day, pairing the aggressor with a different dog. If this second attempt fails, the dog(s) that showed aggression in both trials shouldn’t be kenneled together. Depending on the level of aggression, they may be suitable to include in a playgroup with constant monitoring.

Beginning at Points A and B (Figure 2), walk Dogs A and B clockwise on the circle to complete 1 rotation. Reinforce calm behavior with praise, treats, and petting at each point on the circle. Repeat three times.
Beginning at Points A and B, walk dogs to the 1A and 1B points (just inside the circle), ask for a sit, reinforce with treats, petting, praise. Return to Points A and B. Repeat three times.
Beginning at Points A and B, walk dogs to the 2A and 2B points (a little farther inside the circle), ask for a sit, reinforce with treats, petting, praise. Return to Points A and B. Repeat three times.
Walk dog at Point B around the edge of the circle to Point A, approaching Dog A from behind. Stop about 10 feet behind Dog A.
Walk dogs around the circle three times, with Handler A and Dog A in front and Handler B and Dog B closely following but not allowing physical contact between the dogs.
Stopping at each of the four points on the circle, allow Dog B (the follower) to smell Dog A for 1 second. Reinforce calm behavior with treats, petting, praise. Dog A (the leader) moves forward immediately after the 1-second sniff.
After three rotations, Dog B at the back passes Dog A to the front. Walk around the circle again, stopping at the points for a 1-second sniff, and reinforcing calm behavior. Repeat three times.
Handlers walk dogs together, side by side, with 6 feet between them, for 2-5 minutes.

Phase 3 (5 minutes)

This can take place in a fenced area outdoors to form playgroups, or in an enclosed space (minimum of 20 x 20 feet) to form kennel groups.

With dogs in close proximity, remove the head collars and attach leash/dragline to collars. Handlers watch (calm, silent, avoiding abrupt movements) for any signs of aggression.
When dogs interact for 5 minutes with no signs of aggression, they’re an official “safe pair.”

Repeat the process and pair all dogs for the desired playgroup. Supervise the playgroup at the level you deem necessary based on the number of dogs and the behavior you observe in the group (ranging from constant supervision with multiple handlers to occasional supervision by one handler).

Our goals and implementation

Our goal was to form groups of two or three dogs that could be left in the yard with periodic check-ins by one staff member. Staff would observe the dogs’ behavior through the shelter’s windows, which allowed for an almost full view of the yard, or by walking outside. Since every part of the kennel was within a 15-second walk to an exit door, check-ins would take about 45 seconds. Staff members were knowledgeable enough to identify agonistic behavior in dogs and had experience intervening in fights.

Staff used the protocol for about one month to pair dogs with no incidents of fighting. Since they ruled out pairing dogs who showed aggressive behavior at any point during the protocol, they abbreviated it and began using a shortened version with the same success. Because they sometimes only have one staff member and no volunteers, they prefer to pair the dogs rather than group them. They sometimes group dogs in threes and use the abbreviated protocol with each dog in the group prior to selection.

Introducing adult dogs

I’ve also used the protocol with adopters looking to add another dog, including dogs with what I would consider mild-to-moderate dog-aggression issues. For example, a family was looking to adopt a 1-year-old female Australian shepherd mix who was leash-reactive to all dogs, manifesting as barking and lunging at other dogs from a distance of approximately 30 feet. They had a 3-year-old neutered male Rhodesian ridgeback mix who had mild leash-reactivity issues and was “dog-selective” when interacting with dogs off leash. On leash he would bark and growl, no lunging, at approximately 50% of dogs who approached within 15 feet. Off leash he would typically greet other dogs with mild interest, then growl and snap at dogs who attempted to play with him using physical contact. He sometimes enjoyed brief, reciprocal chase games with other dogs.

We introduced the dogs using the protocol, took them on a 30-minute parallel walk together with a minimum of 6 feet between them, then allowed them to greet on-leash. The Rhodesian showed no signs of leash reactivity and tolerated the Aussie’s brief attempts to engage him in play after greeting. The family adopted the Aussie. They used the protocol daily for five days, along with slow introductions inside the home and additional training protocols. After approximately three weeks, the dogs were spending most of the day together while supervised, without incident.

Integrating dogs with dog aggression issues

A client was fostering an adult female Labrador retriever mix who had whelped and weaned a litter in the home. They kept her in a back bedroom because she attacked the family dog when the two were introduced in the home. Both dogs sustained puncture wounds around the head and neck that did not require veterinary care. She was also leash reactive and began barking, growling, lunging, and pulling toward dogs she would see walking at a 100-foot distance. And, she tried to attack one of her pups who returned to the client’s home. Both dogs were leashed, and she did not make contact.

The client wanted to integrate her into the home with the family dog. We used an extended version of the protocol with additional counter-conditioning protocols and careful management, outlined below, over a five-week period, and assimilated the two dogs in seven weeks.

Basic protocol for mitigating inter-dog aggression within the home

Day 1: We (the owners and I) completed phases 1 and 2 of the protocol twice, omitting steps 2 and 3 of phase 2. We walked the dogs parallel to each other 20 feet apart for 10 minutes, then 15 feet apart for 10 minutes. We separated dogs upon return to the home.

Day 2: We completed phases 1 and 2 of the protocol twice, omitting step 3 of phase 2. We walked the dogs parallel to each other 20 feet apart for 10 minutes, then 15 feet apart for 10 minutes. We separated dogs upon return to the home.

Day 3: We completed phases 1 and 2 of the protocol twice, omitting step 3 of phase 2. Then we walked the dogs parallel to each other 20 feet apart for 10 minutes, then 15 feet apart for 10 minutes, then 10 feet apart for 10 minutes. We separated dogs upon return to the home.

Day 4: Owners completed phases 1 and 2 of the protocol twice. They walked dogs together 20 feet apart for 10 minutes, 15 feet apart for 10 minutes, then 10 feet apart for 10 minutes. They separated dogs upon return to the home.

Day 5: Owners completed phases 1 and 2 of the protocol twice. They walked dogs together 15 feet apart for 10 minutes, then 10 feet apart for 10 minutes, then 6 feet apart for 10 minutes. They separated dogs upon return to the home.

Day 6: We completed phases 1 and 2 of the protocol twice. We walked dogs together 10 feet apart for 10 minutes, then 6 feet apart for 10 minutes, then together for 10 minutes. Dogs were returned home and introduced with a barrier between them (a tall pressure gate) for 10 minutes. We reinforced calm behaviors in both dogs.

Day 7: We completed phases 1 and 2 of the protocol once. We walked dogs together 10 feet apart for 20 minutes, then 6 feet apart for 20 minutes, then together for 20 minutes. Dogs were returned home and introduced with a barrier between them (a tall pressure gate) for 15 minutes. We reinforced calm behaviors in both dogs.

Day 8: Owners completed phases 1 and 2 of the protocol once. They walked dogs together 6 feet apart for 20 minutes, then together for 20 minutes. Dogs were returned home and introduced with a barrier between them (a tall pressure gate) for 20 minutes. Owners reinforced calm behaviors in both dogs.

Day 9: Owners completed phases 1 and 2 of the protocol once. They walked dogs together 6 feet apart for 20 minutes, then together for 20 minutes. Dogs were returned home and introduced with a barrier between them (a tall pressure gate) for 30 min. Owners reinforced calm behaviors in both dogs.

Day 10: Owners completed phases 1 and 2 of the protocol once. They walked dogs together 6 feet apart for 20 minutes, then together for 20 minutes. Dogs were returned home and introduced without a barrier, both dogs dragging leashes, for 10 minutes. Owners reinforced calm behaviors in both dogs.

Day 11: Owners completed phases 1 and 2 of the protocol once. They walked dogs together for 30 min, then returned home and introduced without a barrier, dragging leashes, for 20 minutes. Owners reinforced calm behaviors. Owners then introduced an additional 10-minute period of protected contact (one dog crated in the living room, the other loose).

Day 12: Owners repeated a truncated version of the protocol (one trip around the triangle, one around the circle), they walked dogs together for 30 minutes, then returned home to the living room, dragging leashes, for 30 minutes. Owners reinforced calm behaviors. Owners did an additional 15 minutes of protected contact using the crate, with the dogs reversed.

Day 13: Owners repeated a truncated version of the protocol (one trip around the triangle, one around the circle), they walked dogs together for 30 minutes, then returned home for 30 minutes of contact in the house. Owners then did two sessions of protected contact using the crate, 15 minutes each.

Day 14: Owners repeated a truncated version of the protocol (one trip around the triangle, one around the circle), they walked dogs together for 30 minutes, then introduced in the LR, dragging leashes, for 45 minutes. Owners also did two sessions of protected contact using the crate, 15 minutes each.

Day 15: Owners walked dogs together for 30 minutes, then introduced in the LR for 1 hour. Owners also did multiple sessions of protected contact using the crate, 15 minutes each.

Day 16: Owners walked dogs together for 30 minutes, the introduced in the LR for 1 hour twice per day. Owners also did multiple sessions of protected contact using the crate, 15 minutes each.

Day 17: Owners walked dogs together in the morning for 30 minutes, then introduced in the LR for 1 hour, then separated for 3-4 hours, the introduced for 1 hour, then separated for 3-4 hours, then walked for 30 minutes and introduce for 1 hour.

After day 17, the owners continued walking dogs together daily and gradually increased time intervals together and decreased separation time until the dogs were coexisting together all day. The owners did this over two weeks.

We used the adapted protocol longer than necessary because the client believed the dogs enjoyed it. Based on my observations, I agreed. According to the client, the dogs became “best friends” and had no incidents of aggressive behavior post-intervention.

Conclusion

I’ve found this protocol invaluable for introducing dogs with potential dog-directed aggression issues under three different circumstances. It is a simple, detailed plan that involves minimal preparation and equipment. It requires two large outdoor areas (about 900 square feet each), although I’ve used the same area for phases 1 and 2 when space was not available. If you use it in your work, I hope you’ll reach out and share your experience with me.

References

Kimm Hunt, CPAT-KA, CSC-D, owns Kimm Hunt Dog Training in Alpharetta, Georgia. She is working on an MSc in clinical animal behavior at the University of Edinburgh.

TO CITE: Hunt, K. (2020) Using research-based protocols with shelter dogs. The IAABC Foundation Journal 18, doi: 10.55736/iaabcfj18.7