Assessing the value of bacteria, plants, fungi and arthropods characterized via DNA metabarcoding for separation of forensic-like surface soils at varied spatial scales

Abstract

Geological materials, including soil and dust, are ubiquitous and often inadvertently transferred during crime events. Forensic geologists use a range of particle-based analytical approaches to characterize the inorganic fraction for sample-to-sample comparisons. While analysis of the inorganic component often provides sufficient information to answer case related questions, there are inevitably cases where analysis of the organic fraction could be beneficial. In this study we assessed whether the biological communities associated with the organic fraction of North Carolina surface soils could allow for sample separation at four spatial scales: 1) sub-site samples – three samples collected 1 m apart, 2) small – A and B sites at a single location in close proximity (median distance of 66 m apart), 3) moderate – multiple locations within the same research station (median distance of 1.1 km apart), and 4) large – multiple locations within each physiographic region (Coastal Plain, Mountains, Piedmont) separated by up to 250 km. DNA metabarcoding coupled with Illumina® sequencing was used to amplify ∼200 bp regions of the following in each surface soil sample (n = 180): 1) 16S for bacteria, 2) ITS1 for fungi, 3) ITS2 and p6 loop of the trnL-UAA intron for plants, and 4) COI for arthropods. Generated sequences were taxonomically classified and subsequently organized into eight taxonomic combinations each analyzed at four data levels (unique sequence level and family rank, with and without removal of the 25 % most common taxa). Beta diversity was used as input for ANOSIM and PCA analyses to determine the optimal biological communities and data level to permit separation across the four spatial scales. The biological communities that permitted separation between samples varied based on the spatial scale being examined. At small spatial scales, datasets that included bacteria had the highest success of separating samples. At moderate spatial scales, the combination of plants + fungi or plants alone provided the highest success of separation. Finally, at the examined largest spatial scale, all combinations of biological communities tested performed comparably, likely due to the large diversity in habitat, land-use and geologic parent material between locations. Across all examined spatial scales, the optimal biological community combinations performed best when analyzed at the most granular unique sequence level. The results of this study demonstrate the value of biological communities for differentiating soils across varying spatial scales, providing a complementary analysis method for samples where the inorganic fraction lacked exclusionary differences.