The forensic unit of the University of California, Davis, Veterinary Genetics Laboratory is housed in a small cluster of run-down, double-wide trailers surrounded by silos and fields. The day I visited, director Elizabeth Wictum and her team had just finished analyzing dog hairs found on the clothes of two Atlanta murder victims. They compared the specimens to hairs collected decades ago from the dog of the man convicted of the murders in 1982, reasoning if the hairs didn't match, he may be innocent.
The dog-eared lab has a noir charm that is reflected in Wictum's world-weary nonchalance. The crimes that come across her cluttered desk are often brutal enough to harden the kindest heart. "Two times last year I had to testify about dogs burned in ovens," she says, shaking her head slowly. "We have cases where we have to test knives, drill bits, chainsaws, and swords and compare the DNA found on them with that of an animal corpse."
Sometimes people contact the lab for less high-profile reasons, for instance to figure out whether the neighbor's dog killed their cat, by testing DNA from saliva, hair, or blood retrieved from the remains of the victim. It can cost as little as $500 to run such an analysis. A big case, like the one involving the Atlanta murders, can involve a lot of evidence and cost as much as $5,000, still a bargain for a victim or defendant.
In one case, a burglar put two Pekingese dogs in an oven and turned it on to keep them quiet while he ransacked the home for valuables. When the police traced the crime to a suspect and searched his apartment, they found running shorts with blood on them. They sent the shorts and tissue samples from the remains of one of the dogs to Wictum, who tried to isolate DNA from the samples. "We weren't able to get any nuclear DNA, possibly because it got destroyed by the heat," says Wictum.
Wictum's lab has developed a series of species-specific primers that help analyze mitochondrial DNA, more numerous and durable than nuclear DNA. MtDNA does not provide as specific a signature as nuclear DNA, but in this case, the dog's tissue and the drop of blood on the gym shorts revealed a haplotype found in only about 1 in 75 dogs. That evidence alone may not have been sufficient to seal the case, says Wictum, but together with other evidence it convinced a jury to convict.
Wictum didn't suspect her career would take this turn when she studied zoology at UC, Davis, in the 1970s, nor when she went to work at the VGL, doing protein electrophoresis for parentage testing on racehorses and other livestock. In the early 1990s she helped develop the DNA techniques for establishing parentage for horses and other livestock, which have become the industry standard and the basis for the lab's forensic work. In 1999, responding to an increase in requests for forensic animal DNA sample analysis from police departments, lawyers, and district attorneys around the country, Wictum was asked to start the forensic testing service at the lab.
The great bulk of the lab's work is still testing horsehairs to establish parentage for breeders so they can prove the lineage of their horses (the lab can conduct 1,000 tests a day), but it spends more time every year on forensics and the research that supports it.
As for the Atlanta murders, the lab's analysis of the dogs' mtDNA concluded that the convict was a possible contributor. Though all the tested hairs shared the same haplotype, it was too common to say much about the likelihood of the dogs' specific identities.