Saturday, June 27, 2009
A professor and a member of the San Bernardino police department determined to test the tracking abilities of bloodhounds that had received more than 18 months of training against bloodhounds that had received less training. They obtained four dogs in each category and performed the following test. Subjects down trails from a half mile to a mile and a half long in five separate areas at five separate times. Two of the areas were in regional parks, a third area was a college campus, and two of the areas were in downtown San Bernardino. The subjects laid down the trails 48 hours before each dog was brought to the starting point of a separate trail. The subjects were given maps which showed them where to begin and where to end. With each subject was a partner whose scent would not be given to the dog. About 50 feet from the end point of a trail, the subject and the partner would separate and find a place to hide behind an object, such as a tree or a building. After standing in this location for about ten minutes, the subjects and the partners were driven away by a car which did not cross the trail they had laid at any point. Two days later, when the dogs were brought to the beginning point of each trail, the subject and the partner were brought back to hide in the locations where they had previously stood for ten minutes. Neither the handlers nor the researchers watching the dogs work knew where the trails went or which subjects the dogs were supposed to alert to. Scent had been taken from each subject using a scent transfer unit (STU-100, see picture), which put the subject’s scent on a gauze pad. The dog was allowed to smell the gauze pad (with no other contact with the subject) at the beginning of the trail, then commanded to TRAIL. Most of the trails had been crossed by hundreds if not thousands of people between the time the trails were laid and when the dogs began to track. At one regional park, a trout fishing contest meant more than 1,000 people had crossed the trails. The trails on the campus and in downtown San Bernardino involved tracking on cement and asphalt. Rain had fallen before the first trails were laid. Of the 20 trails that the younger, less experienced dogs followed, 12 were run to completion, with the subject found by the dog. One dog identified a partner rather than a subject in one trial. Of the 20 trails followed by the veteran dogs, however, 19 were run to completion with the subject alerted to. The researchers noted the following concerning the single veteran dog failure. “The one veteran bloodhound that did not make the find was distracted by a foul smelling dumpsite that was located about 300 yd off the trail. The handler was not able to keep the dog’s attention on the trail and decided to stop his dog. Once the dog was walked a substantial distance from the dumpsite, she continued the trail and made the find.” The researchers suspected that the poorer performance of the younger dogs was in part due to their less mature neurological systems. The two youngest dogs in the experiment were only 10 and 11 months old, respectively. Within three months of the test, both had improved to a level of 100% on subsequent tests. Lisa M. Harvey and Jeffrey S. Harvey, “Reliability of Bloodhounds in Criminal Investigations,” 48(4) Journal of Forensic Science 811-816 (July 2003).
Tuesday, June 23, 2009
A German research team was asked by the Hamburg Police Department how long a dead body would have to have contact with a mattress or a carpet for a cadaver dog to detect that a body had been on the item. This question arose from a real case involving a married couple that went sailing in their yacht. The husband returned alone and reported his wife missing, but the police soon regarded him as a suspect. They brought a cadaver dog onto the yacht, which alerted in the cabin. Although this would not have been enough for a conviction, the researchers argued that it justified further investigation. They designed an experiment to answer the question of the Hamburg police. Two men who had died only two hours before were wrapped in cotton blankets. Carpet squares were placed under their backs, touching the blankets but not the bodies. The squares were left under the men for either two minutes or ten minutes, then removed and placed in sealed containers. Carpet squares were also placed under living men, also without direct contact. The carpet squares were then used in a test of the skills of three cadaver dogs of the Hamburg Police Department, two Malinois and a herding dog. One of the Malinois had five years of experience, while the other two had about a year and a half. They were tested the day of the exposure, but up to 35 days later for the two-minute exposed squares and 65 days for the ten-minute exposed squares. The dogs sometimes failed to identify the two-minute squares, but two of the dogs were perfect on the ten-minute squares. Collectively the dogs alerted accurately 86% of the time on the two-minute squares and 98% of the time on the ten-minute exposed squares (with only one dog have any incorrect alerts). The message to the criminals apparently is, if you’re going to move a body, do it quickly or the dog will find you out. L. Oesterhelweg, S. Krober, K. Rottmann, J. Willhoft, C. Braun, N. Thies, K. Puschel, J. Sildenath, and A. Gehl, “Cadaver Dogs—A Study on Detection of Contaminated Carpet Squares,” 174 Forensic Science International 35-39 (2008).
Additional Note. A recent article discusses a dog trained on detection of bones hundred of years old. A scientific evaluation of this report would be useful.
Thursday, June 18, 2009
A few blogs ago, Liz Stroter talked about one of the reasons that people do not adapt to dogs and give them up to shelters: the expectations that are created by televisions shows that a stern hand and a few rough words will change the dog’s behavior in a matter of weeks. When the expectations are not fulfilled, the dog goes to the pound. Another reason people give up on dogs is that TV and film depictions lead people to believe—particularly people without experience in raising dogs—that dogs are merely rather simple happy people who cannot speak. This has the effect that many people may believe that a dog that is destructive of furniture is not only mischievous, but actually bad. If the dog is continually bad, perhaps he is evil. This means that leaving the dog at a shelter where he will ultimately be euthanized may actually be deserved. “He was just a bad dog. There was nothing we could do about it. We tried. We really tried. George took him out for a walk twice a day. He didn’t care. He still pooped on the kitchen floor in front of the refrigerator. He just didn’t care.” Recently on a transcontinental flight I watched Firehouse Dog. Aside from the improbabilities of a typical Hollywood story, a dog that is adopted by a firehouse is taken to an agility trial, where without any apparent experience, he runs through most of the trial following the commands of a boy who has never worked with him, performing each feat flawlessly. This is not a good lesson for a teenager who has been given his or her first dog. Hollywood depictions of animals has become something of a sociological topic. See Marla V. Anderson and Antonia J.Z. Henderson, “Pernicious Portrayals: The Impact of Children’s Attachment to Animals of Fiction on Animals in Fact,” 13(4) Society & Animals (2005).
Sunday, June 14, 2009
For some years, my consulting practice has allowed me to live in Phoenix in the winter and spring and Stone Ridge, New York, in the summer and fall. This creates a problem in volunteering with Chloe for therapy dog assignments. There comes a time when we must say good-bye for six months. I was particularly worried about this for the nursing home we visit in Youngtown, west of Phoenix. Many of the residents are hospice patients (we visit through the auspices of Hospice of the Valley, a large non-profit hospice system in Arizona), which means that many of them are not expected to live for more than six months. When we return after Thanksgiving, some of Chloe’s friends will no longer be there. Her biggest fan in the home is 99 years old, a veteran of several wars. Another is a woman of 105. In reading one of the earlier studies on pet facilitated therapy in nursing homes, I got a sense of the quantitative effect of our visits, and of the quantitative effect of our ceasing to visit. Ira B. Perelle and Diane A. Granville, Assessment of the Effectiveness of a Pet Facilitated Therapy Program in a Nursing Home Setting, 1(1) Animals & Society (1993). Perelle and Granville used four cats, two small dogs, and one rabbit, and visited 18 men and 35 women in a nursing home for ten weeks. They measured the sociability of the residents using what they called a Patient Social Behavior Scale designed for the study. The scale attempted to measure various social and self-maintenance behaviors. They found that the patients visited improved in their sociability while the pets were visiting, but men improved quickly and women more gradually over the period of the study. Unfortunately, all declined after the study was completed, though not back to the point where they had been before the study began. Eleven residents had to be dressed by others prior to the start of the study, but only three had to be dressed by the end of the visitations. This number rose to seven again in the four weeks after the visitations were over. The staff of the home were particularly impressed by how men interacted more with other residents because of the study. My experience has also been that men often sit in their wheelchairs at favored places in the hallway but do not interact. The women go to the activity room and interact with the staff and each other. Chloe often became a subject of conversation between the men and sometimes seemed to allow them to talk to each other. Perelle and Granville noted that they were able to conduct their study because of a cooperative administration and staff. They said something that caught my attention at the end of their paper: “Experience and the literature indicate that most institutions are not at all amenable to the introduction of animals of any type.” They wrote this in 1993. I doubt that this statement would be made now. The nursing home where I take Chloe in Arizona has been trying to find a replacement for Chloe for the months we are not there. Unfortunately, there is much more demand for therapy animals now than there is supply.
Wednesday, June 10, 2009
A man crosses a football field. Twenty minutes later a tracking dog is brought to the midpoint of the trail and commanded to TRACK. The dog will sniff the area, perhaps back-track a few steps, and then move in the direction that the man walked across the field. How does the dog know which direction the man was moving in? How many footsteps does the dog need to be able to smell to determine the direction the man was walking? Two researchers in Belfast asked these questions and devised an elegant experiment that may have narrowed the possible answers to the first question. As to the second question, the dog cannot reliably tell direction if there are only three steps but it can do so if there are five. Here’s how the experiment was conducted: A dark beige wool carpet was cut into 18-inch squares. Twenty-one squares were lined up and a man walked across them, stepping once on each square. As in the football field, the dogs were brought to the midpoint and told to TRACK. Neither the handler nor the dog had seen the man walking on the squares. The dogs, which had been trained as trackers, reliably went in the direction the man had gone. Two squares were removed so that there were only 19 squares. The dogs were still reliable. Then two more were removed, and so on. The dogs retained their ability to detect the direction of the trail down to five squares, but at three their choice of direction became random. They needed more than three footsteps. Could it be the direction of the footsteps that the dogs used to determine direction? After all, the heel hits before the toe and perhaps this provided a cue. So the researchers mixed up the squares, using a computer program for randomness, but kept the direction of the heel-toe axis of each square facing the same direction as the man had walked. The dogs were not successful in finding the correct direction. They did not seem to be able to rely on general body deposition (the plume of skin flakes, etc.) to determine the direction of the target. The researchers theorized that there is an odor gradient that results from either a stronger smell for more recent tracks that gives the direction to the dog, or alternatively, there is a decay factor which is stronger in the footprints from the direction that the man began at, which the dog used as a signal to move in the opposite direction. The researchers are working on additional experiments to answer more questions about the tracking skills of dogs. D.L. Wells and P.G. Hepper, “Directional Tracking in the Domestic Dog, Canis familiaris,” 84 Applied Animal Behaviour Science 297-305 (2003).
Saturday, June 6, 2009
The first formal service dog organization in the United States was The Seeing Eye, founded by, among others, Dorothy Harrison Eustis and Morris Frank. Eustis was training German Shepherds in Switzerland after World War I and wrote an article for the Saturday Evening Post (“The Seeing Eye,” November 5, 1927; the picture on the left is from the original article) about a program of the Shepherd Dog Club of Germany, which was training dog guides for blinded veterans of the Great War. Morris Frank was the first American for whom she trained a guide dog. Her article is famous for describing how guide dogs work, but it is also worth reading, and rereading, for its description of the effect the dogs have on the blind men who were paired with the dogs after they had been trained by skilled trainers. She describes how the initial encounters can be disappointing.
"In the four months of school [the dog] has become attached to his teacher and works perfectly for him and he is puzzled and thrown off by the exchange. The first days with the new master are difficult. The blind man is nervous, distrustful and supercritical, as well he might be. The dog works unevenly, often looking back at his old teacher, and the blind man has a disturbed mental picture that this is the way he is always going to be led and he states his opinion in no uncertain terms."
Then both man and dog beging to adapt to each other:
"Gradually the rehabilitation takes place. First, the uncertainty becomes less uncertain, a glimmering that perhaps here is eyesight; then the acknowledgment that here at least is ever pleasant, ungrudging companionship and protection. Then the putting out of feelers: "Can this really mean Independence?" And then comes the whole great realization that the future holds freedom. No longer a care and a responsibility to his family and friends, he can take up his life where he left it off; no longer dependent on a member of the family, he can come and go as he pleases; and as these thoughts and possibilities gather strength in his mind, despair and loneliness give way to happiness and companionship, and these qualities can be seen developing from day to day.
"A comparison of the men completing their course with those just commencing is the proof. The men arrive forlorn, with lined, anxious faces and drooping bodies, thin or over-fat from inertia. In four short weeks they are remade; life takes on a new interest; shoulders lose their droop, backs straighten up and feet forget to shuffle. The thin have won back their appetite through their daily exercising walks and have put on weight and muscle, and the fat ones have trained down. Occasionally, a chuckle is heard which is the opening wedge for a laugh, just as the birds' early morning twitter presages the full song to the sun."
The final paragraph of the article is the trumpet call to action that in many ways led to the creation of The Seeing Eye.
"No longer dependent on a member of the family, a friend or a paid attendant, the blind can once more take up their normal lives as nearly as possible where they left them off, and each can begin or go back to a wage-earning occupation, secure in the knowledge that he can get to and from his work safely and without cost; that crowds and traffic have no longer any terrors for him and that his evenings can be spent among friends without responsibility or burden to them; and last, but far from least, that long, healthful walks are now possible to exercise off the unhealthy fat of inactivity and so keep the body strong and fit. Gentlemen, again without reservation, I give you the shepherd dog."
We shed 30,000 to 40,000 skin flakes each minute. (Picture shows electroni microscope image of a desquamated skin flake, at least a thousand of which you'll shed by the time you finish reading this piece.) Clothing is permeable to this particle stream and it comprises between 70% and 90% of house dust. These skin flakes spread up and out from the body by heat and air currents in a pattern known as a plume, and are a major part of the scent trail that a tracking dog follows. (See Gary S. Settles, “Sniffers: Fluid-Dynamic Sample for Olfactory Trace Detection in Nature and Homeland Security—the 2004 Freeman Scholar Lecture,” 127 Journal of Fluids Engineering 189 (March 2005); Robert Hunt, “The Benefits of Scent Evidence,” FBI Law Enforcement Bulletin, p. 15 (November 1999).) How the olfactory systems of dogs work is a study in itself, as are the fluid dynamics that describe the dissemination of the plume—how far and how quickly it spreads. Dogs can generally find scent trails within 48 hours after the target has been in an area but much longer periods have been reported. An FBI newsletter describes a case where a bloodhound followed a trail from a car abandoned 17 days earlier to a suspect’s apartment building. This dog was even able to follow the 17-day old trail into a commuter station to a kiosk where the suspect caught a bus. The dog was thus able to ignore the scents of thousands if not tens of thousands of commuters in tracking the suspect. (Guy J. Hargreaves, “Detection Dog Lineup,” FBI Law Enforcement Bulletin (January 1996).) Another area of research on dog scenting concerns the dog genome. In 2008, a team of Polish scientists determined that variations in olfactory receptor genes “might affect the olfactory ability of service dogs in different fields of specific substance detection.” Twin dogs with a specific gene variant were able to track better than other dogs with different variants of the gene. (A. Lesniak, M. Walczak, T. Jezierski, M. Sacharczuk, M. Gawkowski, and K. Jaszczak, K., “Canine Olfactory Receptor Gene Polymorphism and Its Relation to Odor Detection Performance by Sniffer Dogs,” 99(5) Journal of Heredity 518-527 (September/October 2008).) Thus, it might soon be possible to predict, both for cancer sniffers and other detection dogs, whether a puppy is a good candidate for such work. It might even be possible to select for genetic predisposition to high tracking skills in breeding programs. Whether that kind of dog breeding will be a good thing I am not sure, but I have no doubt that someone will make it part of a breeding program sooner or later.