Crime Library: Criminal Minds and Methods

K-9 Forensics

Hunting Bodies

According to the authors of the Cadaver Dog Handbook, locating the remains of a missing person can be "one of the biggest challenges a forensic investigator may face."

Sometimes the death was particularly brutal, the discovered body is that of a child, or the victim has been missing a long time and the remains are in bad condition.

Cadaver dog searches water (AP)
Cadaver dog searches water (AP)

Occasionally the terrain to be searched is difficult to navigate or the corpse has been carefully hidden.

In fact, it was the presence of decomposition odor that led cadaver dogs in Philadelphia in August 2001 to the body of Kimberly Szumski, 36, a woman missing from her home three months earlier. She had been wrapped in plastic and duct tape, buried under cinder blocks, and cemented into a wall reinforced with steel bars. Nevertheless some odor escaped, throwing suspicion on the estranged husband, who had done construction work in that building. The dogs gave the signal, the walls were torn down and the body was located.

Unlike trailing scent dogs that stick with an odor on the ground, cadaver dogs find the scent in the air as well. The use of such dogs in police work began in 1974 for victims that had been buried in a forest. A dog was brought from Texas to New York for the task, and her first find, as documented in The Cadaver Handbook, was a college student buried under four feet of dirt. Now police agencies train dogs for this and there are numerous volunteer search dog teams as well.

Sandra Anderson from central Michigan handled Eagle the dog in the Kupaza case. Eagle was also featured on an episode of Unsolved Mysteries in which he was taken into the home of Azizul Islam, a physician suspected of killing his wife. The woman went missing on December 20, 1999. Two days later, parts of a woman's body were found in a dumpster and a field. Apparently feeling confident, Islam allowed a search of his home. Anderson noticed strong bleach odors everywhere and she feared that this would hinder Eagle's work. However, the dog immediately gave the signal on a paint roller and soon indicated that there was something associated with human decomposition on the floor that the doctor had been painting. It turned out that there was blood in the paint and it matched the DNA found on the missing woman's toothbrush. Islam was arrested and convicted of murder.

Anderson offers demonstrations of Eagle's uncanny abilities, and after one such demo, she agreed to an interview.

Her work began when she was just starting to decide what to do with her life. "I actually began assisting in the training of narcotics dogs as early as 1978, when I was 18," she recalls. "I needed some extra money and I'd grown up with dogs—my father was a veterinary technician and one of the founders of the anti-cruelty association in Detroit, Michigan. Of everyone I've ever met, my father has had the most impact on my life because of his incredible compassion toward all creatures great and small.

"From there, I continued into narcotics training and obedience work. Around 1984 or 1985, the Mexican earthquake hit. It was a change for me because what I heard being discussed was the dogs that specialized in finding live people. God bless 'em, there are still dogs that do that. And I remember thinking, you know, we have dogs that only do narcotics. Perhaps we should try to see how well a dog might do with only detecting human remains. That was the beginning for me. I went to seminars and I ended up taking courses from wonderful people in Search and Rescue, as well as law enforcement and the military. They've all had an impact on where I am today."

Eagle is her third dog to specialize in human remains, yet she's assisted in training hundreds of other dogs as well. "Right now I'm working with three dogs of my own, in the hope that one of them will be successful. We've got Gator, who likes finding birds more than he likes finding human remains, so we don't know if he'll be successful. And I have a wonderful golden retriever crossbreed and a German shorthaired pointer. They're all neck-and-neck, although I'm seeing a little more genius out of the retriever."

While she does a lot of training at home, she also travels to meetings with anthropologists, pathologists, coroners, and medical examiners around Ohio, Michigan, and Wisconsin to continue her education, and she's the director of the detection division for Canine Solutions International, (, located in Charlottesville, Virginia. "It's an educational and service-oriented company, specializing in consultancy work," Anderson explains, "as well as offering specialist canine services to law enforcement and other agencies, like narcotics, explosives, executive protection, forensics, disaster, and missing persons. Death investigation for me is a fulltime job and sometimes it overlaps into the weekends. It's a commitment that one doesn't walk away from just because the clock or calendar says it's a certain time."

Asked what traits she looks for in a dog with potential for this work, she lists:

  1. A sociable temperament. "I don't have the time, energy, or respect for a dog that is unsociable or shows any aggressive qualities. I'm not adverse to mixed breeds. Twenty years ago I think the pedigree meant more than it does these days. There are reputable breeders but there are also breeders who aren't. The papers and the breeding do not make a dog a better search dog."
  2. High hunt drive. "If I throw a ball in some tall grass and the puppy won't come back until he finds that ball—-even at eight weeks of age—-he'll make a good search dog."
  3. Focus: "While he's hunting for that ball, if a car backfires or I clap my hands, he won't care. He's got a job to do."
  4. Retrieves a toy. "That means I've got a reward system already. That tells me the dog has possessive qualities. And 'though I need to teach the dog to possess scent, it's a head start because he's already showing me that he has that quality."
  5. Methodical. "I don't necessarily want a type A dog and I don't want the laziest dog. I don't like the dog that's hyperactive or goes into overdrive. I may choose the puppy that I notice is always sniffing around on the ground vs. bouncing off the walls."
  6. Observant: "I don't ignore the dog that's sitting under the table watching everything. There's a lot going on with a dog that watches other dogs."

She feels strongly that a dog needs to have these skills and to want to give its best effort. "With the level that I work at, it's a needle in a haystack. It's difficult and time-consuming, but the victim deserves better than adequate."

To test for these traits, Anderson will do several different things. "Naturally we'll use a toy like a ball to test for hunt drive. We'll just walk among the puppies and clap our hands, and if any run away or don't hunt for the ball, we put them aside. They won't have the traits we're looking for. If any puppies are shy, they won't have it. We can't make that better.

"Then, once we get the litter down to one or two—-and sometimes none—-we bring out a large box containing very smelly human remains in some state of decomposition. [She gets this from an anthropologist or pathologist.] Then we watch the dogs for a certain attraction or for being averse, which a lot of puppies are. We won't even look at a dog like that again, because it's not going to outgrow that. If we have a dog that smells the box and that doesn't want to leave it, even if we make noise or walk away—-if he's obsessed with it, we probably have a winner."

Asked her opinion of the use of synthetic cadaver scent that many trainers employ, she's dismissive: "I think that until we're out there looking for pseudo-people, we're kidding ourselves using those chemicals. As a matter of fact, anytime there's a chemical that I can use to start a fire, something's wrong, because I've never been able to start a fire with human remains. I have a real problem with it, and I think that a lot of dogs have problems in training because we've become such a fast food society: Whatever is the quickest way to train the dog is the way we go. That's not a good idea. I know, from experimenting with numerous dogs over the past decade, that if I put out cadaver chemical pseudo-scent and I have an arson dog along, the arson dog will find it. That tells me that something's wrong.

"I expect my dog to be a human remains specialist, so in the training program and throughout our lives together, I'm not going to let him down by not being a specialist, too. That means he will only detect human remains and not something else. Nothing else is satisfactory. People who use synthetic scents aren't ready to train a dog to this level, because you need to be able to work with highly educated forensic specialists that can assist you in all ways. If you aren't ready for that, then you need to do a lot more work to get there."

The way she gets a dog to imprint on a specific type of scent is to fall back on the Pavlovian technique of association and reward.

"The puppies are in a crate," she says. "It's time to eat. Twenty minutes before they're due to eat, we bring human remains out in a sterile paint container or jar, and put it on the crate. It doesn't take long for them to associate that smell with getting fed. Once we've got that down, then the puppies get fed regularly. But then we get a new toy. It might be a towel or a squeaky ball. Ten or twenty minutes before we let them out to play with it, we bring out the scent. This will go on for about six or seven months. Then we begin to teach them the search process. That means they have to hunt for it and then something good happens.

"Then in about a year, we go back to imprinting on skeletal remains. Then at two and a half, we start imprinting blood and trace evidence. That's it in a nutshell. It's a very basic procedure."

When on a case, Anderson takes the dog out and looks for specific behaviors that indicate the dog has found the target scent. "I expect a bark. If I don't have a bark, he's not on the target. That's key to everything. Excitement levels dictate the strength of the odor. It's like a good hunting dog. A big bird excites the dog more than a small one. Dogs are honest creatures. They can't contain their excitement if the smell's really good. Then there's also the ease with which he works on a problem; if he goes into a room and within two minutes he can tell me there's scent, I know there's probably a great deal of odor present. If he's working hard, then there's not a lot. With skeletal remains, because we teach the dog that they can only possess the scent and not the object, he'll go down [adopt a prone but alert position]. He knows he can't touch anything, so he has to go down because that's how he controls himself. Preservation of the remains is everything, so they have to be able to hold the point. But he's got the scent and then there's the anticipation of 'What are we going to do now?'

Anderson says that the dogs are in constant training. "I might come home at night from a job and go out and work them for an hour. It's obsessive. There's no end. There's scent work, and then obedience and manners, and handling planes, trains, and automobiles. He has to learn to be a good dog while someone's eating dinner, and when a place is noisy. There are so many parts to the training process."

Given how complex and demanding the training is, inevitably there are problems.

"Most dogs fail the program because they can't differentiate. At around one year of age, we start introducing distractions, sometimes earlier depending on the dog's progress. Some dogs just like dead raccoons better than bodies. Some would rather find a hotdog than a body. Some would rather find a tennis ball. So when we encounter any of that, that dog's probably not going to make it through the training. Usually if a dog's in the program as long as a year and a half, we know he's good to stay."

And then there are the handlers. They, too, have to be trained.

"It can be difficult. A lot of them don't do it as a career. They do it because they're on a search-and-rescue team and they want to get called out with their dog once in a while. Yet I've met numerous passionate individuals who work very hard at it and are able to accomplish some very good work with their dogs. They're the kind of people who are willing to say, 'The victim deserves better than adequate. If this dog doesn't make the cut, I'll get another dog.' At no time do we look at our partners as pets; we see them as friends and as working associates. We never own the nose. We share the nose. We translate. We're the communicators. We're the folks who are trying to transfer what the dog is telling us to the public at large, and we have to be very good at that."

During training, some people "clue" the dog, or give signals to point the dog in a certain direction.

"If a dog can be clued," Anderson says, "it's already the wrong dog. If you have a dog that we call a deceiver or a dog that is more into getting the toy or pleasing the handler than into hunting, you already have the wrong dog and you need to find one that is honest, confident, and above all else loves hunting and the odor of human remains.

"If you look at a really good hunting dog, there is nothing in the world that is more important to that dog than finding that pheasant or quail. You could have a tennis ball fall from heaven, but nothing is more important than finding the bird. We're hunting human remains, but we still have to have all the same energy and desire as those hunting dogs."

She points out that there is a difference between an ordinary cadaver dog and a dog like Eagle, who is a human remains specialist.

"A human remains specialist is not a cadaver dog. Calling it that is like calling the best investigator in the world a security guard. A cadaver dog will find anything that is actively decomposing. A human remains specialist will only find human remains, not just in the decomposition stage, but also the skeletal and trace evidence stages. I call Eagle a death investigator dog. He lost the title cadaver dog at four months old.

"From among my personal choices, a dog like Eagle is about one in a thousand. It's like finding a great painter or concert pianist. In three years, I've probably tested over thirty dogs for myself. None have made the cut. There are thousands of cadaver dogs around the country, but if you want to know about human remains specialist dogs, from very strict testing procedures we know that among dogs that have been evaluated by my company, CSI, we've passed only eight rock solid dogs. Not every dog can do this, but the need is great, and for that reason, we cannot lower the standards.

"I feel strongly that when people lower standards by telling an agency they can do something that they can't, they give a family false hope. Saying you 'might' be able to do it is cruel. I've seen where dog teams have come through and told an agency that an area is all clear, meaning there's nothing there. And then guess where I find it? Right there where they said nothing was there.

"To thine own self be true. Don't do a job that you can't do. If you have a cadaver dog, great. That's important. But you should always be progressing it. If you can't progress it, then be honest with the agency using you. Many agencies are frustrated with dog teams because people have misrepresented themselves. I have yet to meet a forensic pathologist or anthropologist that anyone's going to kid. They're sharp; they know what they need and they know what to look for."

Eagle has been on the job for over eight years. "We've done probably close to 1000 cases, and last year alone, we did 177." His first case, when he was just over ten months old, was finding a drowning victim in Michigan's Muskegon River. "The only reason he was used is that the dog I had then was old and getting tired. We'd been out for three days, and the sheriff asked if I could use Eagle. I said he was only a pup, but he figured we had nothing to lose. Within about 15 minutes, Eagle had run up this trail to this little cliff, literally jumped off the cliff and dropped about twelve feet into the river and started barking. And there was the body. At that moment I thought, wow, he's really good! In water, he was a natural."

When asked to explain how a dog can scent a body in water, Anderson said, "We have to remember that the oils from the body come to the surface. As it decomposes, there's a great deal of odor that comes up, and at some point the decomposition gases began to rise. If there's been a stabbing, we even have blood odor."

Her most difficult case, and the only one in her recollection that she has not solved, was the case of a missing girl in Michigan. "It's going on the fifth summer now and I still have not found her body. That's why it's the hardest. It plagues me. I've never discovered any clues. If I find even one clue or suspicion, I consider it a good case. On this case, we had a murderer in prison that admitted it, but he couldn't remember where he put the body. I spend my days off, infrequent as they are, trying to find her. I'm very committed to her family. She's a victim and her parents need closure. I have the talent and means to give them that. No one else is going to give them that body. It's going to take someone just finding her while they're out walking, or the dog."

No matter what line of work these dogs are trained for, it's clear that they've become essential to forensic investigation.


We're Following
Slender Man stabbing, Waukesha, Wisconsin
Gilberto Valle 'Cannibal Cop'