High frequency environmental DNA metabarcoding provides rapid and effective monitoring of fish community dynamics

Long‐term monitoring is critical to measure the response of biodiversity patterns and processes to human‐mediated environmental pressures. This is particularly pertinent in freshwaters, where recent estimates indicated a third of all fish species are threatened with extinction, making ongoing biomonitoring essential for conservation management. High frequency annual monitoring is critical for identifying temporal changes in fish community composition; however, traditional survey methods are typically less practical over such timeframes. While environmental (e)DNA measurement represents a potentially powerful tool for monitoring temporal community dynamics, studies are lacking. To address this deficit, we generated a high frequency time‐series dataset of entire fish communities using eDNA metabarcoding, to directly assess the repeatability and sensitivity of this method for detecting annual population trends. We targeted two differing environments (freshwater vs. intertidal) within the Thames catchment, UK, where detailed historical records from traditional monitoring were available for comparison. To test how robust eDNA data is for inferring the known community, we applied a hierarchical, nested design encompassing short and longer‐term variation in eDNA data. Our analyses showed that irrespective of environment, eDNA metabarcoding represented known seasonal shifts in fish communities, where increased relative read abundance of eDNA coincided with known migratory and spawning events, including those of the critically endangered native species Anguilla anguilla (European eel). eDNA species detections across a single year included over 75% of species recorded in a ca. 30‐year historical dataset, highlighting the power of eDNA for species detection. Our findings provide greater insight into the utility of eDNA metabarcoding for recovering temporal trends in fish communities from dynamic freshwater systems and insight into the potential best sampling strategy for future eDNA surveys