Abstract
Sensitive and efficient monitoring of species’ distributions is crucial for understanding biodiversity and conserving species. Environmental DNA (eDNA) metabarcoding enables rapid detection of biological communities and its use is increasing in freshwater ecosystems. However, variations in eDNA sampling method may influence results. We compared anuran species richness estimates derived from two eDNA filtration strategies: (i) five Sterivex 0.22 µm filters from five locations per wetland using a manual syringe, and (ii) one Smith-Root 5 µm filter applied to a pooled sample from five locations per wetland using a computer controlled eDNA sampler (halved for analysis). Additionally, we assessed whether the number of sampling locations influenced species richness estimates using a pooled sampling approach and 5 µm filter. There was no difference in the total sample volume of the two filtration methods, yet the likelihood of detecting anuran species was higher when using the 5 µm system compared to 0.22 µm. Species richness also increased with the number of sampling locations, and was higher when using an amphibian-specific assay. These findings demonstrate that eDNA filtration technique can impact anuran detection, and that the use of 5 µm filters enables sample pooling which will be particularly cost-effective for large-scale or long-term applications.