Soil environmental DNA metabarcoding can quantify local plant diversity for biomonitoring across varied environments

Biomonitoring of plant communities is essential when evaluating outcomes of ecological restoration, but morphology-based inventories can impede consistency across spatial and temporal scales. The Eden Project botanic garden (Cornwall, UK) offers an opportunity to test whether soil environmental DNA (eDNA) reflects plant species richness, community composition, and diversity. Furthermore, it enables the exploration of whether community complexity, environmental conditions, or sampling scale influence the similarity of soil eDNA-derived results compared with those derived from classic botanical inventories. Survey plots were established encompassing a mixture of high- and low-complexity plant assemblages, under either hot and humid or cool and dry conditions. Botanical inventories were conducted in quadrats up to 7 m around a soil eDNA sample plot. Soil samples were processed by metabarcoding of trnL p6 loop to determine plant eDNA identity and relative abundance. Differences in diversity and community composition were calculated before determining whether (1) inventory scale; (2) plant assembly complexity; (3) ambient conditions; (4) soil chemistry; (5) bulk root biomass and (6) soil respiration influenced the similarity of soil eDNA and botanical inventory methods. Inventory scale significantly determined the dissimilarity in both diversity and composition, with soil respiration, phosphorus, and C:N ratio also accounting for variation across sites. Our results demonstrate that soil eDNA metabarcoding is an effective approach for quantifying relative diversity in plant communities across broad habitat types and could be integrated into restoration monitoring strategies, but that its application requires an appropriate a priori knowledge of survey sites to best tailor soil eDNA survey implementation.