Can Airborne eDNA Be Used to Assess the Impacts of Invasive Species on Biodiversity?

Anthropogenic impacts on landscapes create opportunities for invasive species to spread, causing community biodiversity changes. Biodiversity alterations contribute to ecosystem degradation, creating a need to better understand the impacts of invasive species on ecosystems. Hemlock woolly adelgid, Adelges tsugae , is an invasive insect responsible for devastating millions of eastern hemlocks, Tsuga canadensis , in North America. Eastern hemlocks are a foundation tree species found in sensitive habitats and riparian zones. Hemlock woolly adelgids feeding on hemlocks can cause tree mortality within several years. As hemlock mortality increases, we expect a shift in plant and arthropod communities in hemlock woolly adelgid‐infested areas. We used airborne environmental DNA to investigate hemlock woolly adelgid impacts on plant and arthropod community compositions in west Michigan coastal forests. We deployed 100 environmental DNA traps across 15 sites in the summer of 2023 to sample plant and arthropod taxa. Our sites included hemlock stands with no known infestations and stands with varying treatment and infestation levels. We sequenced ITS2 and CO1 regions to identify plant and arthropod taxa at the genus level, respectively. We found significant differences in the biodiversity of observed plant and arthropod genera across all sites. When comparing genera detected between high‐risk and infested sites, there were significant differences between arthropod genera but not plant genera. For the plant communities, this lack of difference may be because the infestation is relatively recent, and we are only starting to see changes in hemlock health across some sites. For the arthropod communities, some of these differences may be due to the presence of hemlock woolly adelgid and/or treatment in the infested sites. Overall, environmental DNA could be a powerful tool to evaluate how invasive species and their management alter biodiversity, but further optimization of these methods is needed.