Who Killed Mellory Manning?

David Wolman
Published in
20 min readApr 13, 2015


Dallas Mildenhall used his microscope to crack cases all over the world. Then a murder took place in his own backyard.

By David Wolman
Illustrations by Cam Floyd


It’s been a tough year, but Mellory Manning’s turning things around. These days she’s steering clear of the people, places, and habits that have given her so much trouble.

Mellory is twenty-seven, and she’s kind of striking: a wide forehead, high cheekbones, thick, curly auburn hair. The people who know her around Christchurch know she’s loyal to those close to her — a sympathetic listener, a friend. But life’s not simple, and Mellory’s got difficulties of her own, not least a longtime drug problem that she pays for with sex work.

Then, in August — the middle of wintertime in New Zealand — Mellory’s older sister Jasmine kills herself. The loss is a jolt. Her life’s on a downward slide, she realizes, same as Jasmine’s. It’s enough to catalyze a change: She quits the street, goes clean, starts thinking about the future. Maybe she’ll go back to school. She talks to her boyfriend about starting a family. But when Christmas arrives, Mellory’s down again. Short on money, she sees no option but to hook again. One last night.

On the evening of December 18, 2008, Mellory stands beneath the old oak at the intersection of Peterborough and Manchester, a crossroads just north of downtown Christchurch. It’s the same corner she’s been working for years. It’s summer. She’s in a pink miniskirt, blue bikini top with polka dots, a thin gray sweater, and a hooded cardigan over the top.

Mellory meets at least two customers that night. The first man drives her to the parking lot of a vacant hotel just a few miles away. The second encounter happens at around 10:00 p.m., in another nearby parking lot. It doesn’t take long: By around 10:20 p.m., she’s back at her corner.

Eight minutes later, she gets a text message from the client who just dropped her off. He writes:

See you again sometime for mutual fun.

She replies.

We will one day.

The next morning, a kayaker paddling through the shallow waters of the Avon River spots something entangled in the branches and grass along the riverbank. It is Mellory’s body. The autopsy reveals that she has been strangled, beaten with a metal pole, and stabbed three times before being dumped in the river.

The chief investigator on Mellory’s case was detective Greg Williams, a tall, brawny fifty-two-year-old who wears wraparound sunglasses. The killing put Williams, like the rest of Christchurch, on edge. Two other prostitutes had been murdered in the previous couple of years. Without a swift indictment, the community would inevitably start whispering about a serial killer, even though the police never thought the murders were connected.

Williams also knew the numbers were against him: Only a tiny number of investigations of murdered sex workers result in successful convictions. There are plenty of reasons these cases are difficult to solve, and some of them are understandable: the work itself, which happens in the shadows; the intimate interactions with many potential suspects; the complicating factors of alcohol and drugs.

But Williams chafes at the suggestion underlying those numbers — that society values some lives less than others. He set out to prove that such callousness, deliberate or otherwise, has no place in his department, or for that matter in his city. To find whoever murdered Mellory, Williams spearheaded what would become one of the biggest and most expensive forensic investigations in New Zealand’s history.

Cell-phone data and witness information helped paint a detailed picture of Mellory’s movements that night, providing police with numerous leads to pursue. There were the two clients, of course, and a Ford Falcon in the area that the cops wanted to know more about.

After meeting her second client, Mellory was picked up in another car at 10:40 p.m., and she sent that final text message at 10:43 p.m. Water damage to her watch caused it to stop just before 11:00 p.m., which corresponded with a witness report of a loud splash in the river at roughly the same time.

The medical examiner found small scratches on Mellory’s body that indicated contact with prickly vegetation. Analysis of plant material on her clothing suggested that she had come into contact with an area dense with weedy grasses. And the fact that the exit wounds on Mellory’s back were small and similar meant that the knife used to kill her likely hit against a hard surface. He also noted that Mellory’s left arm was probably positioned up and in front of her head when she was murdered, a reflexive act of self-defense.

Williams dispatched officers to examine dozens of sites around the city where Mellory had been or might have been in the weeks prior to her death. He also made sure he wasn’t missing anything in the river itself. Instead of the typical procedure of using oranges to determine how far the body might have floated, Williams had investigators build two adult-size plastic dummies so that he could more accurately model how Mellory’s body had traveled.

In the months that followed, hundreds of people were interviewed by the police, and suspects’ names were gathered, chief among them the members of a local gang, the Aotearoa Mongrel Mob. The group was financed by drug money and prostitution. Sex workers in Mob territory have to pay a protection fee, and one hypothesis was that Mellory had defied Mob members and was killed as punishment.

In the neighborhood where Mellory’s body was found in the river, the Mob rented a hybrid garage/warehouse, as well as the overgrown vacant lot next to it. Located on Galbraith Avenue, the 1.5-acre property is just a short walk — perhaps four hundred yards — from the edge of the river.

The Mob connection was a reasonable guess, nothing more. The police redoubled efforts to pinpoint the location of the murder in the hope that it would shake loose new information about a perpetrator. But the case was stuck. Williams’s best shot now was a breakthrough by the pollen detective.


An Expert In The Invisible

Dallas Mildenhall unlocks a filing cabinet and crouches to open a thin drawer revealing rows of glass slides. Each one is labeled with impossibly small lettering. “Found it,” says the seventy-year-old scientist, gingerly removing slide L25854 from the tray.

We’re in the reference collection library at GNS Science, a geosciences consulting firm north of Wellington, New Zealand. Mildenhall has large, almost bulging blue eyes and wispy white hair and wears wool sweaters and sneakers. He has worked here as a palynologist for more than forty years.

If you aren’t professionally invested in the sex lives of plants, you probably don’t know what a palynologist is. In fact, you probably don’t think about their specialist subject — pollen and spores — at all. Pollen might earn your attention during a sneezing fit, or when you hear it mentioned in weather reports. That’s about it, though.

But palynologists like Mildenhall can’t wipe pollen from their minds. It’s too ubiquitous. Whereas you and I look through the air to the streets, parks, farms, beaches, and buildings around us, Mildenhall sees into that air. Even a seemingly tranquil setting — the manicured gardens outside GNS, for instance — is actually a blizzard of the microscopic, fueled by bursting flowers, wind vectors, zooming insects, and clouds of pollen grains swirling, rising, and drifting.

Amid this tempest, pollen grains are constantly falling out of the storm and settling on the ground, your windshield, and pretty much every square inch of the tactile world. If you know how to read them, the grains of this pollen rain can tell stories. And, sometimes, those stories can shed light on criminal investigations.

Pollen grains have characteristics that make them especially useful to investigators. They are tiny, of course, and often quite varied by type. That means experts are comparing specific, different structures, not just analyzing one set of blobs against another, similar set of blobs. Pollen grains are also outrageously tough: A heavy-duty laundry cycle with your favorite stain remover might successfully wash away a spilled pinot, but don’t expect it to rid your clothes of pollen. Sporopollenin, the substance that constitutes a pollen grain’s outer wall, is one of the most damage-resistant organic compounds on earth. “The acids we use to take them out of rocks would kill us,” Mildenhall says.

Mildenhall is one of the world’s best pollen detectives. He has worked on more than two hundred investigations and consulted on dozens more, helping to crack cases involving art forgery, narcotics trafficking, bank robberies, counterfeit pharmaceuticals, arson, and more.

The first thing to know about forensic palynology is that it’s almost never about single pollen grains. Why any individual grain is in a particular location is nearly impossible to know. But concentrations can be telling. Just as no two snowflakes are ever alike, at least in practical terms, no two pollen samples — a tiny pinch of material that might contain hundreds of individual grains — will ever be the same.

If you know enough about what grows where, what the surrounding vegetation is like, seasonal flowering times, dispersal patterns, local winds, and other factors that influence pollen rain, you can draw links among what was, until then, disparate information. Forensic palynology has been used to connect suspected individuals or items to a crime scene, to debunk or back up alibis, to locate illicit manufacturing operations, and to provide other key details. Mostly it is used to provide geographic information. Think of pollen as a form of fingerprinting that doesn’t tell you who or what — “I don’t prove that people did things,” says Mildenhall — but can tell you where.

For example, he explains, just because a pollen analysis of dirt taken from inside a suspect’s shed matches the profile of pollen taken from the victim’s hair, it doesn’t mean the suspect did anything criminal to the victim. But those stowaways can go a long way toward putting the victim in the shed at the date and time when the crime is suspected of occurring. And that can be enough to unlock a case. “If I were advising criminals,” Mildenhall says, “I’d tell them to confess to everything except the final act, because placing them at the scene is only a matter of time.”

L25854, that slide that Mildenhall has scooped out of the archive, is one of the hundreds made from samples that police sent him during the Mellory Manning investigation. He carries it back into his office and places it under a microscope. After decades spent carefully holding and turning the focus wheels of microscopes in the lab, he has developed tendinitis in his hands. Sometimes he does slow-motion exercises with his thumbs to try and alleviate the discomfort, but the exercises are of questionable benefit. He shrugs and leans in to the eyepiece.

Slowly scanning, focusing, and scanning again, he finally zeroes in on what he wants to show me. It’s a grain of pollen from Bromus diandrus, a grass commonly known as ripgut brome.


Napier. Rakaia Gorge. Avon River.

The basic forensic approach used at crime scenes today centers on Locard’s exchange principle: When any two things come into contact, there’s always some transfer of material, whether it’s tire tread on a pavement, bodily fluid on a bedsheet, or a jacket placed in a tree. Perpetrators and victims always transport objects with them to a crime scene, take things with them when they leave, and likewise leave things behind.

Some types of trace evidence — bodily fluids, clothing fibers — are used regularly in criminal investigations. But pollen, despite the fact that it exists almost everywhere, has been adopted more slowly.

In the late 1950s, it was used successfully in a murder case in Austria, where the police had a suspect and a motive but no way to link the man to the crime. The break came when investigators took a pair of muddy boots from the suspect’s home and sent them to a renowned geologist, who found pollen from willow, alder, and spruce, as well as rare hickory. Only a small sliver of land in Austria had soils characterized by this particular pollen brew, and when the suspect was told that police had evidence of where he had walked — and mentioned this specific area — the man sang his confession.

In the decades that followed, pollen analysis was used in cases here and there, but for the most part forensic palynology simply wasn’t, and isn’t, a thing anyone has heard of. One limiting factor is that just a handful of people on the planet can bring to bear the necessary analytical acumen. “There are only about 350,000 different pollen grains. It takes some skill to differentiate them,” quips Vaughn Bryant, professor and director of the palynology lab at Texas A&M University.

As a kid growing up in the hilly rain forests of New Zealand’s North Island, Mildenhall often went exploring in the woods or ventured into abandoned gold mines or made his way deep into the mountains, listening to the calls of native birds. From an early age, he loved the taxonomic.

At university, he studied geology because it seemed like the easiest way to bring his studies outdoors. His knack for mental cataloging made a palynology specialty a natural fit, and soon after graduation he took a job as a staff scientist with GNS. At the time, he spent most days looking at fossilized pollen from ancient forests.

One day in 1973, local detectives asked one of Mildenhall’s colleagues to look at some soil samples. “I thought, Oh — I can use palynology for this!” He conducted an analysis, but nothing remarkable came of it.

His first big case was in 1983. A fourteen-year-old girl named Kirsa Jensen had taken her horse, Commodore, for a ride along the beach near the town of Napier, on the North Island. The horse was later found wandering by a nearby river, but Kirsa had vanished.

She had been riding near an historic gun emplacement, built during World War II in case the Japanese decided to invade New Zealand. Mildenhall used pollen analysis to determine that a rope found at the site had substantial concentrations of pumpkin, beech, and broad bean pollen grains. Mildenhall then showed that this pollen profile matched a second rope, found on the farm where the primary suspect was employed.

Until then, most of Mildenhall’s forensic work had taken place in the lab, with samples sent back and forth in the mail. For the Jensen investigation, however, he became deeply involved, visiting the site along the beach, collecting samples, mapping vegetation, and examining photographs. It was detective work, minus the badge. “I just got so emotionally invested in it,” he says.

The police and the Jensen family were hopeful that this barely known forensic technique would strengthen the case. But in the end, the court concluded that prosecutors still lacked enough evidence. The family was devastated, and the suspect later committed suicide. “They will never have closure,” Mildenhall says.

Fifteen years later, a fifteen-year-old girl named Kirsty Bentley took her Labrador for a walk not far from Christchurch and didn’t return home. Kirsty’s body was later found in Rakaia Gorge, some fifty miles north. “I can still see her feet poking out from under the shrubs that had been placed on top of her,” Mildenhall says. How did she get to the gorge? Who took her there? “Nothing on her shoes, clothing, or anything gave us anything beyond what we knew,” Mildenhall says.

Bentley’s murder was never solved, either. Mildenhall says those two cases still haunt him. Certain words bring emotions and memories bolting to the forefront of his mind. The triggers can be anything, but often they are place names or pivotal details. Napier, Rakaia Gorge, rope. “Anything like that brings back the frustration about the cases.”

Greg Williams, the detective heading up the Mellory Manning investigation, had worked with Mildenhall on the Bentley murder. Even though the earlier case remained unsolved, Williams was still bullish on forensics. Almost as soon as he started trying to find Mellory’s murderer, he had investigators relaying pollen samples to Mildenhall to see what he could find.

The first sample was extracted from Mellory’s nasal passages. As we breathe, our respiratory systems capture pollen grains, which get cycled in and out of our noses about every twenty minutes. These grains persist in the body long after death, and Mildenhall found small quantities of pollen from grass, birch, and tree fern in the sample. If a victim is facedown on the ground just before death, hundreds, if not thousands, of grains will typically be found in nasal samples. The relative dearth of samples from Mellory suggested that she had been on her back at the time of the fatal attack. This was potentially useful for understanding the murder itself, but it did little to help the police figure out where she had been killed.

Mildenhall received sample after sample to compare with the material from Mellory’s body and clothing. It was slow going, and after a period of more than a year, he had helped the police eliminate dozens of possible murder locations.

But as the months passed without an arrest, Mildenhall began to feel as though he was reading a depressingly familiar script. He’d helped the authorities win convictions in cases all over the globe, yet every time a young woman was murdered at home in New Zealand, the world’s most talented pollen sleuth couldn’t do a damn thing. He grew increasingly pessimistic. Was Mellory’s case destined to go the same way as that of Kirsa Jensen and Kirsty Bentley? Napier. Rakaia Gorge. Rope. Avon River.

One sample, at least, held some promise — even if it wasn’t conclusive.

“In my opinion, the evidence does not support the contention that Manning took her last breaths at the sites represented by the comparator soil and vegetation samples from Dallington Terrace, Avon River bank, Caledonian Hotel carpark, or 26 Gresford Street,” Mildenhall wrote in a report to the Christchurch police. However, “The evidence does not exclude the possibility that Manning took her last breaths at 25 Galbraith Avenue.”

It was the address of the Mongrel Mob’s pad.


Grain of Truth

Two weeks after Mellory was murdered, a young detective named Gabrielle Thompson had visited the Mongrel Mob warehouse on Galbraith to talk with any occupants. No one answered when she knocked on the door, so she turned to leave. That was when she noticed that some of the grasses in the vacant lot next door looked similar to some of the seed heads that had been plucked from Mellory’s coat. Thompson took a few samples with her, just in case, and sent them on to Mildenhall.

The grasses were ripgut brome. Mildenhall gave them a look but found nothing out of the ordinary: pollen from the grass itself, as well as from other weeds commonly found in vacant lots in that part of the world.

Williams also asked Mildenhall to look at samples taken from the jacket and compare them with material from around the Mongrel Mob crib. He didn’t think it was a total waste of time, but the pollen was so widespread that its presence (or absence) couldn’t signify much of anything.

Yet when he examined sample BDX004, from Mellory’s coat, Mildenhall noticed something astonishing. Although pollen grains are dazzlingly diverse, they have some things in common: hard outer shell, reproductive cells, and a single pore. This pore is the hole through which the gametes — basically sperm — are carried outward to do their business.

Under the microscope, ripgut brome pollen is roughly spherical, like textbook drawings of single cells, with a distinctive smaller shape enclosed within: a pore. This particular grain, however, had not one but two pores. At first Mildenhall thought it was a trick of the eye, a lump, a bump, a dark spot that just looked like another pore.

But soon after, he saw a second one just like it. Then a bunch more. Now there was no mistaking it: Something like 5 percent of the ripgut brome pollen in the sample from Mellory’s coat contained two-pored grains — an unusual, unexpected mutation. “It was just incredible!” Mildenhall says. “Something must have changed it at the genetic level!”

To a pollen geek, this is dramatic in the same way that, say, a five-legged dog is dramatic. Mildenhall contacted Vaughn Bryant in Texas for a second opinion. Bryant took a look at the images and immediately e-mailed Mildenhall to confirm what the New Zealander already knew: It was mutant pollen, all right. If the cops could find matching samples from a location in Christchurch, they could potentially unlock the case.

Mildenhall believed the aberrant pollen grains had been caused by an herbicide. It was just a theory, but he thought a weed killer sprayed on grass just when it was going to flower could have caused the genetic mutation. He mentioned the idea to Williams. Could Mellory’s coat have come into contact with an area that had recently been sprayed?

Less than two weeks later, police confirmed that the vacant lot next to the warehouse had been sprayed with an herbicide just a month before the murder.

So Williams sent Mildenhall back to rework the samples taken from Galbraith Avenue, to specifically look for more mutant pollen. Finding some would all but confirm that the killing had happened there.

Mildenhall was dubious. “What were the odds I’d see it at this possible crime scene? I thought completely zero.” Zero because the pollen on the coat — whether it was mutant or normal — had come from just one or two plants. Zero because any signal from the parent plant would be swamped by pollen from other grasses and bushes in the area. Zero because ripgut brome releases pollen for only a short period each year, so there was no chance of capturing any more. And zero because of the rarity of two-pored pollen to begin with.

Mildenhall proceeded nevertheless — better to look for a needle in an epic haystack than watch another case go cold. “I was skeptical I’d find anything, but hoping like hell that I would.” He spent three days scrutinizing sixteen slides full of pollen grains taken from Galbraith Avenue. On the afternoon of the third day, he spotted one. “This was the jackpot!”

He immediately called Williams. The find was consequential, no question. But it was only a start. Mildenhall needed to find more, which he did. Still more mutant pollen on the coat would eventually seal it. The grains were so deeply embedded in the fabric that “direct and forcible contact” was the only explanation for how they had gotten there. Mellory had been on her back, partly on a pad of concrete, partly on the grass, struggling for her life. The police had their location.

On Mildenhall’s suggestion, the police asked a botanist to go to the site and check for other potential sources of two-pored pollen. There were none. Mildenhall triple-checked by looking again at samples from other locations around the city, to demonstrate that the two-pored grains had come from material outside the warehouse. Although it’s possible aberrant pollen could have come from other places, the numbers were just too unusual. In all of Mildenhall’s experience, and likewise that of other top palynologists he consulted, mutant grains in such concentration had never been seen.

When I asked if he missed the aberrant pollen the first time around, Mildenhall said no. “It was one of those truisms of science,” he said. “It’s much easier to find something once you know what you’re looking for.” When profiling samples, it’s a matter of identifying and counting grains to determine concentrations of various types. Once he had discovered the aberrant grains and discussed their possible significance with Williams, his mind and eyes shifted frequencies. The new search was literally more granular.

“The pollen evidence was a keystone,” Mildenhall says. Williams and his team were able to use the location specifics in conjunction with other information — river-flow tests, cell-phone data, witness accounts — to question suspects. Under questioning, a young member of the Mongrel Mob named Mauha Fawcett ended up revealing details he could have known only if he’d been part of the killing.

On a bright, windy afternoon last June, Mildenhall and I took a taxi from Christchurch International Airport to the neighborhood of Dallington. The magnitude-6.3 earthquake that struck Christchurch in 2011 devastated this part of the city, leaving thousands of homes uninhabitable because of the shifty ground. The neighborhood’s empty silence calls to mind a plague or zombie movie. Mildenhall could barely recognize the place.

After stopping at the site along the Avon where Mellory’s body had been found, we drove to Galbraith Avenue. On the way, we talked about earlier cases — about Kirsa Jensen and Kirsty Bentley. Both of those investigations are technically still open, but Mildenhall says they’ll never be solved. “At least that’s not Mellory’s story,” he says.

Last spring, Mauha Fawcett went on trial. Mildenhall’s testimony lasted two hours. Fawcett had chosen to represent himself, and when the opportunity arrived for him to cross-examine Mildenhall, Fawcett simply said: “I’m confused about this one. I might just pass.” The investigation itself is not yet over — Williams is tight-lipped about specifics, but it’s no secret that the police suspect other people were involved in Mellory’s death. Fawcett, for his part, was convicted of murder and sentenced to life in prison.

Mildenhall stepped out of the taxi at the location where the Mongrel Mob warehouse once stood. “They killed her here and probably carried her down Galbraith and dumped her in the river,” Mildenhall told me, pointing toward the Avon.

Many of the surrounding houses, although unoccupied, are still there, forming a perimeter around the gang property. I could see how easily weeds and grasses, the ones that ended up breaking the case, would have encroached on the yards of these homes.

We stepped through a gap in a temporary chain-link fence. Hands jammed in the pockets of his fleece sweatshirt, Mildenhall scanned the ground as he navigated rubble, weeds, and piles of dirt. At one point, he crouched and felt a browned, tired-looking grass that was about as high as his shin. “It’s not feathery enough to be ripgut brome,” he said.

As we searched the area, the palynologist was somber. He had finally helped solve a murder case on his own doorstep, yet the feeling was anything but triumphant: His one-in-a-million forensic revelation doesn’t matter to Mellory and can’t ease her family’s pain.

The murder site itself has completely disappeared, buried beneath mounds of debris and earth. “A final insult to her memory,” Mildenhall said, kicking a chunk of concrete. “You couldn’t even put flowers on the place if you wanted to.”

This story was written by David Wolman. It was edited by Bobbie Johnson, fact-checked by Eric Wuestewald, and copy-edited by Will Palmer. Illustrations by Cam Floyd for Matter.

Log in to Medium and “recommend” this story.
Follow Matter on Twitter | Like us on Facebook | Subscribe to our newsletter



David Wolman

Author of ALOHA RODEO and other nonfiction. Contributing editor @ Outside. Exec. Editor @ Atellan Media.