Framed by a fly
DNA seemed the ultimate proof of guilt until one obsessive scientist uncovered a fly in the ointment. Liz Porter investigates.
The body of a woman is found in a luxury Gold Coast apartment. She has been raped and murdered. The odour of decomposition fills the apartment; homicide detectives estimate she has been dead for at least two weeks. The smell of bleach also lingers. The killer, it appears, intended to leave no trace, but was not careful enough. A crime scene examiner discovers a small cream-coloured stain on a windowsill. It tests positive for semen and the lab returns a lucky strike: the DNA profile matches that of the man from the flat next door.
The neighbour, bewildered and terrified, is adamant that he has never set foot in the place. But he is nevertheless arrested, charged with sexual assault and murder, and committed to stand trial.
Given the chance of a false DNA identification is less than one in 10 million, it seems like an open and shut case, but fortunately the neighbour’s lawyer knows a scientist who is ready to challenge these overwhelming odds.
Forensic biologist Annalisa Durdle will argue that the incriminating stain may not be what it seems. Rather than being semen, it may instead be a spot of fly poo. The reason it contained the neighbour’s DNA was the fly had fed on his semen.
Durdle’s groundbreaking research has shown that human DNA can be extracted from the faeces of flies, but crime scene examiners cannot always distinguish between tiny spots of creamish fly poo and splatters of semen.
As the trial date approaches, the defendant’s expensive lawyer immerses himself in the fine detail of Durdle’s research, hoping it will be a game-changer. He certainly needs one against the formidable scientific odds and a prosecutor well-versed in wielding them.
In his opening statement, the prosecutor skates over the fact that the incriminating evidence is thin on the ground. True, the accused has no criminal record, there is no history of a passionate relationship between him and the victim, and no video footage of him entering the flat. For a jury of 20 years ago, there would be no case. But a modern jury works on modern evidence. The tiny spot of semen – presumably missed when the accused was cleaning up the murder scene – is all the evidence this jury needs to incriminate the neighbour. He tells them a DNA expert will testify that the chance of a DNA test producing a false identification is infinitesimally small. The DNA in the semen spot by the window definitely belongs to the neighbour.
The first witness is the crime scene examiner who photographed and swabbed the crucial windowsill stain.
The defence lawyer now puts his diligent study of Durdle’s scientific papers to good use. He bombards the crime scene examiner with questions about human bodily fluid stains and fly artefacts at crime scenes. He drills down with detailed questions about the shapes and common locations of these artefacts.
Taken aback, the witness keeps repeating that these subjects are not his area of expertise. But he is clearly flustered; he blushes and stammers. Meanwhile, a furrow has appeared on the prosecutor’s brow.
Of course Durdle’s research has just been published and the crime scene examiner couldn’t be expected to be familiar with it. But it’s a major coup for the defence lawyer who has made the crime scene witness look like a man who cannot tell the difference between semen and fly poo.
The defence opens its case with Durdle as the star witness.
Dark-haired, 40-something and dressed in a smart grey suit, she approaches the witness stand, eyeing the courtroom as she goes. Today it is packed with reporters who have headlined the case “real-life CSI”. Durdle is unfazed; she is accustomed to entertaining audiences with vignettes of her bizarre research. But not on this occasion. This defence barrister keeps his expert witness on a very short rein.
Piece by piece, with his long questions and her short answers, the lawyer builds an alternative scenario which steadily dismantles the prosecution’s case.
Modern DNA analysis, the jury hears, can indeed achieve results that would have looked like science fiction 20 years ago. Back then a bloodstain the size of a 20-cent piece was needed to extract enough DNA for even a partial profile. Now the entire DNA profile can be extracted from samples smaller than a pinhead.
The DNA profile will certainly identify its owner. But there is no tried and true method by which crime scene examiners can determine how the sample of human DNA arrived at the crime scene.
The barrister shows Durdle a photo of the small, round cream-coloured stain that yielded the accused man’s DNA.
“Is it possible,” he asks, “to tell whether the human DNA in this stain was left at the flat by the suspect, or, rather, by a passing fly that had earlier fed on the suspect’s body fluid at an entirely different location?”
“It is not,” she replies.
The defence barrister asks her to elaborate. She explains that blood and semen stains and fly poo can look very similar; in particular the kind of splatter produced by gunshots or blunt force trauma can easily be mistaken for fly droppings. A crime scene examiner without specialist expertise might well confuse them. And flies also love to eat semen, especially dried semen.
“So is it possible,” the lawyer asks, “that the ‘semen stain’ that yielded the accused’s DNA was actually the excrement of a fly?”
“It is,” she agrees.
The lawyer pauses.
“So it’s also possible that the owner of the semen in that stain was never at the crime scene. Instead, the fly that deposited that stain dined off a soiled sheet in the man’s bedroom and then, attracted by the smell of a recently dead body, flew next door?”
“It is,” she agrees.
The lawyer notes the looks on the faces of the jury. They are gobsmacked. The bewildered, petrified neighbour could indeed be innocent. His DNA could have been planted in the apartment by a fly.
Three days later the jury acquit the hapless neighbour.
This murder case is a hypothetical, but Dr Annalisa Durdle is very real. And she believes that a miscarriage of justice like this could well take place.
“People see DNA as the Holy Grail now, especially juries, so it is a worry if a conviction is obtained without other evidence to back up the DNA,” she says.
According to Frank Vincent, a retired Supreme Court judge and chair of the Victorian chapter of the Australian Academy of Forensic Sciences, Durdle’s work highlights the risk of miscarriages of justice caused by modern technology. DNA has become “extraordinarily dangerous evidence”, he says. “We have been learning over the last few years that as the size of the DNA sample decreases, the risk of contamination increases, and the risk of injustice increases.”
Take the case of Melbourne student Farah Jama, who in 2008 was jailed for a rape that, it was later revealed, never happened. The teenager was alleged to have assaulted a middle-aged woman in a Doncaster nightclub. But the “victim”, who had passed out in a locked toilet cubicle, had not been raped by anyone. The only evidence linking Jama to the crime was the fact that his DNA profile had been extracted from the one intact sperm taken from the woman’s vagina. This in itself should have aroused suspicion: an average ejaculate contains over 40 million sperm.
Jama served almost 16 months before it was discovered that he had been falsely incriminated by contamination at the hospital suite where the alleged rape victim was examined. The previous night, a young woman had been examined in the same suite. It turns out she’d recently had oral sex with Jama. With his dried semen still in her hair, a fleck must have ended up on one of the slides used to store the older woman’s sample.
Forensic DNA analysis is a niche art but Durdle, a world expert on fly excreta, has carved out a niche within a niche. Like most forensic scientists, she is obsessive and punctilious about her work, which involves analysing mind-numbing amounts of data on what most people would find a repulsive subject. But not to the New Zealand-born scientist. She admits to a taste for the quirky. She also has a wicked sense of humour. Her black curls coil and spring as energetically as her Kiwi-accented conversation and it’s no surprise to hear that she has hijacked more than one dinner party merely by giving a straight answer to the question: “So what do you do for a living?”
Indeed, for the masked ball that concluded last year’s Australia and New Zealand Forensic Society Symposium she fashioned herself a jewelled Venetian mask made from dead flies. “I’m always delighted to be able to share the fly poo love,” she laughs.
At 42, Durdle is older than the average newly minted PhD because she took a winding route to her vocation, arriving via a commerce degree and a 15 year stint as a graphic designer before training as a forensic scientist at Auckland University. Her master’s thesis was a harbinger of her inclinations. Titled “How Clean Is Your Evidence?”, it examined how DNA samples could be contaminated.
To support her research, Durdle took a job cleaning the labs at the New Zealand government’s forensic laboratory. It was there that she was smitten by “fly poo love”. Every day she used chemical agents to clean the surfaces. Then, at night, ultra-violet lights bathed the lab, destroying any residual DNA. But flies and insects still broke through the barricades. She dutifully cleaned them away. But one morning, an insect belly-up on the floor got her wondering whether insect invaders could actually pose a problem. Could a fly, for example, actually transfer human DNA into a “sterile” lab?
It wasn’t idle speculation. In a 2005 Italian murder case, scientists found a squashed mosquito on the wall of a suspected murderer’s flat. They extracted DNA from it and discovered that the profile matched the victim’s DNA – the only evidence linking the suspect to the corpse. Since this species of mosquito had a very small foraging range, the victim must at some stage have been in the vicinity, possibly in the flat itself. There were other examples floating around too, including cases where human DNA had been obtained from pubic louse excreta or the stomach contents of fly larvae.
Durdle decided to investigate, incorporating the topic into her master’s degree. She bought larvae from a pet food company, hatched a small colony of flies, and fed them blood, which flies were known to like. The textbooks were less clear on flies’ predilection for semen, so she made up a meal of semen flavoured with blood stripped of its DNA.
The fly poo tested positive for human DNA. Durdle’s suspicions were confirmed.
A full investigation however, would require PhD-level commitment. With funds for forensics research in short supply in New Zealand, she headed across the Tasman to the Victoria Police Forensic Services Centre, where scientists Roland van Oorschot and Maxwell Jones had made world headlines with their extraordinary discovery that “trace DNA” from multiple people could be recovered from fingerprints on a pen or a set of car keys.
With van Oorschot as a supervisor, she enrolled at Melbourne’s La Trobe University and began work in a tiny lab. She established a colony of Lucilia cuprina, the Australian blowfly, using maggots from Sydney’s Westmead Hospital, where the maggots were being tested as wound-healing agents.
She divided her research into two main areas. The first was about the appearance of fly poo and vomit – so-called “fly artefacts” – and the second was the DNA they contained. It was also vital, she decided, to understand flies’ food preferences. Flies consume blood and semen, but would they consume them if alternatives were at hand? Information like this would help crime scene investigators prioritise their tests. At a crime scene where flies could feast on pet food it was probably not worth testing fly faeces. On the other hand, knowing that flies love semen would be crucial in a situation such as our Gold Coast murder case.
Durdle enlisted her less squeamish friends to provide semen, blood and saliva. These were essential ingredients for a series of eight-hour experiments in which she videotaped her flies’ behaviour when offered a smorgasbord of blood, semen, saliva, pet food, honey and tuna. In other tests, she filmed flies choosing between dry and wet blood, and between dry and wet semen. She also tried peanut butter, but the flies wouldn’t go near it.
Her fly buffet was served in a transparent take-away food container. Placing her flies in the fridge to temporarily paralyse them, she would take each one out by the legs and check its gender. Then, dipping tweezers into a mixture of chalk and water, she would mark the backs of five males and five females, using a dot for females and a line for males. Thus dressed for dinner, they were offered the buffet.
Each experiment was done twice, with one-day-old and three-day-old flies, because food requirements vary with age. Durdle then spent hundreds of hours analysing the findings of her fly focus groups. She counted the number of times each fly visited each food source, and how long it stayed there. She stopped and replayed each video several times to track occasions when three or four flies fed on the same food. All up, she went through 128 hours of such riveting footage. “It drove me mad,” she admits, “but don’t tell my flies that.”
The results were clear. The favourite bodily fluid was semen. If semen was around they gorged themselves on it, preferring dry to wet. Flies continued to feed on it despite the fact that something in it, possibly the high zinc content, eventually paralysed and killed them. “I did an experiment where they ate only semen,” says Durdle. “I’d go away and after a day and a half they’d all be dead. Once I was watching a fly trying to clean its front legs by rubbing them together. One leg kept going one way, and one the other – and it kept on falling over. Eventually it just sat down on its bum. Then it went back and ate some more.”
Her flies also fed avidly on pet food and honey. As to whether they would forgo semen for pet food or honey is a little complicated; it depends on the fly’s gender and age.
Close to the bottom of the list of their food preferences was blood (they preferred dry to wet) and their least favourite was saliva. Durdle’s findings differed to what she’d read in the textbooks. While she says flies were known to feed on blood at crime scenes, no mention was made about semen. “That was completely novel.”
Durdle also released groups of 30 and 40 flies into large containers, then filmed them as they fed, defecated and vomited. Running the video backwards she could zero in to record which end of the fly produced the spots, and noted that, contrary to the assumption recorded in many forensic texts, most fly spots were faeces, not vomit. Both turned out to be rich sources of human DNA.
Durdle’s experiments also required her to spend long hours in a derelict house in the Melbourne suburb of Macleod used by Victoria Police’s Forensic Services Centre for crime scene training. She tracked the journeys of flies at simulated crime scenes under various light and climatic conditions, and with the food in different locations.
According to the texts, fly artefacts should be found on sun-warmed areas of ceilings and walls, on windowsills and near light-fittings – all information used to help investigators assess whether a stain is fly poo or human blood. Durdle decided to put the conventional wisdom to the test.
Her experiment involved covering all the walls with butchers’ paper and closing every gap with masking tape. Then she set down plates of blood, sugar and water, and released 1,000 flies. “On a hot day, I would open the box and run,” she says. “On a cold day, I’d be looking into the box and asking ‘are you guys alive’?” Seventy-two hours later she and a colleague would divide the area into grids, photograph each one, and then count and record the location of thousands of fly stains.
As before, her findings challenged traditional assumptions about fly behaviour at crime scenes. Besides being found in warm, well-lit areas, fly artefacts were also located far away from food sources, in poorly lit areas, and lower down in rooms.
All along, the hardest part of the work was keeping her fly colonies alive. Her first disaster, 18 months into the project, was an attack by the La Trobe campus population of Argentine ants. “The ants annihilated everything,” she recalls. “They were in the breeding cages, they’d eaten the flies and the maggots. In my lab I can’t use insect spray so I was boiling jugs of water and pouring it over the floor – which caused leakage downstairs.”
Durdle had to reboot her colony with new larvae, a process that took six weeks. She was forced to repeat the process twice. Once she came back from holidays to find her lab full of escaped flies. On another occasion, a colleague phoned to report a mass maggot escape. “There was this procession of maggots going down the corridor and I was running along trying to squash them before anybody saw,” she says.
The sign on her lab door – “Do not enter: forensic science experiment in progress” – also worried some of her Psychology Department neighbours. Their anxiety intensified after a maintenance person adjusted the ventilation system, which pumped her lab’s air, redolent with rotting liver maggot food, through the whole building. First came a phone call asking her if her experiment involved a dead body. Then she was encouraged to move her fly colonies to the university’s “animal house”.
Durdle’s trials have ultimately paid off. Articles in forensic journals and papers delivered at international conferences have secured her reputation as a leading light in the field.
German forensic biologist and freelance expert witness Mark Benecke singles out Durdle’s research for the way it bridges different disciplines. “This is a very rare thing. The true improvements come from mixing neighbouring fields, especially with real life forensic biological stains.”
According to Melanie Archer, the forensic entomologist at the Victorian Institute of Forensic Medicine, says: “Annalisa’s work is the first of its kind. This will be crucial for cases where we must try to distinguish confidently between blood spots and fly-generated artefacts.”
The practical ramifications of Durdle’s work are immense. Ian Freckelton, co-author of Expert Evidence: Law Practice Procedure & Advocacy, says future defence lawyers will soon look at ways they can use Durdle’s findings to provide an innocent explanation for the presence of their client’s DNA at crime scenes.
“Too many lawyers do not appreciate the value of learning about such developments,” he says. “Just as Dr Durdle’s discoveries have resulted from endless hours of punctilious attention to detail in her research, so too are miscarriages of justice only avoided by determined exploration by counsel of the potential for novel explanations for evidence that at first appears to be devastatingly inculpating.”
Now working as a forensic officer with the Victoria Police Forensic Services, Durdle is hatching plans to continue her research project with a postdoctoral fellowship.
Of course, one of the most compelling matters raised by her work is that two weeks after a sample has been deposited, the tests performed by crime scene workers can’t determine whether a stain is blood, semen, or fly excreta.
At a crime scene, forensic officers carry out rapid, on-site “presumptive tests” to identify which material merits further analysis for costly lab-based tests like DNA profiling. For instance, to determine if a stain is blood, they will moisten it and rub a yellow-tipped “hemastix” strip against it. It will turn green if the blood-specific molecule haem is present. For a semen stain, there is a reagent that turns purple in the presence of the semen-specific enzyme acid phosphatase. Unfortunately haem and acid phosphatase also show up in the poo of flies that have fed on blood or semen.
Durdle’s next quest is to develop a presumptive test for a substance unique to flies, “a test that confirms the ‘flyness’ of a stain”. Her preliminary experiments suggest that a simple pH test for acidity has potential. Most crucially, she needs to establish how likely the Gold Coast scenario really is. What is the likelihood and under what conditions is a fly likely to bear false witness against a neighbour?
The work of the forensic scientist is not everyone’s cup of tea. How does Durdle feel about the prospect of spending another four years loitering in derelict houses, watching hundreds of hours of video of flies defecating, and extracting DNA from fly poo? “Bring it on!”