Dogs, demonstrators and detours: does the shape of the obstacle matter?

Making a detour around a transparent obstacle poses a challenge for dogs; however, social learning from a human demonstrator can make the task easier. From among many influencing factors, the effect of obstacle configuration on dogs' detour performance has not yet been investigated. We predicted that dogs would more easily detour around a straight obstacle, such as a fence, than an angulated one, because they would notice more easily the far end of the fence in the case of the straight configuration. We tested N = 84 companion dogs in four groups: straight and V-shaped fence, with and without demonstration. Each dog had to perform three consecutive trials. In the demonstration groups, the experimenter showed the dogs the solution before trials 2 and 3. We found that the dogs' detour latencies improved in both conditions only when they observed a demonstration. Irrespective of the demonstration, dogs were more successful and made a faster detour around the straight fence. In the case of the straight fence/control condition, dogs’ success rate showed a ceiling effect. These results imply that the spatial arrangement of an obstacle can affect the difficulty level of a detour for dogs. When they were able to see the boundaries of the obstacle more easily, either because of the wider angle of visibility or the shorter distance between the start point and the ends of the straight fence, this may have been more effective in inhibiting (futile) direct attempts to obtain the reward behind the obstacle, therefore resulting in higher success. We can assume that dogs may rely on response facilitation rather than stimulus enhancement when they learn from the demonstrator, as they did not show any differences in their social learning success between the two obstacle configurations. Our findings provide new insight on how the spatial layout of an obstacle can be used to potentially disentangle the social learning mechanisms in dogs during a locomotory detour task.