Reducing read abundance biases. A Processing mock communities (bottle) with defined composition allows determining taxon-specific correction factors, which can be applied to correct relative read abundance of samples with unknown composition, indicated by a red line. Correction factors can only be determined for taxa included in the mock community. B Correction factors can be determined using iterative algorithms and a guess-and-test approach based on a morphological reference data set (not shown). The correction factors can be applied to correct relative read abundance of samples with unknown composition, indicated by a red line. Correction factors can only be determined for samples that show a good agreement in terms of taxa detected between the reference and the DNA metabarcoding data set. C Adding spike-ins, e.g. a defined amount of genomic DNA, to all samples and co-amplifying and co-sequencing the reference material allows correcting raw read counts by simply dividing read counts assigned to taxa (blue and brown bars) by read counts assigned to the spike-in (red bars); abbreviations: RRA – relative read abundance, S – sample, artwork: Alice Scherges.