Assessing fish diversity in the Amazon: The impact of primers and reference libraries in eDNA surveys.

Selecting primers for DNA amplification poses a challenge in environmental DNA (eDNA) studies, particularly in biodiverse tropical ecosystems. Researchers must balance taxonomic coverage, primer specificity, and study feasibility when choosing between a single primer pair and multiple primer pairs. Empirical tests comparing these strategies are essential for improving eDNA protocols, enhancing the detection of aquatic biodiversity, and supporting conservation and environmental monitoring efforts. This study aimed to evaluate how different approaches affect the specificity and efficiency of detecting fish species in eDNA samples from the Teles Pires River basin. We (1) synthesized modified versions of primers commonly used in ichthyofaunal surveys; (2) compared the results generated by these modified primers with the standard primer versions; and (3) evaluated the species detection using two reference sequence databases: a global taxonomic database (GenBank) and a locally constructed taxonomic database containing samples of fish species previously collected in situ. The 12S-V5_mod primers detected the most fish species across the databases (67, 104, and 130 in the Midori2, Local, and joint databases, respectively), followed by the MiFish_mod primers, which detected 47, 101, and 114 species. Finally, the MiFish primers detected 39, 73, and 78 species. However, the 12S-V5_mod primers also had lower taxonomic specificity of fish species and amplified a broader range of vertebrate taxa than the MiFish_mod primers. The modified primers outperformed the MiFish primers, and detectability nearly doubled when the Local reference database was used. These results speak to the critical need for constructing reference databases of regional biodiversity, and the need to incorporate multiple primer pairs in eDNA studies.