Pollen pioneer Dallas Mildenhall - using nature's sex spores to crack crime
Wednesday, 28 October 2015
By the time Dallas Mildenhall got there, Kirsa Jensen was long gone.
Kirsa's horse Commodore was back home with the schoolgirl's grieving family and police had cleared away the rope that tethered him to the gun emplacement, where Kirsa was last seen.
All attention was fixed on that 3.5m length of cord, thought to belong to whoever was responsible for the 14-year-old Napier girl's disappearance. At the time - 1983 - it was New Zealand's most intensively investigated piece of evidence, ever.
The first Mildenhall saw of that Feltex Duralene rope was when it turned up in his lab. The cops had run out of options. His Dad and the case detective had got talking - they belonged to the same club.
Up until then, Dr Mildenhall - palynologist - had mostly spent his time examining fossilised pollen to help date oil deposits. The idea of using pollen to fight crime was novel - to Mildenhall, to New Zealand and, largely, to the world.
Mildenhall chemically treated the rope to extract the invisible pollen caught within its strands. Broad bean, pumpkin and beetroot pointed to a vegetable garden.
Witness William John Russell was a market gardener, but the pollen evidence wasn't enough to send him to court. Russell killed himself in 1992 and the case remains open.
It was Mildenhall's first forensic palynology case and the one that really got under his skin. Forensic scientists shouldn't become emotionally involved in the crimes they're investigating, but he couldn't help it. He'd walked the beach Kirsa had ridden on Commodore. He'd visited the last place anyone saw her. He knew then that this was a field he had to learn more about.
More than 30 years and 300 cases later, Mildenhall was sitting in his Lower Hutt office, eyes glued to the microscope, looking for clues to another death. This time it was Christchurch prostitute Mellory Manning and again the police case had stalled. His hair was whiter and flightier and the wisdom of three decades had allowed him some emotional detachment.
But in all his thousands of hours poring over slides of plant sex spores he had never seen anything like this.
The gun emplacement where Napier schoolgirl Kirsa Jensen was last seen. Photo: HAMISH COLEMAN-ROSS/FAIRFAX NZ
Forensic palynology, using pollen science to solve crime, dates back to a 1950s Austrian murder mystery. A man disappeared while travelling down the Danube. Mud on the suspect's boots contained 20-million-year-old fossilised pollen grains which could only have come from a small area on the river. The suspect confessed and led police to the body, which was just where the pollen said it would be.
It was a promising start for a novel branch of forensic science. But 60 years later forensic palynology remains a little known specialty and Mildenhall is one of only a handful of global experts in the field. Although widely accepted in Britain, it remains rarely used in North America.
Pollen, Mildenhall argues, is perfectly designed to tell tales on criminals. It's everywhere and has a cunning knack of sneaking into folds and orifices and sticking to skin and clothing. Different plants produce pollen and spores with markedly different structures, so you can quickly see what kind of environment a sample has come from. Its incredibly durable, withstanding washing and millions of years of sedimentation. Better yet, it's invisible to the naked, crime-scene-cleaning-criminal eye.
It can tell you whether the place where a body was found is the same place the person took their last breaths; whether a suspect had been in the crime scene area; whether the dung on the car tyre came from the cowpat at the scene of the hit-and-run; whether grass pollen on a dead prostitute's cardigan matches that found at a Mongrel Mob pad.
What it can't tell you is who did it. Even 30 years on, Mildenhall can't say for sure whether police had the right guy in the Kirsa Jensen case.
'I'm pretty sure I had the right rope.'
Crime scene analysis often requires hundreds of pollen samples. Photo: ANTHONY PHELPS/FAIRFAX NZ
Mildenhall nestles into a kneeler, stretching his back, in his office at the Lower Hutt crown research institute GNS Science.
The small workroom is a monument to his career. The yellowing copies of the New Zealand Journal of Geology and Geophysics cramming the bookcase go back to 1959. The stained lab coat hanging behind the door dates from the days of DSIR and the scarlet crime scene overalls behind it have now been superseded by disposable boiler suits. There's an old crank pencil sharpener clamped to the desk but Mildenhall laughs that his pencil is too stubby to fit.
When Mildenhall first arrived here armed with a degree in botany and geology he had never seen a pollen grain under the microscope. A born-and-bred Wellingtonian and son of an import/export merchant, he was motivated to study science by an inspirational Onslow College teacher.
Today, forensic palynology makes up about 30 per cent of his workload and he's considered one of three world experts in the field. He's testified for the prosecution and defence in courts in Australia, Ireland and England. Some of his most valuable work, says fellow palynology expert and British professor Patricia Wiltshire, has been his collaboration with the World Health Organisation and Interpol to identify pollen types in fake malaria drugs to work out where they are being made.
'What he's done that is truly wonderful is the provenancing of illegal drugs. He has really contributed with that.'
At 71, Mildenhall is supposed to be semi-retired, but he's still in so much demand he's here three days a week. Both he and Wiltshire worry that they cannot find properly-trained scientists to take over the job and that those dabbling in the field are producing poor work. It's a job that requires both robust science and a robust personality, Wiltshire says.
'You have to be able to work with the police, which is not easy. You have to stand up in court - the cross-examination in Britain tears you apart. Dealing with the bodies at the crime scene.'
Mildenhall rarely gets to visit crime scenes these days and he is anxious to point out that most of his work is far from glamorous. He'll spend hours at the corner microscope manually counting different types of pollen grains to estimate the relative percentages of different plants. That might be enough to tell you whether a sample came from a wetland, a forest, a bog or even the sea.
In one case, Mildenhall was able to establish that a body plucked from the sea was not a case of accidental drowning. The pollen in the lungs came from a freshwater environment.
'This person had actually been drowned in a freshwater lake and then been carried out to sea and dumped, with the perpetrators trying to disguise the fact that a murder had taken place.'
But the Mellory Manning murder brought the genuine lightbulb moment scientists dream of.
Manning was strangled, beaten and stabbed to death before being dumped in Christchurch's Avon River, where she was found on 19 December 2008. Police had cut squares of cloth from different areas of Manning's cardigan to allow Mildenhall to map the pollen distribution, in the hope it would tell them something about how she was killed.
It's a common use of the science - if you've been sitting under a tree you'll have pollen in your hair, on your forehead, on your shoulders. If you've had your face ground into the earth under that same tree, the distribution will be quite different.
The samples confirmed that Manning had died on her back. But they also showed something else. Something startling.
Pollen grains, such as these examples from the dracophyllum and restio plants, can be identified by their unique structures. Photo: SUPPLIED.
What makes pollen such a good forensic tale-teller is different species not only look different, they behave differently, too. Pine pollen, Mildenhall says, is about as useless as it comes when trying to pinpoint a sample's origin. Wind whips through the tall trees, carrying the thousands of pinus pollen grains considerable distances, so they end up pretty much everywhere.
'All you can tell from pinus is that you're dealing with the right planet,' Mildenhall says.
Among the thousands of pollen grains that might appear on a single microscope slide, there could be up to 200 different types. But most will be of just one or two common varieties. The real clues lie in pollen from rare plants that grow only in limited areas, and in plants whose pollen and spores are designed to be distributed by animals, so the grains are few and they fall close to the source.
'You look at those and you know that whatever you're looking at, the environment that you're looking for has those plants growing there.'
For example, Mildenhall has helped solve at least two crime investigations by matching rare silver beech pollen found at a crime scene. In a hit-and-run, a cowpat on the road contained an unusually high percentage of pollen from the mountain-dwelling tree. Cow dung embedded in the tire of a suspect's car was a perfect match.
In another case, Mildenhall used silver beech pollen to question a suspect's alibi. William Jan Haanstra was accused of murdering Terri King, who was found in an area of the Tararua Ranges heavy with silver beech. Haanstra claimed his shorts and jacket had never left Wellington, where the tree does not grow. But when Mildenhall extracted the invisible pollen tucked inside the weave he found silver beech's characteristic grains. Nonetheless, Haanstra was acquitted of the murder.
It was one of those uncommon pollen types that proved critical in the Manning case.
One of the pollen species Mildenhall found in Manning's cardigan came from ripgut brome grass. Like most pollen types, it has a characteristic structure, which in this case includes a single pore. But about 10 per cent of the grass grains Mildenhall was seeing had two pores, the second being smaller and not symmetrically arranged as in most multi-pore pollen structures. It was a genetic mutant.
He knew at once that if he could find the same structures in any of the samples from possible crime scenes he would know where Manning was killed.
'That was essentially a eureka moment. I was in here, looking at it under that microscope. I thought…funny. It gave me something to look for, without the necessity of doing huge counts, because I knew this was something so unusual. I asked all of the palynologists in New Zealand 'Have you seen this before?' and they all said 'No'. Then I asked the two people I work with quite closely overseas, 'Have you seen this before?' One said 'No, never', the other said 'Yes, once and only one grain'.
'As soon as I saw that I thought 'Can we find it at the crime scene?'.'
The mutant double-pore ripgut brome grass pollen that helped crack the Mellory Manning murder case. Photo: SUPPLIED
With a new sense of urgency, Mildenhall fished out a slide he hadn't yet analysed. It was a sample randomly taken by a conscientious police officer from a Mongrel Mob gang pad near to where Manning's body was found.
'I got that sample and looked at it and, lo and behold, there they all were,' Mildenhall says.
When he re-examined other samples from that weedy yard, he found more mutants that he had missed the first time around.
The discovery was a breakthrough for the Manning case. Mob prospect Mauha Huatahi Fawcett was eventually found guilty of Manning's murder, although he is appealing that conviction. But it raises a red flag for a branch of science under increasing scrutiny for a phenomenon known as confirmation bias - seeing what you expect to see, rather than what is actually there.
Contrary to popular belief, forensic science is not black and white, but is open to interpretation. The best illustration of that was a study asking five fingerprint examiners to confirm that two sets of fingerprints the FBI had mistakenly found to be matching were not from the same person. The prints were in fact pairs the experts had themselves judged as matching five years earlier. Only one of the five came to the same conclusion as their earlier self.
Mildenhall says only finding the mutant grass spores once he knew what to look for was not confirmation bias. It was simply that, at the first look, he had been examining the slides at a low resolution.
'When you're counting them, you're counting under a low power. You would see a grass grain with one pore, but the other pore would be in a fold and you wouldn't pick it up.'
Palynologists note the co-ordinates of anything unusual, so another expert can then peer review their findings - something Mildenhall has done in cases all around the world.
'If you start finding things that are not there, you're in serious trouble.'
Mildenhall does acknowledge, however, that there is a danger of developing an 'I must get this person' mentality.
'What you're looking for is truth. You're trying to work out exactly what happened. I'm not interested in who did it, why it was done or who was involved, I just want to know what happened. If you look at it like that you don't get emotionally involved. In spite of the fact that if you appear in court and you give evidence and you walk out of court you are approached by friends of the victim, who say 'Thanks very much for putting him inside'. I have to quietly say 'No, that's not my job. My job is to work out what happened'.'