Red River Girl Read online

  The pathologist continued to examine Tina’s skull, looking for areas of bleeding that could indicate a significant blow to her head. There were none. Moving down her body, he paid careful attention to her neck. His experience told him that strangulation was a common cause of death in female homicide victims, and assailants typically used more force than necessary to kill. This tended to leave evidence in the form of bruising, spots of blood, or damage to the larynx. He carefully dissected each layer of skin and muscle, looking for something out of the ordinary, but again found nothing unusual. There were no fractures or dislocations in Tina’s vertebral column, and the tiny C-shaped bone in the front of the neck, the hyoid bone, which is often broken under pressure, remained intact. Rhee knew it was possible for a perpetrator to strangle a victim and not leave a mark, but that was rare. He concluded that he couldn’t rule out the possibility that Tina had suffered a minor assault to her neck, but he could not say with any certainty that it had happened.

  Moving on to the internal examination, Rhee saw that Tina’s internal organs were still mostly intact, with the exception of the soft tissue of the brain, which had already disintegrated from being submerged in water. The teenager’s heart was healthy, and her cardiovascular system appeared to have been functioning normally. He examined Tina’s lungs, considering that she may have been unconscious but not dead when she entered the river. As there was no specific autopsy finding to confirm drowning, he knew it would be a diagnosis of exclusion, made only if all other possibilities were ruled out. But the lack of evidence left him unsure. There was also the possibility that she had been suffocated, but again he could find nothing to prove it definitively.

  Rhee’s next task was to look for evidence of sexual assault. The fact that the teenager had been found fully dressed with no rips in her clothing suggested she had not been involved in a struggle before death. But knowing that Tina had been a young runaway at risk of exploitation, he considered it likely that her killing was sexually motivated. Examining her, he found nothing to suggest she had been raped or sexually assaulted, but the pathologist knew that river water may have washed vital evidence away. Wanting to be sure, Rhee took swabs to be analyzed in the RCMP forensic laboratory in Ottawa to see if DNA from a second person could be identified.

  With no anatomical injuries on Tina’s body, Rhee considered the possibility that she had died from an alcohol or drug overdose. Testing would normally mean collecting vials of blood for analysis, but the state of Tina’s body made that impossible. Instead, Rhee collected samples of fluid from the chest cavity and removed a section of liver, sending everything to the RCMP lab in Vancouver for analysis. The findings came back promptly. Tina’s blood alcohol level was moderately high, just over the legal limit for driving. But the lab urged caution in interpreting the result, explaining that it may have been falsely elevated because of decomposition. Even if the reading was accurate, the alcohol in Tina’s blood would not have been enough to kill her.

  The lab conducted a second test for THC, the active component in cannabis. The results showed the presence of cannabis but this would not have been lethal. Tina tested negative for crystal methamphetamine and cocaine, and there was no evidence that she had used opioid prescription drugs. Nor was there any trace of gabapentin, a drug originally developed to treat epilepsy, which had become popular on the streets for its ability to give a mild cannabis-like high.

  In his final report for the Winnipeg Police, Rhee concluded that the cause of Tina’s death was “undetermined.” There was nothing in his examinations or the toxicology results that gave a clear indication of how she had died. The pathologist agreed that the circumstances in which she had been found suggested that she had been killed at the hands of another person, but he could not with any certainty say she had been murdered. “There are possibilities other than homicide, such as accidental death from drowning, after which someone could have concealed her body,” he wrote, adding that his senior colleagues agreed with him.

  Reading Rhee’s report, O’Donovan felt a deep sense of frustration. The detective had hoped that at the very least, the pathologist would conclude that Tina’s death was suspicious. He asked Rhee to consider ruling that the schoolgirl had died from “homicidal means,” pointing out that she “didn’t put herself in the duvet cover, fill it with rocks, tie a knot at the top, and hop into the river.” But Rhee maintained his position, and O’Donovan was not able to bring any pressure to bear.

  The detective moved on to his next question. Could the pathologist give an opinion on the post-mortem interval, the time between Tina’s death and her discovery? A precise answer was impossible, but Rhee was able to give an estimation based on his understanding of decomposition. He knew that the environment surrounding a body—the temperature, the surface on which it rests, its exposure to oxygen, and whether it’s in water—affects the speed at which it decays. He was aware of Casper’s Law, an observation used by doctors since the 1800s, which holds that a body left in the open air will decompose twice as fast as one immersed in water and eight times faster than one buried in the ground. Following this, he concluded the decomposition rate would have been slower than normal because the body was submerged.

  The water temperature may also have been significant. O’Donovan asked the Winnipeg harbour master for a record of the temperature at the time Tina was found. He was told that it wasn’t recorded in Winnipeg but measured two hundred kilometres upstream, due south across the US border in Grand Forks. There, the water had been 24°C to 25°C, not cold enough to have significantly slowed the rate of decay. Putting those factors together, the pathologist estimated a post-mortem interval of between three and seven days, with the caveat that it could be as little as two days or as much as nine. Rhee also told O’Donovan that he didn’t think Tina’s body had been lying exposed in the shallows for long.

  At home after work, O’Donovan conducted his own research on rates of decay by reading papers written by scientists and FBI agents working on “body farms,” research facilities where the decomposition of human bodies was studied in a variety of environments. He also turned to Kevin Pawl, a fellow sergeant in the Winnipeg Police Service and an expert in underwater search and recovery. Pawl had been a member of the unit for almost twenty years and had worked on numerous cases, including once recovering the body of a woman who was weighted down with ropes and patio blocks in the Assiniboine River. He had been the supervisor in charge of the dive team searching for Faron Hall the day Tina was found. Pawl had directed his divers to look in the area where Hall was last seen, because he knew bodies could sink and resurface quickly in the same place. This was known as a “first float” or “gastric float,” and it happened when gases trapped in the stomach expanded and dragged the body back to the surface. It normally took between two and five days, but warm temperatures could accelerate the process to as little as one day, which was why Pawl had sent out a team for Hall as soon as he could. It was also possible for a body to remain submerged for longer until it had reached the “full body float” stage, which happened when the limbs and abdomen bloated up from decomposition, typically seven to ten days after being submerged.

  Looking at the autopsy photos, Pawl could see that Tina’s body looked far larger than her death weight of 35 kilograms (77 pounds), suggesting that she had been at the “full body float” stage when found. Pawl knew that the weight needed to hold a body underwater in this advanced state of decomposition would need to be equal to half its body weight. The 11.5 kilograms of rocks placed inside the duvet cover with Tina would not have been enough to keep her submerged beyond a week. It was possible that her body had been trapped for a few days in the heavy silt at the bottom of the Red, but at some point her body would have become too buoyant to remain weighted down and she would have broken free. Knowing this, Pawl told O’Donovan he was certain that Tina had been in the water for no less than seven days, and more likely around ten.

  O’Donovan was keen for Pawl to tell him where on the river
bank Tina had been put in the water. But the sergeant was unable to give an answer. Putting together all the information from the river search and autopsy, O’Donovan believed the most likely entry point was the wooden wharf of the Alexander Docks. His theory was that Tina had been thrown into the water there and that her body stayed weighted on the riverbed until the combined factors of bloating and the churning action of the River Rouge‘s engines had set her free. She would have floated with the current until becoming snagged in the reeds in shallow water less than a hundred metres away.

  * * *

  —

  From the start of the investigation, O’Donovan’s detectives had been collecting security camera footage from around the Alexander Docks area. Not all the cameras had been helpful, with many set to overwrite within a day or two. The most promising was on the Mere Hotel, pointing towards its parking lot next to the docks. O’Donovan had already assigned the task of going through its footage to an enthusiastic young constable. Now he was able to tell the junior officer the time frame within which to search for suspicious activity.

  Meanwhile, the Forensic Identification Unit, more commonly known as Ident, had begun its work. Constable Susan Roy-Hageman had been the original Ident officer on the scene when Tina’s body was recovered, and now she would coordinate the forensic examination. Roy-Hageman had been with the Winnipeg Police Service for twenty years and had spent the last eight of them building up an expertise in crime scene investigation. In real life, forensic work was never as quick and conclusive as people thought, “not like CSI,” the popular TV drama. If anything, she would describe her work as painstaking, slow, laborious, and often inconclusive. It was normal for teams to go to great lengths to gather and preserve evidence only to find it too damaged to be useful. Test results rarely came back within the hour, and DNA was not always identifiable. In fact, Roy-Hageman had worked on several homicides where her team found no trace of a perpetrator at all.

  In this case, without a crime scene, Roy-Hageman and her team knew they had little to go on. Even with what they did have—a body, some rocks, and a duvet cover—it was likely the river had washed away evidence or it had become overwhelmed by decomposition. Nevertheless, the officer ordered a meticulous examination of everything found. Her team’s first task was to identify exactly where the rocks had come from. In total, they counted four large pieces of stone, a piece of brick, and a number of smaller stone fragments. Roy-Hageman recognized the stone as Tyndall limestone, which was native to Manitoba and used as riprap along the length of the Red River to protect its banks against erosion. The stone was present just north of the Alexander Docks but was also common farther upstream and so not much help in placing where the duvet cover had been loaded.

  The cover itself was saturated with mud, river water, and organic debris from Tina’s body. Roy-Hageman realized she would not be able to process it unless it was washed, which was not ideal for preserving evidence. But the officer figured she could limit the damage by washing the outside only and making sure that any evidence caught inside remained. When it was clean, she could see that it had once been white or pale beige, decorated with a pattern of embroidered leaves in gold, brown, and rust that ran along its bottom edge, rising up towards its centre. Roy-Hageman positioned it face down on a large sheet of paper and used a scalpel to cut it open into three panels. Her team carefully checked each of these for hairs and fibres, marking and photographing their finds in situ before lifting them off with clear tape. In total, they found twenty-nine individual hairs, seven of them grouped within a small cluster. They also identified four kinds of fibre that had come from elsewhere, coloured red, black, and red-and-blue.

  Before sending the hairs to the RCMP forensic lab in Ottawa for DNA testing, Roy-Hageman examined them under a microscope and selected the ones most likely to have a root from which she hoped a nuclear DNA profile could be extracted. It was the same RCMP lab to which the swabs from Tina’s autopsy had been sent. During the process of examining these exhibits, lab technicians had found a hair in the vaginal swab, which they had added to their list of items to be tested. Keen to move his investigation forward, O’Donovan asked the lab’s senior management to expedite the analysis, explaining that the case was particularly sensitive because it involved a child. The lab obliged. But when the results returned, they were disappointing. There had been too little biological material on any of the hairs or from the swabs to extract a profile.

  Frustrated, O’Donovan made the decision to send the material to a private lab in Laval, Quebec, that specialized in testing for mitochondrial DNA. This type of testing had been used by crime scene analysts around the world since the 1990s as a complement to nuclear DNA testing, but it was still regarded as a somewhat imprecise science by the Canadian justice system. Mitochondrial DNA was more abundant than nuclear DNA and easier to find, making it useful for cases where samples were damaged or degraded or there was limited biological material. But the profile it gave was based on a shared maternal lineage rather than matched to one individual, and this meant that results were less conclusive and often challenged in court. It also came with a hefty price tag. O’Donovan would be spending close to $200,000, which meant he needed approval at the highest level. He explained to his superiors that, with so few leads, this testing was their best and perhaps only option. They agreed, and the preliminary results were promising. The lab was able to prove that the hair found in the vaginal swab had the same maternal lineage as Tina and so belonged to her. They were also able to match three of the hairs found in the duvet cover to the teenager. But from the remaining hairs they were able to identify at least eight different unknown profiles. There was no suspect to match these to yet, but O’Donovan considered this a significant development.

  Later, Roy-Hageman sent the lab pieces of the duvet cover that had fluoresced under an ultraviolet light, believing they might show seminal fluid, saliva, or urine. The results, when they arrived, showed that the biological material on the cover matched Tina and no other profiles could be found.

  With the forensic testing exhausted, the next step was to find out where the duvet cover had been sold. The manufacturer’s label identified it as being from Costco Canada, a stroke of luck for the investigation team as the company was a members-only store that kept a record of each item bought by its registered customers. The barcode showed they were dealing with a queen-sized duvet cover in a pattern called Chloe Green. O’Donovan believed he could use Costco’s records to track down every Chloe Green cover sold in the city. It would be a laborious task, but his detectives would be able to check whether the owners were still in possession of the bedding and whether any had the sort of background that might make them a suspect.

  The search was complicated by the fact that the item barcode used to identify the Chloe Green cover was shared by three other duvet designs, and it wasn’t possible to identify them separately. Company records showed that from February 2013 to the date of Tina’s discovery in the Red River, Costco’s Winnipeg stores sold 864 duvet covers of all four patterns. In addition, twenty-nine duvet covers with the same identification code had been marked as surplus and sent to a discount store on Main Street. Again, it was impossible to say how many of these had been the Chloe Green pattern.

  Feeling he had no choice but to try to track down all 864 duvet covers, O’Donovan began to plan the most ambitious canvassing operation of his career. He comforted himself with the thought that although the search would probably take months, it would still be possible to complete and the situation would have been worse if Tina had been found wrapped in a plain white sheet. Dozens of detectives and police shift workers were drafted in to help. Teams working in pairs were assigned names of duvet cover buyers and asked to visit them personally to make sure they still had the bedding in their possession. Owners were not told why they were being asked, and for the most part the teams themselves did not know that the effort was connected to Tina Fontaine. The media had reported that the schoolgirl was found in some
sort of bag, which over time became widely assumed to have been a plastic bag. It was a discrepancy that suited O’Donovan, who felt the detail could be useful when a suspect was finally found.

  Despite the odd request, duvet cover owners were more than happy to help. In cases where the bedding had been sent out of town—to friends or relatives living elsewhere or to summer cabins—purchasers went out of their way to make sure they could still account for it. One drove to his lake house specifically to take a photo of the cover with a copy of that day’s newspaper lying on top of it. Others emailed pictures from as far away as Greece and the Philippines. Slowly, the Homicide Unit was able to build up a picture of how many of the 864 duvet covers were Chloe Greens and where they were in the city. Out of all those canvassed, only one hundred customers had bought the Chloe Green pattern. Of these, eighty-nine were able to show the police that they still had the cover. Five said they had returned it to Costco. Four said it was at a cabin or elsewhere but couldn’t confirm that, and two said they had given their bedding away to the thrift store Value Village, where donations were left outside in large steel bins that were often picked through by the homeless.

  Detectives had less luck identifying the owners of the twenty-nine covers sold at the discount store on Main Street, which didn’t keep records. Nor could they guarantee that the cover used to wrap Tina had not been brought into Winnipeg from another province. But even with these unknowns, O’Donovan considered the exercise successful in eliminating potential suspects. His teams checked names against the police register for serious offences or previous convictions for sex crimes and found no one of potential interest. And the project established that the Chloe Green design was relatively rare, a fact that was likely to become important later in the investigation.