Hair vs. fabric: substrate-dependent microbiome shifts dictate the accuracy of forensic body-fluid identification.

Hair and clothing are among the most frequently recovered evidentiary items at crime scenes. Compared with human soft tissue, they resist environmental degradation and frequently retain perpetrator-derived biological deposits; accurate identification of these traces is therefore pivotal for case qualification and investigation. Microbiome profiling has emerged as a promising forensic tool for body-fluid attribution, yet body fluids like blood and semen contain only a sparse indigenous flora and are highly vulnerable to environmental or substrate-borne microbial overwrite. To date, systematic evaluations of how deposition surface and ambient microbiota influence the reliability of microbe-based fluid identification remain scarce, especially with respect to hair-a substrate that inherently carries the victim's resident microbial community and may obscure fluid-specific markers. In this study, four forensically relevant body fluids (blood, semen, vaginal fluid and saliva) were deposited on hair shafts and cotton fabric and aged for 30 days under indoor conditions. Amplicon sequencing of the V3-V4 hypervariable region of the bacterial 16S rRNA gene revealed that fabric-hosted stains retained a stable, fluid-specific microbiota across all sampling intervals. In contrast, hair-associated traces underwent a rapid and persistent compositional shift toward the native scalp/hair community, resulting in significant loss of fluid-identifying signals. Consequently, the prediction accuracy of our random-forest classifier decreased to 84.2% when hair samples were included. Saliva and vaginal fluid proved exceptional: a subset of oral-associated taxa (Streptococcus, Gemella) and vaginal associated microorganisms (Lactobacillus) remained detectable on both substrates, preserving a degree of fluid specificity. Collectively, these findings demonstrate that substrate-derived microbiota can compromise microbiome-based body-fluid identification, underscoring the necessity of matrix-specific marker panels and cautious extrapolation of signatures derived from pristine laboratory simulations to real-world evidentiary samples.