State- and Context-Dependent Learning Effects in Different Dog Populations

Animal shelters have limited resources, especially when considering the time and resources required for implementing behavior change protocols for resident dogs. This issue of resources is particularly relevant to the question of whether training should be carried out while dogs are living in shelters or if shelters should allocate their training resources to when the dog has been adopted and is in its new home. Given that a shelter is a different context from an adoptive home, a context in which dogs experience much greater stress levels than those living in homes, their learning could be impacted by both state- and context-dependent effects. These effects could result in reduced retention and recall of behaviors learned while living in the shelter. Research in other species has shown state-dependent and context-dependent learning effects, little is understood about the retention of learned behaviors by shelter dogs. In this study, we trained 20 owned and 20 shelter-living dogs in a training room within the animal shelter to acquire a novel behavior: touching their nose to a traffic cone. One month later, we tested both owned and shelter dogs’ performance of this behavior in their homes. We then assessed their response latency, and whether dogs correctly responded in their homes at the same training level reached in the shelter, or if behavioral criteria needed to be lowered for them to respond correctly. Results show that acquisition and recall by former shelter dogs closely correspond to that of owned dogs. Shelter dogs had much longer initial response latencies than owned dogs, but both groups were able to demonstrate fluency and test at higher criteria levels in the home than at the shelter. Dogs also demonstrated generalization of the behavior to their owners.