eDNA replicates, polymerase and amplicon size impact inference of richness across habitat

Environmental DNA-based monitoring has been increasingly used in the last decade to monitor biodiversity in aquatic and terrestrial systems. Molecular-based surveys now allow quick and reliable production of baseline knowledge of species community composition on a large scale, allowing better understanding of ecosystem function and mitigation of stressors linked to anthropogenic activities. Despite this, technical hurdles often remain, and the impact of replicates, PCR polymerases and amplicon size on the recovered species richness is still poorly understood. Here, we conducted a large controlled experiment, with bulk samples collected from terrestrial, marine and freshwater environments to assess the impact of natural and technical replicates, PCR polymerases with different degrees of fidelity or proofreading activity, as well as amplicon size on species richness recovery across habitats. In this study, we consistently found variations in sample species richness depending on PCR polymerase choice. We further demonstrate the dissimilar impacts between natural and technical replicates on species richness recovery, and the necessity of increasing natural replications in eDNA based surveys. We highlight the benefits and limitations of replication strategies, polymerase choice and amplicon size across terrestrial, marine and freshwater habitats, and provide recommendations to increase the reliability of future eDNA-based metabarcoding studies.