09:45 05 April 04
An investigation of the way insects colonise corpses left decomposing in the open has cast doubt on one of the key techniques used to estimate when a murder victim died.
Along with assessments of the body's state of decomposition, insect analysis is the most common means for estimating time of death. Many species of flies and beetles may live on a human body as it decomposes. By identifying their stage of development, and comparing them to those on a pig or human body deliberately left to rot in a similar environment, forensic entomologists can work out how long a corpse has been lying dead.
These estimates are often claimed to be accurate to within months or weeks - or even days if the body has been dead for less than a month.
In an experiment to test the accuracy of insect analysis, Melanie Archer of the Victorian Institute of Forensic Medicine in Melbourne, Australia, placed five piglet carcasses in scavenger-proof wire cages in bush land.
At least once a week for four months she identified the insects living on the carcasses and on the ground underneath them. She repeated the experiment each season for two years.
Archer found unexpected variations. For example, blowfly maggots left the carcasses after nine weeks in the first winter, but after only four weeks in the second winter. A beetle species, Creophilus erythrocephalus, arrived after four weeks in the first winter, but after only two in the second (Australian Journal of Zoology, vol 51, p 569).
The differences could be due to several factors, such as varying weather or changes in insect numbers. But until now, Archer says, none of the published studies on reference corpses has been repeated over successive years, so no one takes the variations into account.
Another potential source of error lies in the way temperatures are estimated, Archer found. Insect infestation of a corpse is strongly affected by temperature.
To estimate the temperatures at a site where a body has been found, entomologists usually go back to the site and monitor temperature changes there, and compare them with temperatures at nearby weather stations. They then go back to weather station recordings from the time the body was decomposing to estimate the temperature of its surroundings.
When Archer put this technique to the test at sites of hypothetical murders, the estimates were usually within 1 degree Celsius of actual temperatures.
However, if the weather had changed markedly between the monitoring period and the period the imaginary body was decomposing, the estimates were out by as much as 8 degrees Celsius. That could lead to errors of several days in the estimated time of death (Journal of Forensic Science, DOI: 10.1520/JFS2003258).
"The estimates are not as tight as some forensic scientists imply in court," Archer concludes. "We need to introduce some rigour."
Forensic entomologist Lee Goff at Chaminade University of Honolulu, Hawaii, agrees. He says many lawyers, and even forensic scientists, are unaware of the limitations. "The legal community want things to be precise, but they have to remember that we are dealing with estimates," he says.
Rachel Nowak, Melbourne