Criminals don’t always pay attention to detail: a speck of soil on the sole of a shoe or a spatter of mud on somebody’s skin can allow soil analysis specialists to confirm whether a person or object was part of a crime scene. And they do not need much material to work with. As techniques have evolved over recent years, an amount of soil equivalent to a grain of rice is now enough to identify the origin of a sample. The main advances in this field—known as forensic soil science or forensic soil analysis—have been made over the last two decades. Much progress has occurred in countries where the techniques are more widely used, such as Australia, the UK, and the USA, and is generally motivated by the need to give the police access to new and more accurate tools for analyzing the elements of a crime scene or suspect. More recently, police and forensics officers in Brazil have begun to adopt these strategies as a result of collaborations with agronomist Vander de Freitas Melo, a retired professor at the Federal University of Paraná (UFPR), and other researchers taught by him.
Melo, a specialist in soil chemistry and mineralogy, was approached by Leila Barbar, an expert from the Criminalistics Institute of Paraná, in the mid-2000s. In several investigations she was involved with at the time she found soil on fabrics and shoes, and she was interested in finding out whether it would be possible to create a map of the soil profiles in the Metropolitan Region of Curitiba that could be used to identify the origin of samples found on objects potentially used by suspects. Melo was fascinated by the idea and began the meticulous work of collecting soil samples from areas with high crime rates. He teamed up with another specialist in the area and a chemist, and in collaboration with Barbar, they developed a methodology for characterizing soils specifically for forensics in the state capital of Paraná.
Described in an article published in the journal Forensic Science International in 2008, the strategy involved assessing the soil’s physical characteristics, such as color and texture, as well as analyzing its chemical and mineral properties. In a crime scene simulation, it was used to analyze samples taken from two neighborhoods in Curitiba and from two nearby cities: Colombo and São José dos Pinhais. In a blind analysis of a minimal amount of material—just 1 gram (g)—the researchers were able to identify where they were from with a high level of confidence. Soil is generally classified into three size categories: sand, silt, and clay. In the study, the group found that by analyzing the smallest particles (clay, which is highly responsive to the analyses due to its size and ability to inherit the characteristics of the minerals present in the rocks where it forms), they were able to group samples from the same location more accurately and distinguish them from samples collected in other areas.
“The work was seminal to the study of soils for forensic purposes in Brazil. It established well-defined technical criteria for soil collection and analysis,” says geologist Fábio Salvador, who worked at the São Paulo Institute for Criminalistics and was technical and scientific director of the Federal Police in Brasília from 2019 to 2020. Salvador, who is also a member of the Initiative on Forensic Geology, an international group of researchers and experts interested in encouraging and disseminating geology studies that aim to help obtain and analyze criminal evidence, learned about Melo’s work in the middle of the last decade and began working in partnership with the UFPR team. “Previously, these analyses were carried out sporadically on the individual initiative of the investigator, with no widespread or discussed criteria,” says the geologist, who currently works or the technical and scientific department of the Curitiba Federal Police.
Samara Alves Testoni
Electron microscopy image of sand collected from a simulated crime scene: the shape and size of the particles are used to identify its originSamara Alves TestoniEven after successful simulations, it was some time before the methodology was evaluated in practice. The opportunity for a real-world test only presented itself years later, when the scientific department of the Paraná Police asked Melo and his team for help analyzing soil found on a safe stolen from a bank in Araucaria, part of the Metropolitan Region of Curitiba, in December 2016. The safe had been discarded in a rural area of the municipality a few days after being opened. Soil samples were taken from the surface of the safe, from a vehicle suspected of having been used to transport it, and from a property where it may have been opened. A comparison of the three samples, however, did not result in a clear conclusion. “The case showed that the sample collection strategy needed to be improved and the procedures standardized,” says agronomist Samara Alves Testoni, who analyzed the case during her doctorate, which was supervised by Melo and included a spell at the James Hutton Institute in Scotland, where she worked under geographer Lorna Dawson, an internationally recognized researcher in the field of forensic soil analysis.
With the help of Dawson, Salvador, and other Brazilian forensics experts, Testoni and Melo adjusted their collection strategy, perfected their analysis methods, and drafted a standard operating procedure (SOP) for what to do when soil traces are found at a crime scenes. In the procedure, the researchers describe in detail how and where to collect the sample, the equipment to use, and the appropriate storage methods. When the initial version was completed, it was presented at forensic science conferences and refined based on feedback. At about 20 pages long, the final version was published in the journal Revista Brasileira de Ciências do Solo in 2019. In another test using simulated crime scenes, researchers and investigators independently collected soil samples following the SOP guidelines. A member of the UFPR team who did not participate in the field expeditions then examined the material without knowing which group collected which samples and reached the same conclusions.
Another real case, this time a homicide, provided an opportunity to validate the procedure and verify that it was viable in real-world situations. In July 2016, a man was killed and dismembered in Colombo and his body parts buried in different places in rural Araucaria. The victim’s ex-wife later confessed to the murder and told police where she had buried the body parts. Soil stuck to the shovel she used matched samples taken from the burial sites, backing up her version of events, the UFPR researchers said in an article describing the case published in the Journal of Forensic Sciences in 2020. The technique seemed to work well and was put to the test once more in another murder, a femicide that took place in 2018.
In the early hours of May 9, a woman disappeared in Colombo. Her body was found almost a month later, about 50 kilometers from the city, in a ditch at the side of a road between Curitiba and the coast of Paraná. On May 19, the victim’s ex-boyfriend, a military police officer, was arrested on suspicion of her murder. Security cameras from the building where she lived captured her leaving in the night, accompanied by her ex-boyfriend. The accused, however, denied the allegations and claimed he had not been on the road where her body was found. Called in to help search for evidence, the UFPR team looked for soil samples inside the suspect’s vehicle, where they can often be found. “The car pedals and carpets were clean,” recalls Testoni, who helped collect the samples and is currently doing a postdoctoral fellowship in forensic sciences at the Federal University of Pelotas (UFPEL).
On the outside, however, there was a small amount of soil inside one of the fenders and on one of the wing mirrors—about 0.5 grams in total. Preliminary mineralogical analyses revealed that the samples contained potassium feldspar, a mineral typical of granite soils, a rock commonly found in the area the victim’s body was dumped. Soil samples were then taken from different points along the road and analyzed at the UFPR laboratory. The chemical and mineralogical profile of the soil found on the vehicle matched the samples taken from near where the body was found, the researchers reported in an article published in the journal Science & Justice in 2019. The defendant remains in prison awaiting trial.
Researchers refined their analysis methods and wrote a standard operating procedure for crime scenes
“Fingerprints and DNA samples allow us to identify a person who has commited a crime, while forensic entomology—the study of insect larvae on corpses—helps us establish a time of death. Soil analysis, meanwhile, enables us to connect the suspect or objects the suspect has used to a crime scene,” explains Rodrigo Studart Corrêa, an investigator from the Federal District Civil Police. Corrêa met Melo in 2015 and did a postdoctoral fellowship at the UFPR laboratory. When he returned to Brasília in 2017, he began incorporating forensic soil analysis into police procedure. “We are implementing it very cautiously because no matter the outcome, there is always an impact on someone’s life. A false positive result is damaging to the accused, while a false negative affects the victim’s family,” says Corrêa, who has already helped to obtain evidence in a number of investigations, from low-level crimes to homicides.
In one case, a man was killed and dismembered after a disagreement at a party in Planaltina, near Brasília, in 2020. His body was buried in a suitcase far outside the city. A dirty shovel and a hoe were found in the house of a suspected accomplice. Soil analysis indicated that the traces on the hoe did not match samples from the burial site, but the soil on the shovel did, providing evidence against the alleged accomplice. “Our report helped link the shovel to the grave,” says Corrêa.
Soil analysis in criminal investigations is not new, but the approach has advanced significantly and become more sophisticated in recent decades, earning increased attention in many forensic science institutes. It is said that the ancient Romans were pioneers of this scientific practice, inspecting the hooves of horses from their enemy’s armies to determine where they came from. According to experts in the field, the first documented use of forensic soil analysis in a criminal investigation took place in Prussia in April 1856. A barrel of silver coins was loaded onto a train, but when it arrived at its destination, it was instead filled with sand. German naturalist and geologist Christian Ehrenberg (1795–1876), who was called in to assist with the case, took soil samples from every station the train had stopped at and examined them under a microscope to identify when the exchange had been made, ultimately helping to catch the culprit.
Between then and the early 2000s, little changed. “The use of forensic soil analysis was rare, usually involved a simple visual analysis, and was highly dependent on the knowledge of the specialist,” says Corrêa. Major advances were driven by the work of academic researchers, such as Lorna Dawson and Australian agronomist Robert Fitzpatrick, founder of the Centre For Australian Forensic Soil Science, who adapted techniques used in agriculture to investigative forensics and helped solve crimes that had received international coverage. In one, known as the World’s End case in reference to a pub in Edinburgh, Scotland, that two young women had visited the night they were killed in 1977, Dawson helped to describe the sequence of events surrounding the murder using soil traces collected from the foot of one of the victims and kept as evidence for 37 years. His report, produced in 2014, contributed to another conviction for serial killer Angus Sinclair (1945–2020). “If we can establish soil analysis as routine in police work, maybe it could become as widespread as DNA testing,” says Corrêa.
Scientific articles
MELO, V. F. et al. Chemical, physical and mineralogical characterization of soils from the Curitiba Metropolitan Region for forensic purpose. Forensic Science International. Aug. 2008.
TESTONI, S. A. et al. Validation of a standard operating procedure (SOP) for forensic soils investigation in Brazil. Revista Brasileira de Ciências do Solo. Sept. 4, 2019.
TESTONI, S. et al. The use of a sequential extraction technique to characterize soil trace evidence recovered from a spade in a murder case in Brazil. Journal of Forensic Sciences. June 24, 2020.
MELO, V. F. et al. Can analysis of a small clod of soil help to solve a murder case? Science & Justice. Nov. 2019.
