The past 20 years or so have seen a remarkable revolution in the science of forensics. But while genetic analysis is now routine in criminal investigations, the use of soils – known as geoforensics – has lagged behind. Now, a new initiative is combining hi-tech analytical techniques with GIS databases to create a powerful tool that could eventually rival DNA. Louise Murray reports

 

Ace fictional sleuth Sherlock Holmes was 30 years ahead of his time when he linked the soil on a pair of shoes to the scene of a crime. And even now, another century later, the science of geoforensics – the use of particles of soil or rock in criminal investigations – is still in its infancy. However, it’s about to grow up.

 

The first recorded use of geoforensics in criminal investigation took place in 1908, when a German forensic scientist, Georg Popp was called in to help investigate the murder of a certain Margarethe Filbert in Bavaria.  The police discovered that the prime suspect had three layers of soil on his shoes. The outer contained traces of brick and coal, matching the area around a castle where the murder weapon was found. The middle layer matched soils found next to the body, and the innermost, or oldest, corresponded with soils on the walkway outside the victim’s house. The suspect’s alibi was quickly discredited because his shoes were telling a very different story.

 

Six years later, Dr Edmond Locard, a pioneering French criminologist, established one of the basic principles of forensic science – that when two objects come into contact, each leaves a trace of the encounter on the other, whether it’s dust, fibres, hair or soil. This trace evidence can yield vital information about where a person has been, where they live and work, their diet, gender and, often most important, with whom they’ve been in contact. And when gathered at the scene of a crime, it can provide information about those involved in its commission.

 

So far, so CSI. But although the use of trace evidence collected from boots, shoes, tyres and the like has a long history, it’s only recently that researchers have begun to focus on finding ways to extract the maximum amount of information of one particular form of evidence – dirt.

SoilFit is a new cross-disciplinary initiative coordinated by Dr Lorna Dawson at the Macaulay Institute in Aberdeen. The project is attempting to link, very precisely, soil samples collected at a crime scene or from a suspect with specific geographical locations and the associated vegetation.

Far from being homogeneous brown dirt, soils are extremely complex and variable, even over short distances. Forensic scientists currently use numerous conventional soil-analysis techniques, noting characteristics such as colour and texture, and making microscopic observations (including palynology, the study of living and fossil pollen and spores) and mineralogical measurements. The SoilFit team’s job is to integrate these established methods with the very latest chemical and biological identification techniques.

 

Over the past 20 years, a great deal of forensic research effort has been put into the use of DNA fingerprinting and matching in criminal investigations, to the point where they have become extremely powerful techniques. “This has meant that other evidence types have become the poor relation,” says Dawson, “particularly when it comes to innovation and development. SoilFit will contribute to the ever-growing arsenal of hi-tech weapons in the fight against crime. Who knows, soil evidence could become the new DNA.”

 

However, there is still a place for the old DNA in geoforensic analysis. One way of characterising a specific soil is to create a DNA profile of the microorganisms present within it, whether bacteria or fungi. These profiles are retrievable even after drying of the soil in the lab or after a rainstorm at a crime site.

 

Plant debris in the soil also retains the signatures of the waxes that were once present on the cuticle of the living plants, allowing scientists to differentiate between, for example, soils from heath and heather moorland, grassland or pine forests. These biochemical markers can persist in the soil for thousands of years, and have been verified by carbon dating and cross matching to pollen profiles.

 

A combination of these techniques can direct police investigators to areas of, say mixed birch and scots pine woodland with an understorey of heather. Early results have shown that individual gardens can be identified by soil alone, extending the hope that soil evidence could become much more important in so-called volume crimes such as burglary.

When these techniques are combined with GIS mapping – where a standard Ordnance Survey Mastermap base, traditional soil archive data and underlying geological maps are integrated – a powerful new database can be assembled to assist police.

 

Dawson demonstrated a prototype in Aberdeen. “We focus on the unusual aspects of a soil that might allow us to discriminate it from others,” she explains. In this case, the mineral horneblende was found in an acid soil sample from a hypothetical criminal’s shoe in the Edinburgh area. Geological map layers enable large tracts of land to be eliminated and acid clay-based soils highlighted. Pine resin was also present, indicating that the sample came from an area of coniferous woodland. When these are overlaid on the previous layers, and combined with road and travel time data within the range of the suspect’s alibi, a very small search area results.

Dawson admits, however, that there are gaps in the national soil archives. “They were largely developed to support agriculture,” she explains. “So they focus on rural areas, which isn’t where most crimes are committed.” To fill this gap, degree students are being deployed in urban areas around the country to sample soils in parkland, gardens, commons and industrial scrubland. Early results have shown that individual gardens can be identified by soil alone, extending the hope that soil evidence could become much more important in so-called volume crimes such as burglary.