Search for a command to run...
Abstract Increasingly, molecular methods of species monitoring are integrated into freshwater biodiversity surveys and fisheries management. Inferring organism abundance or biomass from sequence counts derived from metabarcoding data has been an exciting but contentious concept in the biomonitoring community for some time. Although demonstrating a strong correlation with abundance has proven difficult, many researchers have assumed that quantitative metabarcoding data can at least provide broad‐scale abundance ranking of a species across sites. However, robust field validations of this widely held assumption remain scarce. Here, we analyse metabarcoding read counts of fish eDNA data derived from 20 lakes and use beta‐binomial mixed effects models to compare this to rank abundance generated from long‐term fish survey data. Rank abundance data for 18 species was generated within‐species across sites, meaning that ranks compare the abundance of the same species in different lakes. We also investigated a possible allometric effect on eDNA production by analysing a subset of data for effects of fish body mass on the proportion of eDNA sequences. We found a significant relationship between species‐specific eDNA sequences and within‐species rank abundance categories for fishes, with rare fish producing 2% of sequences in a library, moderately abundant fish producing 7%, and abundant fish producing 22%, according to model predictions. We found a small negative effect of body mass on the amount of eDNA sequences, where the proportion of reads recovered significantly decreased with increased mean body mass of the population. Synthesis and applications . This approach provides a practical tool for managers to rapidly assess rank abundance of freshwater fishes, offering particular value for monitoring smaller species that are often missed by conventional surveys.