Environmental DNA (eDNA) metabarcoding has become a powerful tool for examining fish communities. Prior to the introduction of eDNA-based assessments into regulatory monitoring contexts (e.g., EU Water Framework Directive), there is a demand for methodological standardization. To ensure methodical accuracy and to meet regulatory standards, various sampling, laboratory and bioinformatic workflows have been established. However, a crucial prerequisite for comprehensive fish monitoring is the choice of suitable primer pairs to accurately identify the fishes present in a given water body. Various fish-specific primer pairs targeting different genetic marker regions were published over the past decade. However, a dedicated study to evaluate the performance of frequently applied fish primer pairs to assess Central European fish species has not yet been conducted. Therefore, we created an artificial 'mock' community composed of DNA from 45 Central European fish species and examined the detection ability and reproducibility of five primer pairs. Our study highlights the effect of primer choice and bioinformatic filtering on the outcome of eDNA metabarcoding results. From the five primer pairs evaluated in our study the tele02 (12S gene) primer pair was the best choice for eDNA metabarcoding of Central European freshwater fish. Also, the MiFish-U (12S) and SeaDNA-mid (COI) primer pairs displayed good detection ability and reproducibility. However, less specific primer pairs (i.e., targeting vertebrates) were found to be less reliable and generated high numbers of false-positive and false-negative detections. Our study illustrates how the careful selection of primer pairs and bioinformatic pipelines can make eDNA metabarcoding a more reliable tool for fish monitoring.
