Microbial and Microbiome Evidence in Forensic Cases

Within minutes of cardiac arrest, the body begins to lose the immune and physiological barriers that normally keep the gut microbiota contained. Bacteria from the gastrointestinal tract, primarily anaerobes such as Bacteroides, Clostridium, and Enterococcus species, begin to breach the intestinal wall and spread through lymphatic vessels and blood into tissues. This early microbial translocation is followed by a broader colonisation as putrefaction advances and body tissues liquefy, creating nutrient-rich anaerobic environments that favour successive waves of bacterial species.

The key insight for forensic science is that this succession follows a broadly predictable trajectory. Work by Metcalf and colleagues (2013, 2016) using a mouse model and later human cadavers showed that microbial community composition at death-proximate body sites changes with PMI in a statistically consistent way. By profiling 16S rRNA amplicons from tissue samples and fitting the community data to a regression model, the researchers estimated PMI to within approximately 3 days across a 48-day window, even with variation in temperature and body size. Subsequent studies on human outdoor cadaver facilities in the United States, Germany, and Australia have broadly replicated this finding, though environmental conditions substantially affect the rate and pattern of succession.

The body site matters. The gut, rectum, oral cavity, and skin each show distinct succession patterns and have been studied at different levels of resolution. The rectum has received particular attention because it is largely protected from environmental contamination and tends to preserve a consistent anaerobic community for longer than skin surfaces. Nasal and oral sites are colonised more rapidly by environmental bacteria, making their microbial signals noisier but potentially informative about the environmental context of decomposition.

When a body decomposes in soil, it releases large quantities of nitrogen compounds, lipids, volatile fatty acids, and water, fundamentally altering the local soil chemistry. This chemical change drives a shift in the microbial community: nitrogen-fixing and ammonia-oxidising bacteria increase, and the overall diversity pattern changes compared to undisturbed control soil at the same site. The altered zone, called the cadaver decomposition island (CDI), retains a detectably different microbial signature for months to years after the body has been removed.

Researchers at several outdoor decomposition facilities have demonstrated that soil microbiome analysis can distinguish grave soil from control soil at the same location and can detect the CDI long after the surface has been recolonised by vegetation. Carter and colleagues (2015) showed that CDI-affected soils maintained altered fungal and bacterial community signatures at least a year post-decomposition. This has practical applications for cold cases where remains have been removed or scattered: the microbiome of the soil retains an evidence record that is not visible to the naked eye.

Geographic provenance is a separate application. Soil microbial communities differ systematically between regions, driven by climate, vegetation type, and geology. Studies using machine learning classifiers trained on large soil microbiome datasets have achieved geographic assignment of soil samples to broad regions with accuracy far above chance. If soil found on a suspect's clothing or vehicle can be characterised by its microbial community, it may be possible to indicate whether that soil is consistent with originating from a particular type of environment or region. This is complementary to physical and chemical soil forensics covered in forensic geology and is beginning to appear alongside soil mineral analysis in case reports.

Every person carries a partially individualised microbiome on their skin, in their gut, and in their oral cavity. The Human Microbiome Project, which sequenced microbiome samples from hundreds of healthy volunteers, demonstrated substantial inter-individual variation in community composition, particularly in the gut. This variation has raised the question of whether a microbiome profile could identify an individual in the same way a DNA profile does.

The current evidence suggests partial identifiability over short time periods. Franzosa and colleagues (2015) showed that combinations of microbial strains from the gut microbiome could re-identify individuals from a dataset with high accuracy over a period of one year, and that skin microbiome profiles on hands were partly consistent across samples collected weeks apart. However, the microbiome is considerably more variable than the human genome: it changes with diet, antibiotics, illness, travel, and ageing, and two people living together rapidly converge in their microbial communities because they share food, surfaces, and each other.

The implications for forensic science include: skin microbiome transfer from a suspect to a victim or surface, microbiome analysis of touch deposits that lack sufficient nuclear DNA for STR profiling, and post-mortem microbiome analysis to infer lifestyle factors such as diet or antibiotic use history that might assist identification. None of these applications has yet achieved the validation standard required for primary identification evidence in court, but they are active research areas. The parallel with touch DNA, covered in Touch DNA and Trace Biological Material, is instructive: transfer, persistence, and secondary contamination are the same fundamental challenges.

The dominant analytical approach in forensic microbiome research is 16S rRNA amplicon sequencing. A hypervariable region of the 16S ribosomal RNA gene, most commonly the V3-V4 or V4 region, is amplified by PCR from total DNA extracted from the sample. The amplicons are sequenced on an Illumina platform, producing hundreds of thousands of short reads per sample. These reads are grouped into operational taxonomic units (OTUs) or amplicon sequence variants (ASVs) and matched against a curated reference database such as SILVA or Greengenes to assign taxonomic identities. The result is a table of taxon relative abundances that describes the microbial community composition.

Shotgun metagenomics is a more comprehensive alternative. Instead of amplifying a single marker gene, all DNA in the sample is fragmented and sequenced. This generates data from bacteria, fungi, viruses, archaea, and any host DNA present. Shotgun metagenomics can detect finer-scale genetic variation below the species level and can characterise the functional gene content of the community, not just its taxonomic composition. It is more expensive and produces larger datasets, but it avoids the PCR amplification biases inherent in 16S approaches and can detect pathogens or other organisms of forensic interest that might not be captured by universal 16S primers.

Bioinformatic interpretation is a major challenge. The reference databases used to assign taxonomic identities are incomplete and unevenly curated, and many forensically important environments contain high proportions of taxa without close database matches. Machine learning models trained to estimate PMI or classify geographic origin are sensitive to the database and parameter choices made during processing, and small changes in analytical pipeline can produce meaningfully different outputs. This variability is a core obstacle to standardisation.

Forensic microbiome evidence faces the same admissibility standards as any novel scientific technique: the methodology must be validated, the error rate must be characterised, and the technique must be generally accepted by the relevant scientific community. In the United States, these criteria derive from the Supreme Court's decision in Daubert v. Merrell Dow Pharmaceuticals (1993) and its progeny. In England and Wales, the Criminal Procedure Rules 2020 and the Law Commission's 2011 report on expert evidence impose equivalent requirements. In India, the Bharatiya Sakshya Adhiniyam 2023 (which replaced the Indian Evidence Act 1872) retains the framework for expert opinion evidence under which novel scientific methods are evaluated by the court on the basis of reliability and relevance. The European Union's approach is governed by member-state procedural law, but scientific evidence standards generally follow similar principles of reliability and transparency.

Microbiome evidence does not yet meet these standards as primary identification evidence. The obstacles are specific: no certified reference database for forensic microbiome analysis, no standardised collection and extraction protocol across laboratories, no published proficiency testing programme, and no large-scale validation study that establishes error rates across diverse environmental conditions and victim populations. Published PMI estimation studies have operated in specific geographic settings and decomposition environments, and it is not known how well their models generalise to different climates, burial conditions, or cause-of-death scenarios.

The practical position in most jurisdictions is that microbiome evidence can legitimately inform investigative direction: it might narrow the window of death, suggest that a body was moved from one environment to another, or indicate geographic origin of soil trace evidence. These uses are analogous to behavioural profiling or forensic geology in that they generate hypotheses rather than proof. If and when standardised protocols, reference databases, and validated error rates emerge, the evidentiary weight will increase. Several research groups and national forensic science agencies in the United States (through the National Institute of Justice), Germany, and Australia are actively working on this validation infrastructure.

Forensic microbiome science does not operate in isolation. Its most direct neighbours are forensic entomology and forensic botany. The same decomposition processes that drive microbial succession also attract insect colonisation and alter plant growth around a grave. PMI estimates from entomological data and from microbiome data can be combined to produce a narrower confidence interval than either method alone, and their complementary failure modes make them mutually validating: entomological methods fail when insect colonisation is prevented (sealed containers, deep burial) while microbiome methods may be degraded by extreme temperatures that preserve insect evidence well. The interactions between decomposition ecology, entomology, botany, and microbiology are covered in more depth at Forensic Entomology and Forensic Botany.

Wildlife forensics represents another convergence point. Illegal wildlife trade cases frequently involve determination of species identity and geographic origin of biological material. The microbiome of gut contents or hide samples may reveal geographic provenance independent of morphological or genetic species identification, and metagenomic approaches are already being used in environmental DNA (eDNA) work to detect the presence of protected species in water or soil samples without requiring recovery of the animal itself. These applications are an extension of the same sequencing and bioinformatic workflows used in human forensic microbiome analysis and are treated in Wildlife Forensics.

Forensic anthropology intersects with the microbiome when remains are skeletonised or fragmentary. At that stage, the thanatomicrobiome that was active in soft tissue is largely gone, but the bone surface and internal cancellous architecture may retain microbial activity traces that can be examined histologically and molecularly. Post-mortem interval estimation from skeletal remains using microbiome methods is at a very early research stage compared to the soft-tissue work but represents a logical extension of the field. Bone, teeth, and tissue preservation are covered in the context of Forensic Anthropology.