Lake sediment DNA as a proxy in fish population studies: analytical challenges and opportunities

Abstract Sedimentary environmental DNA (sed-eDNA) coupled with metabarcoding is increasingly exploited for ecological studies, but application of the method to resolve fish dynamics in lakes still needs better validation. This study (1) evaluated the sed-eDNA yields from the commonly used DNeasy PowerSoil DNA Kit from mineral-rich and organic-rich sediments and (2) examined the viability of fish sed-eDNA recovery and detection in surface sediment samples from 13 Swedish mountain lakes, with organic contents of 18–52%, by using conventional PCR and droplet digital PCR. Based on concurrent fish-population surveys these lakes contain arctic char and brown trout. We show that, compared to other specifically designed lysis buffers, the DNeasy PowerSoil DNA Kit is less effective to recover DNA from organic-rich sediments and almost 50% of the extracted DNA was lost during purification steps. The amplification of fish sed-eDNA using conventional PCR with teleo primers failed to detect positive signals; whereas ddPCR assays enabled quantification of amplifiable DNA in all the extracts. However, further molecular cloning of the positive ddPCR droplets from one sediment sample revealed amplified sequences of unidentified origin that cannot be aligned well to fish. Thus the performance of the teleo primers for quantification of fish sed-eDNA detection requires further examination. For detection of fish sed-eDNA for ecological studies, we suggest that DNA extraction methods and primers should be carefully selected and the performance of ddPCR to detect DNA at low quantities needs to be further scrutinized to circumvent the pitfalls of false positives.