ABSTRACT Environmental DNA (eDNA) is a non‐invasive monitoring approach increasingly used to detect marine organisms; however, misunderstandings of the temporal variability in eDNA detection have limited its integration within management decisions. A clearer understanding of the periodicity (e.g., seasonality) and duration (e.g., weeks, months) of species eDNA detection is essential to optimize sampling design and data interpretation. As such, this study aims to provide a representative assessment of optimal eDNA detection windows across diverse taxonomic groups, primers, and geographic regions using eDNA metabarcoding. Coastal marine presence‐absence eDNA data were collected along the Northwest Atlantic coast, in the Bay of Fundy, Scotian Shelf, and Baffin Island. eDNA detection window(s) were defined as unimodal, contiguous months having greater than 75% detection probability and were calculated for each taxon for each primer in each region. Most marine species exhibited short eDNA detection windows (1–2 months). The optimal sampling periods and durations were conserved among closely related species, highlighting the importance of considering biological traits when designing and interpreting eDNA studies. Additionally, primer choice influenced the optimal detection periods, with higher seasonal variation in community composition and detection rates using universal COI and 18S primers compared to fish 16S and 12S primers. These results demonstrate that ignoring seasonal variation may cause false negatives, inefficient sampling, and reduced data comparability across independent studies. Thus, we propose a set of guidelines aimed at the development of optimal sampling designs for coastal ecosystems and the interpretation of trends across datasets.
Morrison et al. (Thu,) studied this question.