Forensic science is the application of science to matters of law and within forensic science, there are several disciplines including biology and geology. Traditional forensic geology examinations focus on the inorganic components found with geologic materials, such as soil and dust. Bulk soil morphology, such as color and texture, can be identified during the examination, including the mineralogy found within these materials. Unfortunately, the biological components often go unidentified, as expertise is minimal in the United States. This dissertation sought to use DNA-based methods in the forensic analysis of non-human biological materials. Specifically, this involved using DNA metabarcoding to analyze environmental DNA (eDNA) containing bacteria, fungi, plants, and arthropods recovered from forensically relevant geologic materials. Prior to the implementation of DNA metabarcoding in forensic casework, several gaps needed to be addressed. Firstly, the methods used by forensic geologists to remove soil adhered to evidence for traditional geologic analyses, needed to be assessed for suitability for downstream eDNA analysis. We demonstrated that picking and scraping with forceps and metal scoopulas is an effective method at removing soil from mock evidence and was most suitable for eDNA analysis. Following the identification of this method, five mock evidence items containing soil and dust were collected monthly from two study sites in Raleigh, NC over a one-year period. To characterize the eDNA from these mock evidence items, we developed a complete workflow from DNA isolation to bioinformatic processing of next generation sequencing reads. We employed the use of indexed primers to simultaneous amplify and prepare libraries in a single polymerase chain reaction for five different barcode regions: 16S (bacteria), ITS1 (fungi), ITS2 (plants), trnL UAA intron (plants), and COI (arthropods). Using this workflow, we were able to taxonomically identify bacterial, fungal, plant, and arthropod communities in mock soil and dust evidence in which we found: a) soil and dust can be differentiated using bacteria, fungi, and plants individually or in combination, b) sites can be differentiated using individual taxa or in combination, and c) plant eDNA in dust is stable across a one-year period. To further support the adoption of DNA metabarcoding in forensic laboratories, we developed and validated four dye-based qPCR assays to separately quantify the total amount of bacterial, fungal, plant, and arthropod DNA found in single source specimens and mixed source environmental samples. These assays were found to be specific for the taxon of interest (in vitro and in silico), reproducible, and sensitive. Standards were also developed using synthetic double stranded DNA (gBlocksÔ) and the linear dynamic range was determined. Melt curve analyses were also assessed to help identify any non-specific amplification. This research has paved the path forward in the eventual adoption of DNA metabarcoding in forensic casework as a supplement to forensic geology examinations, but further experimentation is needed to assess these approaches on a broader scale.
(Author abstract provided.)