Shallow Graves

by Maureen Boyle

  Before joining the state police, Andy had worked at a lumber company after returning home from the U.S. Army, where he worked in military intelligence. He was looking for something different to do at the time and joined the volunteer Auxiliary State Police in Connecticut. In that capacity he rode with troopers and got a taste of state police life. He liked the pace, the excitement, and the idea he was helping people. Eventually, he was accepted as a full-time trooper at age thirty, went through the rigorous academy and, after working as a road trooper, joined the state police canine unit. One of his search dogs, a bloodhound named Clem, had a good nose for finding people and received an award in 1977 from the National Bloodhound Association. But Clem had a big problem when it came to searches. Clem didn’t like dead people. Since a number of missing individuals turn out to be suicides, Andy trained his next dog, a German shepherd named Rufus, specifically to find the dead. As the years passed, he would train more dogs.

  Training a cadaver dog takes time, patience, and a working knowledge of scent. The handler needs to know how scent moves through the air, comes up through the ground and how best to guide the dog at the scene. Handlers use the chemical equivalents of decomposing bodies, cadaverine and putrescine,10 to reproduce what the dogs need to find in the field. Sometimes they use soil samples from sites where bodies were found. Early on, some handlers used bones or pieces of decaying flesh. (The son of one Massachusetts state trooper once said he learned as a child to always read the labels carefully on the Tupperware containers in the refrigerator because you never knew what was inside.)

  How much of a smell a dead person gives off depends on the level of decomposition. The freshly dead likely won’t be detected by the human nose, but an animal, such as a search dog, even one not trained to find cadavers, may react to the scent as if it were a person very much alive. But as time passes, the body changes. It bloats and decays. You can smell the stench from a distance. Eventually, the dead begin to liquefy as part of the decaying process and the body finally dries out. A cheesy or musty smell lingers in the aftermath. By the time a body is skeletal or mummified, you may still catch that musty whiff near the remains.11 For the search dog, it is what is left behind that is crucial. The chemicals from a decomposing body seep into the soil, leaving a telltale scent for the well-trained dog. A body, even skeletal remains, can leave clues for these highly trained dogs. The handlers are the ones who make sure the dogs can find those clues.

  By the time he came to New Bedford, Andy had already spent eleven years training search and cadaver dogs using what was called the play-reward system. When his dogs found something, they ran to him, nudged a tennis ball in his pocket then led him to the body. Andy would then bring the dog to a safe area and play ball as a reward.

  Josie, the German shepherd Andy planned to use in New Bedford, was about two years old when he first got her. Originally, she was undergoing training as a guide dog for the blind and visually impaired at the Fidelco Guide Dog Foundation in Bloomfield, Connecticut. She flunked out. “They couldn’t break her of stealing food off the plate,” Andy later joked.12 One thing Josie was good at was finding bodies. She was one of the best—and she loved to play.

  Less than a week after returning home from that first meeting with investigators in New Bedford, Andy was back in Massachusetts with a search plan in a folder and Josie in his SUV.

  The first area he picked was Interstate 195 in Dartmouth, where two bodies had already been found. He searched for five hours that first day. All he found was litter and discarded bald tires. A lot of tires.

  On the second day, November 29, 1988, he slowly walked along the westbound section of the highway in Dartmouth just before the spot where one of the bodies had been found. He was looking at areas where a car or truck could easily pull over, where there were no guardrails and where there was heavy brush or a gully.

  Secluded entrance ramps—like the one where one body on Interstate 195 was found earlier that month—were also on his list.

  It was a warm day for November, and an unmarked cruiser, driven by Trooper Kevin Butler, stayed a safe distance behind Andy and his dog, Josie, making sure no one bothered the searchers. After passing the area where the second woman had been found in July, Andy moved on to the exit and entrance ramps.

  The pair searched the brush by the exit ramp then moved onto the adjoining westbound entrance ramp and the small clump of woods between the two.

  Twenty-five feet in, Josie bolted to a drainage ditch and came back, nudging at the ball in Andy’s pocket.

  “She found something,” Andy called out.13

  State Trooper Kevin Butler got out of his unmarked cruiser a few yards away and walked up the entrance ramp.14

  “Over there,” Andy told him, pointing.

  “Where?” the trooper asked.

  They walked through the brush.

  “Right there,” Andy said, pointing to the drainage ditch.

  Kevin Butler looked down. He saw muck.

  “I don’t see anything,” the trooper said. “Andy, there’s nobody there.”

  “The dog says the body is right there.” Andy pointed straight down. “Right there.”

  Kevin straddled the ditch, bent down and reached into the mud with his gloved hands.

  He felt something and pulled it out.

  It was what appeared to be a skeleton.

  He put it gently back down.

  Body number four was found.

  This was now a crime scene.

  NATURE WILL CHANGE and reclaim a dead body over time. Exposed to dry heat, the skin can take on a leatherlike or mummified appearance. Exposed to the freezing cold, the body can look ashen. Exposed to the water, the body will bloat, and algae may be found growing on it. Sometimes the environment can shield the body from the elements. Sometimes it can give conflicting clues. Sometimes what you see on first sight can be misleading; sometimes it is spot on. There are no quick or simple answers, only questions to be answered through detailed examination in laboratories. That first sight, though, can give a direction. These are all the things police in the field and forensic experts in the lab know from years of work. In 1988, many investigators still relied heavily on what they could see in the field rather than what could be seen under the microscope or in the laboratory. The use of forensic DNA science in law enforcement and in the courts was still evolving. To get a good DNA sample from a suspect, police merely obtained a blood sample. To get good DNA from a scene, however, the sample size often needed to be fairly large and sometimes was destroyed in the testing process. Some cautious judges demanded prosecutors prove DNA testing was not “voodoo science.” Defense attorneys challenged the science as confusing, misleading, and unreliable. Juries were often confused by the scientific verbiage and statistics used to explain the testing and results.

  State police in Massachusetts were also working to catch up with the latest technology in 1988. There was no official “crime scene services” division yet—that wouldn’t come until 1993—and gathering evidence was a fragmented process. At that time, four different units could be at a scene: the primary investigator assigned to the DA’s office, someone to collect ballistics evidence, chemists to collect blood or fluid evidence, and yet another person to photograph the scene and check for fingerprints. The closest thing to a crime-scene unit was the Bureau of Photography and Fingerprints and its five units headquartered in the state police barracks in Middleboro, Yarmouth, Boston, Northampton, and Leominster. While those in the photography and fingerprint unit primarily photographed the scenes as well as checking for fingerprints, some of the troopers in 1988 also looked for such things as tire tracks and the type of evidence the civilian chemists could test in the state lab. However, for the most part, taking charge of evidence often rested primarily with the trooper in the DA’s office investigating the case. This was a system that worked for years in Massachusetts and in other states across the country. But things were changing quickly in the forensic field, and investigators discovered the
y needed to keep up.

  In the state police investigative unit in the DA’s office, one trooper, Kenneth Martin, was fascinated by the latest advances in the forensic field.15 He first considered a career in the sciences but instead joined the Massachusetts State Police in 1980 and started in the Bristol County DA’s Office unit three years later. He wasn’t a scientist back then—the biology degree at Bridgewater State University would come later—but he understood science and knew the importance of finding the type of evidence someone might not realize was viable, the evidence those in the lab could analyze, the evidence that could be crucial in a case. He eventually approached the district attorney and Bob St. Jean about doing more forensic crime-scene work in the unit to supplement what the other state police evidence units were doing. It was considered a cutting-edge and growing field back then. They said yes.

  He went to seminars, workshops, joined professional forensic organizations, and talked with anyone he could about advances in the field. He met people who could analyze things he didn’t realize could be helpful in an investigation. Most of all, he found people in the sciences who were willing to help if needed. By 1988, he was looking at crime scenes with a different eye.

  In the field, he looked for bugs, dirt, bird nests, spider webs, and eggs, the tiniest of blood droplets. He dug up blocks of outdoor crime scenes and sent those sections, intact, to laboratories. He looked for the footprints, the tire prints, and fingerprints as everyone did. But he was also aware that there were clues that forensic science could find that might point to a suspect or a cause of death or information about a crime.

  The woman’s body found by Josie in the muck-like dip off the entrance ramp to Interstate 195 could yield a wealth of information. So could the area around her. Part of the woman’s body was sunk into water and dampness; a small part, a hand, was not. Kenny Martin bagged and gathered the bugs and dirt and sticks and rocks and anything he could find at the scene, hoping it would help in the investigation. But he, like the other investigators, knew it was sheer luck that this woman’s body landed on the ground in just the right way, with her hand clear of the wet. This might give them a break in the case. Unlike the earlier cases, where the bodies were essentially skeletal, the hand and fingers were protected somehow from the elements by a twist of fate. They could get a fingerprint. They might get a name.

  Today, thanks to advances in technology, just about every law-enforcement agency in the country can electronically scan a fingerprint from a human finger and enter it into the computerized Automated Fingerprint Identification System (AFIS), where millions of prints are searched and compared. It can take minutes to do a search. In 1988, it wasn’t that easy. The AFIS system was introduced just two years earlier in Massachusetts, but only a handful of local police departments in the state were entering new fingerprint cards into it. Even entering the prints into the computer system was time-consuming, compared to the process today. Latent fingerprint images from crime scenes were first enlarged five times, then a technician had to trace the image by hand to identify points in the print. Then that tracing was reduced and scanned into the AFIS computer where it was converted into data for classification and comparison with other prints. The process could take one to two hours. It was a little bit easier to enter the fingerprint cards of people who were arrested. Those prints were first scanned into the system, a technician would then review the prints for quality control and make annotations on them before finally entering the information. On a good day, the process would take roughly ten minutes.16

  Most fingerprint comparisons, even if they were entered into AFIS, were ultimately done by hand at the time. That meant spending hours poring through fingerprint cards to visually compare prints to find a match. “You had to do every fingerprint by hand and a lot of the fingerprint cards were not good. Most of them weren’t good,” recalled Richard Lauria, a retired Massachusetts state trooper who worked in the fingerprint and photo unit in the 1980s.17

  A large percentage of fingerprint cards done by officers in the field in the 1980s were smudged, unreadable, and unusable. Sometimes police officers rolled the fingers wrong on the cards or didn’t get a clear imprint or did it so fast the print looked like a blur. “At least 50 percent were not legible [enough] to do anything with,” the retired trooper noted.

  In most cases, police had a suspect in a case and compared the prints on the cards with those lifted at a crime scene. In the case of the dead, authorities had a possible name of the victim and used the fingerprint cards to confirm it. Early identifications through fingerprints were time consuming. For example, the 1984 slaying of a seventy-four-year-old Holyoke woman was solved thirteen years later through fingerprints when state police lt. Brian O’Hara, who felt the AFIS system at the time was missing some prints, asked a young state police technician named Christopher Dolan to manually examine 40,000 prints from the Holyoke police department after authorities theorized the killer likely was local. It took more than four hundred hours over three months to do. The lieutenant later verified the final comparison and discovered the suspect was in the AFIS system, but the computer originally failed to make the match.

  The investigators in Bristol County knew what they were up against as they waited to see if the prints from this latest victim could be matched. The chances they would quickly get a hit on this woman’s fingerprints depended on so many factors: If she had ever been arrested. If she gave her correct name when she was booked. If her fingerprints were taken when she was arrested, since not all departments printed everyone who was arrested. If those fingerprints were clear. If that police department entered the prints into the statewide system. There were a lot of ifs.

  All they could do now was wait.

  IT WAS GETTING LATE in the afternoon, and Andy Rebmann figured he could wrap up the search along this last stretch of Route 140 in Freetown about three hundred yards from the New Bedford line by sunset. Two days earlier, his dog had found the remains of a woman off a highway ramp on Interstate 195 in Dartmouth west of New Bedford, the road to the south T-boning Route 140, and now they were continuing the hunt for more bodies. It seemed unlikely he would find another body in this last area to be checked on this fourth, and final, search day. And it seemed even more unlikely something would be found in this final hour before he drove home to Connecticut. He and Josie spent the day of December 1, 1988, looking, along with Massachusetts trooper Kathleen Barrett and her dog, Syros.

  Three Massachusetts troopers—William Delaney, who worked the first two bodies, Leonardo Solana, and Michael Harding—were keeping an eye on the dog handlers throughout the day. If the dogs found a body, the three would secure the scene and call for more help. In the meantime, if Andy needed anything, anything at all, they kept telling him, just ask.

  Just one more hour—or less—and Andy figured he would be done. He watched as Josie trotted along the side of the road before heading toward the woods. Then she came back. Her tail was up. She nosed the ball in his pocket. He knew she found something. He followed Josie thirty feet off the road, into the thick brush. There, two miles from where the first woman had been found in July, were the remains of what appeared to be a partially clad woman. As he stared into the tall grass and brambles, Andy knew he wasn’t heading home early.

  This was now body number five.

  By the time Jose and Maryann pulled up to the scene, road troopers were already there to keep what would be a growing number of reporters, photographers, and remote-broadcast trucks back. As word spread about the discovery, people began to gather on a nearby overpass and motorists slowed to catch glimpses of the scene. Troopers lit road flares for safety as night fell.

  Bristol County district attorney Ronald A. Pina was fifteen minutes from New Bedford on Route 140, returning home from a meeting in Boston when he saw the cluster of police cruisers. This was now a familiar sight to him. Another body.

  Pina pulled over and was ushered to the body. There, he would see the remains of a woman on her stomach, jeans a
t mid-thigh and the upper body unclothed.18 He emerged a few minutes later, grim-faced. “It’s just crazy,” he told reporters. “It’s the same thing, a body off the side of the road.”19

  It was becoming a pattern, he said. It was one he hoped would end soon.

  They needed to find this killer but first they needed to know who this latest, fifth victim was.

  Investigators had some records of missing women and were getting more. They hoped one might match this woman. If not, the district attorney told reporters the FBI or another agency might have an expert who could reconstruct the faces of the dead to help them in the identification process. An autopsy performed the next day at St. Luke’s Hospital in New Bedford revealed the woman died four to six months earlier. She weighed between 130 and 140 pounds and stood between five feet two and five feet four.

  Outside experts at the FBI and the Smithsonian Institution in Washington needed to be called in again, the investigators knew.

  Someone was missing these women—but yet again the haunting question was: who?

  JOSE GONSALVES slit open the nine-by-nine manila envelope in that day’s mail from the Worcester police and stared at the mug shot on the arrest card inside. Then he slid it over to Maryann. Two more bodies had been found within two days, and the troopers were hopeful they could positively identify one of them today. The troopers already knew a partial fingerprint lifted from the first body found by the dog on November 29 along Interstate 195 in Dartmouth was a match to a woman arrested in Worcester: a woman named Joanna Marie Rose. Just getting that match had been a long shot since so few police departments in the state were in the AFIS system, and if the woman had been local, they would be out of luck. New Bedford wasn’t in the system. Neither were the two cities to the north and west, Taunton and Fall River. Troopers Richard Lauria and Kathleen Stefani, both in the photo-and-fingerprint unit, had been able to lift a fingerprint from the woman’s right thumb, enter it into AFIS, and then confirm the match to the Worcester print. This fingerprint identification was just the first step, though. They had been waiting for this mug shot to help confirm the identification before taking the next step to obtain dental X-rays.