Data compilation
We used R2 values from linear fitting as indices of species abundance estimation accuracy based on eDNA quantity because the values were calculated in most studies collected in our meta-analysis. If the manuscripts only reported Pearson’s correlation coefficients, we squared the coefficients and substituted them for R2values. We then extracted the information on collection strategy (filtration or centrifugation), quantification strategy (species-specific or metabarcoding assay), target taxa, filter pore sizes used for water filtration (μm), volume of water sample (mL), PCR amplicon sizes (base pair; bp), study environments, and abundance metrics (biomass or individuals) from the studies. As all the metabarcoding studies in our meta-analysis collected aqueous eDNA by filtration, collection and quantification strategy were combined as the factor ‘assay strategy’, which was classified as ‘species-specific/filtration’, ‘species-specific/precipitation’, and ‘metabarcoding’. Taxa were classified as fish, herptiles (i.e., reptiles and amphibians), crustaceans, mollusks, and coral and seastars. In studies involving aqueous eDNA collection via centrifugation, the filter pore size was regarded as 0 μm. Study environments were classified as laboratory (i.e., tank, aquarium, or mesocosm experiments), lentic freshwater, lotic freshwater, and marine environments. Multiple R2 values based on different experimental conditions within the same study (e.g., species, filter type, and amplicon size) were treated individually. Moreover, we calculated the sample size (the number of water samples or sampling sites) required for calculating R2 values or coefficients based on figures and/or text in the corresponding literature.