Comparison of established methods for quantifying genotyping error rates in wildlife forensics

Abstract

Several methods have been applied to calculate genotyping error rates (GER) for non-invasive population size estimations. However, there is a lack of comparability between these methods. Here we focused on the comparison of methods for determination of GER within one study using faeces samples of wild boars (Sus scrofa). Error rates were calculated by (1) comparison of reference tissue samples and rectum faeces samples (2) the number of deviations between replicates and the assumed consensus genotypes, (3) re-analysis of a subsample interpreted by allelic and genotype comparisons, and (4) a blind-test of anonymously subdivided faecal samples. The error rates differed widely between these four methods (0–57.5 %) and underline the need of a consensus approach. The blind-test resulted in a GER of 4.3 %. We recommend conducting such a blind-test for estimating realistic GER when starting a pilot study in wildlife forensics.