Effects of single- or pair-housing on the welfare of shelter dogs: Behavioral and physiological indicators

Dogs are often housed alone in shelter settings to reduce injury and disease spread. However, social isolation can be a stressor for dogs. Prior studies have suggested that cohousing can produce behavioral and physiological benefits. These studies have typically focused on laboratory dogs or shelter dogs that have been kenneled for several months. Thus, those results might not necessarily generalize to shelter dogs, many of which have shorter lengths of stay than those dogs studied to date, and might be cohoused soon after intake. In fact, being pair-housed could, in the short term, be more stressful as dogs have to navigate novel social situations in small spaces. We investigated the behavioral and physiological effects of single- or pair-housing shelter dogs, most of which had recently entered the shelter. We collected behavioral data on 61 dogs (30 single-housed; 31 pair-housed) daily across seven days; we also collected urine for cortisol:creatinine analysis on a subset (22 single-housed; 18 pair-housed) for eight days (each day of the seven-day study plus a baseline sample on Day 0, prior to dogs’ enrollment). We found pair-housed dogs engaged in three stressrelated behaviors (lip licking, whining, and ears back) significantly less frequently than single- housed dogs. When we analyzed the change in urinary cortisol:creatinine (Days 1–7 values minus Day 0 value), we found that pair-housed dogs generally showed a greater decrease in cortisol:creatinine levels than single-housed dogs. Pair-housed dogs also had significantly shorter lengths of stay, but we did not detect any effect on dog-dog skills. Overall, we found well-matched pair-housing can have both proximate and ultimate welfare benefits for shelter dogs.