Quantification of total fish eDNA concentration to optimize the detection of rare species and sampling effort

Didier PONT 

eDNA metabarcoding is now a widely used method for monitoring aquatic biodiversity, but detecting rare species from extra-organism eDNA samples remains a challenge. Yet, rare species, whether threatened or newly colonizing organisms, are the primary targets of biological conservation measures. Replication and sample volume, as well as PCR replication, are among the main parameters studied to optimize eDNA sampling strategies and workflows. To test the hypothesis that the total abundance of fish eDNA copies in a PCR aliquot is a determining factor for the detection of rare fish species, we replicated water samples from three sites and applied a metabarcoding protocol combined with qPCR measurement of the total fish eDNA abundance amplified by our universal primer, allowing to estimate the number of specific eDNA copies in the eluate. Simulations of the species-specific probabilities of detection (multi-species occupancy model) showed that, even for a high effective sequencing depth of approximately 380,000 reads per sample, more than around 2000 eDNA copies per aliquot are required to detect very rare species (< 1 ‰ of the total number of DNA copies in the eluate). This threshold is lowered by a factor of around 4 and 20 for rare (1 ‰≤ < 1 %) and abundant species (> 1 %), respectively. Our results highlight that increasing sampling and technical replication efforts are required to detect rare species. The total volume of water samples required for species detection varies from less than one litre to 100 litres, depending on the eDNA concentration in the environment. Estimating the total concentration of fish eDNA per litre in the studied water body prior to a metabarcoding sampling campaign would enable optimization of both the volume and number of water samples to be collected, as well as the volume of eDNA aliquot per PCR.