To Fetch or Not to Fetch – Evidence Search Part III

Deborah Palman

Parts I and II of this series of articles published previously in Dog Sports covered reasons for teaching the passive indication of evidence and a description of the method I use to teach it. This article addresses some of the practical aspects of evidence search for police K-9s and trouble shooting training.

I work for a conservation law enforcement agency, and we encounter many situations where we have to search large areas for evidence such as expended shells, hidden firearms, and animal parts. K-9s have proven the most efficient way to do this. We spend considerable time training with firearm cartridges and shotgun shells, so our dogs find them readily even at older scenes. Comparatively little time is needed to train a dog to be proficient in evidence or article search, and it is an excellent way for a small department to get into using dogs without a big investment in time or money.

I start training article search as soon as a team starts in the training program. The method I use is based entirely on positive reinforcement and easy for the dog and handler to understand. It produces results very quickly and teaches the handler the basics of training with positive reinforcement (See the June, 1995 issue of Dog Sports). I also use articles heavily during tracking training because the articles serve as additional rewards along the track besides the end reward, and help keep the dog focused on the ground close to the path of the tracklayer. In actual situations, finding evidence on the track will sometimes solve the case when the track is so old that there is no hope that the tracklayer will be found or apprehended. At search and rescue scenes, finding something which belongs to the lost person will confirm that the dog is on the right track and not just a contaminating track. Finding such a "clue" (search and rescue terminology) can shift the entire search effort into one area or in a different direction and save much needed search time. I have found that bloodhounds can readily be taught article indication using a food reward, and I would like to see more bloodhound trainers take up this added training dimension.

Environmental Effects

The environmental and weather factors which affect article searches are the same as those which influence tracking and other types of scent work. Conditions which allow scent to rise and disperse to some degree without diluting it greatly are those which enhance searching with a dog. High winds are detrimental because they blow the scent away and dilute it before the dog can sample it. Light winds help because they spread the scent cone over a larger area for the dog to detect, and the dog does not need to search in a tight pattern. If the ground is warmer than the air, the track or article scent will rise and be available.

If the ground is colder than the air, the scent will not rise and will not be dispersed, so the dog may have to put his nose right on the article to find it. I am sure that all dog handlers are familiar with the fact that if they take an object out of their warm pocket (a freshly fired blank from a gun is also hot), a dog finds it readily because the heat of the object lofts the scent up out of the vegetation. This makes for a particularly impressive public demo, but the dog will not do as well on a week old crime scene where the articles are the same temperature as the ground.

Rain seems to have a detrimental effect on scent availability when searching for evidence. I remember a USPCA police dog certification where it started raining lightly after the first dog successfully found both articles during the article search phase of the competition. Every dog after the first, many of them very good at article search (including my first dog), failed to find more than one article. Once I tried to do a practice search during a heavy summer rain (in case you haven't figured it out, I train no matter what the weather is, and that's how I learn this stuff) and my dog could not find the articles even when I tried to show her where they were.

Returning to the same area five minutes after the rain stopped, the dog easily found all the articles. Apparently scent is driven downwards during a rain and is less available to the dog. Scent will also be less available when the ground is colder than the air and the sun is not heating up the ground. This means that early mornings and cold, cloudy days are not the best times to search older crime scenes.

Dogs can detect objects which are under water. The dog will probably not be able to indicate on exact location of the object, but if you are working near water and your dog shows a lot of interest in the water or repeatedly does "double takes" while searching near a pool of water, you should check the water. Metal detectors will work through water. This fall one of our teams was called to an area where a deer was shot illegally by four shooters firing approximately 14 shots. Arriving three hours after the incident, the K-9 found 10 cartridges, most of them in deep, snow covered marsh grass, but one was under four inches of water. The team could have found the remaining four, but they felt they had enough evidence for the case. In another case, this same K-9 located a cartridge which was six days old and covered with snow.

Contamination

Contamination by people, animals or vehicles puts the dog at a disadvantage. Vehicle exhaust temporarily destroys the dog's scenting ability. I have seen my dog stop dead on a track and circle and whine because he couldn't continue on. It took me awhile to figure out that the cruiser running nearby was the problem. Law enforcement officers love to keep their vehicles running, and they don't even think about it when a K-9 team arrives. I have also seen dogs shut down when the weather was calm and extremely cold (well below zero), and exhaust was being held close to the ground by an inversion layer. Otherwise I have not seen cold significantly effect a dog's ability to use his nose, if the scent is available.

To give the K-9 the best chance at finding evidence, officers at the scene should not contaminate it. I can go to a uncontaminated scene with my K-9 and not only find any evidence, but show where the violators or victims have been. Last year I was called to find a single rifle shell in four large hay fields about 100 acres in total. The wardens involved had apprehended two people night hunting on foot after they heard them shoot somewhere back in the fields. Luckily they had not searched the area themselves, probably because they knew how difficult it would have been, even with a metal detector. Starting from where the two were apprehended, I put my dog in her tracking harness and told her to track. She started tracking loosely through the fields. The scene was about nine hours old and an intense morning sun was warming the ground after a cold night.

We tracked along about 200 yards or so through two fields and into the third where she nonchalantly laid down, indicating on the rifle shell. Even I was amazed. Finding the shell took five minutes from the time I arrived at the scene, and that included giving water to the dog. In another five minutes the dog found the deer that was shot and left. After digging in the deer carcass for fifteen minutes, we completed the picture by recovering the spent bullet from the deer. As one of the wardens said when we pulled the bullet out - "Touchdown!"

When a scene is contaminated, the dogs seem to have to scan the area first for the fresher scents, and then go over it again to locate evidence. Give the dog plenty of time to check the area out, and don't expect the dog to locate evidence on the first few passes through a contaminated area. The same rule applies to areas where an animal has been killed or gutted. A well trained and motivated evidence search K-9 will be able to handle this if given the time to check the kill and scents out before going to work.

Practical Training and Trouble Shooting

Basic instructions for teaching the passive indication using motivational methods were given in a previous article. Since writing that article, I have added the use of conditioned reinforcer to the process and find that the dogs learn much quicker, and when properly used, the handler can avoid the problems caused by the timing of the reward. A conditioned reinforcer (CR) is a signal which tells the dog it is correct and the reward will be given for what it did. This is a complex subject, and I plan to write a more detailed article on it in the future. The advantage of the CR is that it eliminates the timing problems when it takes time to get the reward to the dog. This happens when the dog is working at a distance, or when a handler is trying to fine tune behaviors so that they are very precise, such as in competition obedience.

One simple example would be teaching a dog to indicate on narcotics. The dog is motivated to find the narcotics because he gets to play with a toy as a reward. If the handler is pulling the toy from his pocket, at first the dog may not notice this because he is concentrating on the odor of the narcotics. As training goes on, however, the dog begins to associate the finding of the narcotics with the reward coming from the handler. When that happens he finds the scent, indicates by scratching (if that is the desired indication), then looks to the handler to watch the handler pull the toy out. If the handler continues to pull the toy out at this time, the dog may think that the reward is for finding the narcotics and looking at the handler, not the scratching. With time, the scratching behavior fades and the dog just stops searching when he hits the narcotics odor and looks at the handler.

This problem can be solved by using contraptions and hides set up so that the reward comes from the scent source. It can also be solved more simply by adding a conditioned reinforcer. The CR is initially trained by giving a specific signal such as a praise word and then immediately giving the dog a reward (the more and varied the better). This word or noise should be something that is not used randomly in training or speech but reserved for training the dog. I've gotten away from "Good" because I use it too much without thinking about it. "Yes!" is a good word to use. The dog comes to associate the word with the reward, so, in effect, the word takes on the same meaning as the reward. Applying this to the narcotics training example, the drug dog trainer first teaches the CR by exclaiming "YES!" and throwing a toy or playing with the dog.

This is repeated many times until the dog visibly reacts to the word without the toy or reward being present. Then the training sequence (it is assumed here that the dog already knows to find the scent) becomes search, dog finds scent, dog scratches, "Yes!" is given the instant the dog scratches, and the toy comes out. The CR allows the trainer to mark the instant the dog is right and gives him time to present the reward. Another problem encountered by handlers of highly motivated narcotics dogs is destructive behavior. They want the dog to indicate, but not necessarily eat the furniture in the process. Using the "Yes!"the instant the dog scratches reinforces this behavior.

Waiting until he bites is too late, and sometimes this happens if the handler doesn't reward quick enough. If the dog bites, this behavior cannot be rewarded, so during training (on training furniture, not your own good stuff) the handler should wait until the biting stops and a better indication starts before using the "Yes!" and reward. The dog will learn through experience that scratching produces the reward, not biting.

The same technique helps keep passive indication dogs from picking up the evidence or moving and shifting from position before the handler arrives with the reward. Giving the "Yes" when the dog first touches the article with his nose stops the dog from proceeding further because he is now looking for a reward because he knows touching is the right action and he doesn't have to pick it up. The indication can be refined by presenting the CR when the dog looks at the article but does not touch it.

This produces a dog which looks at the article, looks to the handler, then looks at the article again. Of course you can add the down position to this equation, but these are some of the refining behaviors that can be taught the dog. There is also nothing wrong with giving the CR when the handler is at a distance from the indicating dog, then giving the dog a release command and letting the dog come to the handler for the reward. The CR marks the behavior wanted, so it doesn't matter where the reward comes from or what the dog does when he actually receives the reward. In some training methods, the CR is a release command, so it doesn't matter what the dog does after the CR is given.

I've found that many of the dogs who are taught the passive indication still try to dig out small articles when they are hidden in deep grass or snow, or pull articles out of trees if they are hidden up high. They apparently want to lay down by them as taught, and figure out that the best way to do this is to move the article to the ground surface. My current dog will pull down articles and then lie down in a perfect position with the article between her front feet.

I'm not sure this is faulty because the handler would have difficulty locating the article when it is well hidden if the dog did not dig it out. However, but mouthing the article defeats the purpose of not disturbing the evidence. A compromise would be to watch the dog closely when it starts to dig or jump, and stop the dog before it touches the evidence, as those who do a fetch or active indication do.

False Indications

The passive indication is also more prone to false indications than a fetch or active indication, although I have seen fetch dogs start picking up old litter and other objects when frustrated or pressured by their handlers. False indications are created when the dog becomes frustrated or confused, reads the handler's body position and figures that the handler wants him to indicate there. This is caused by poor training, or by residual scent left by articles or people in the area. If someone digs in the grass to try to help the dog, they create a scent hot spot which smells just like an article. How is the dog to know the difference?

False indications can be fixed by setting up problems where the handler knows where the evidence or scent is and rewards only for indications on that location, ignoring any false indications. Handlers also have to watch their body language and strive to perfect an indication by the dog when they are giving no signals such as looking at the dog or ground, bending over, etc. Ideally the dog should be proofed to indicate even when the handler has his back turned.

Maine Moose Season

Often a single piece of evidence makes the difference in a case. Moose hunters in Maine are allowed to shoot only one moose, but some mistakenly or intentionally shoot two or more because moose tend to hang out together during the fall. Every year during and after the legal moose season, Maine Wardens find more than a few scenes where two moose have been shot and one or both left. This fall during the season, a Warden Sergeant and his K-9 were called to a scene where two moose were killed and left in a clear cut. Of course, before calling the K-9 team, four wardens searched for several hours for rifle shells without results. A witness had given the wardens a general description of a vehicle seen in the area. The K-9 team searched for a half hour and located one spent .280 Win. rifle cartridge at the bottom of a brush pile.

On a wood road on the route back to town, the Sergeant saw a truck parked beside the road which matched the description of the suspect vehicle. He stopped and talked to the occupants, a man and his wife, and did a routine check of the weapons (it is illegal in Maine to have a loaded firearm in a vehicle), finding that the rifle in the truck was a .280 Win. The Sergeant told the man that he wanted to talk to him and took him to his vehicle. The Sergeant got in his truck with the man, looked at the man and said, "Now, don't you have something you want to tell me?" The man began to cry and admitted to shooting the two moose by mistake. He and his wife had tried to load one, but when someone drove by, they became frightened and left, picking up all the shells they had fired but the one they could not find. That was the one the K-9 located at the bottom of the brush pile.