Influence of stream characteristics and population size on downstream transport of freshwater mollusk environmental DNA

Many freshwater mussel species are highly endangered. Knowledge of their occurrence and distribution is a crucial prerequisite for their conservation. The use of environmental DNA (eDNA) is a relatively new approach and an effective tool to detect the presence of target species in freshwater environments. This study aimed to establish a molecular identification tool for eDNA for the endangered thick-shelled river mussel Unio crassus Philipsson in Retzius, 1788 and to analyze the relationships between mussel abundance, stream discharge, sampling distance, and freshwater mollusk DNA signal (detectability and DNA quantity) by using binary logistic and linear regression modeling. We additionally correlated turbidity and mollusk DNA detectability. We developed a species-specific eDNA detection system in silico and tested the efficiency and species specificity on water samples collected downstream of extant populations of different sizes (100-42,000 individuals) in Bavaria, Germany. This system allowed detection of very small U. crassus populations over long distances (>3 km) and excluded amplification of DNA from other potentially co-occurring species. Signal strength was most strongly associated with population size and distance of sampling point from mussel population. Additionally, a negative correlation with stream turbidity was evident. This species-specific tool may help in screening of streams for target species that are otherwise difficult to monitor.

Automated summary

The study established a species-specific eDNA detection system for the endangered thick-shelled river mussel and analyzed the relationships between mussel abundance, stream discharge, and sampling distance with freshwater mollusk DNA signal. The system proved efficient in detecting small mussel populations over long distances and excluding DNA from other species. Signal strength was found to be associated with population size and sampling distance, with a negative correlation with stream turbidity.