Evaluation of the capability of oil specific discrimination in detection dogs.

Dogs are used for oil detection to support spill remediation and conservation, but little is known about the effects of weathering and aging of oil odorants on dogs' ability to generalize and discriminate unweathered oil from aged/weathered tar ball oil. Three dogs were trained to detect unweathered oil odorant using a three-alternative choice procedure and automated olfactometers. We evaluated dogs' ability to discriminate unweathered target oil from four different weathered/tar ball samples. All three dogs successfully discriminated the unweathered target oil from the four nontarget weathered oils with an accuracy of 96%, 97%, and 100%. After the oil discrimination test, dogs' ability to discriminate unweathered target oil from novel natural odorants on a beach (plastic bottle lid, bird feathers, and rocks) was tested in a novel discrimination test yielding an accuracy of 95%, 100%, and 100%. These data suggest dogs are successful in discriminating unweathered oil from weathered oil with explicit training.

Development of point following behaviors in shelter dogs.

Pet dogs are known to be responsive to human pointing gestures, but shelter dogs have repeatedly demonstrated poor abilities to follow human pointing, although they can be explicitly trained quickly. This study evaluated the time course in which shelter dogs learn to follow points without explicit training, when given typical interactions with humans. In a longitudinal evaluation, the development of point following was tracked in seven shelter dogs in a training program (enriched human exposure), seven dogs in a traditional shelter (control population), and evaluated once in pet dogs. Twice a week for 6 weeks, shelter dogs' point-following performance was evaluated in 10 probe trials in which an experimenter pointed to one of two containers equidistant from the dog. To avoid direct training, dogs were given a treat for approaching and touching either container; although correct responses were recorded for touching the pointed-towards container within 30 s. Pet dogs were tested in only one session. All shelter dogs initially showed the expected poor performance. However, enriched shelter dogs receiving enriched human exposure showed significant improvements reaching an identical performance to pet dogs within 7 weeks. In comparison, shelter dogs under standard conditions showed an initial improvement, but performance reached asymptote close to chance levels and lower than that of enriched dogs or pet dogs. Together, these results suggest that enriched experiences with humans, typical of pet dogs, is sufficient for dogs to learn to follow points without explicit training.