Human disturbances increasingly threaten freshwater ecosystems in rapidly urbanizing regions, creating a growing need for efficient and scalable biodiversity monitoring. Phytoplankton communities are widely used as ecological indicators, yet traditional monitoring based on net-tow sampling and microscopy is often constrained by limited taxonomic resolution, high labor demands, and declining taxonomic expertise. Environmental DNA (eDNA) metabarcoding has emerged as a promising molecular alternative, but its methodological reliability and potential biases remain insufficiently evaluated in complex urban aquatic systems. Here, we compared three phytoplankton monitoring approaches—net-tow morphology, net-tow metabarcoding, and water-derived eDNA metabarcoding—across 19 river and lake sites in the North Canal Basin of Beijing. eDNA metabarcoding detected 6–8 times more phytoplankton taxa than morphology-based identification and showed higher sampling representativeness and technical reproducibility. Community composition analyses revealed moderate overlap but substantial divergence between eDNA and net-tow metabarcoding datasets, with diatoms contributing most strongly to inter-method differences. Methodological discrepancies were associated with both false-positive and false-negative signals arising from legacy DNA inputs, primer bias, PCR stochasticity, and hydrological transport processes. Despite these uncertainties, the three approaches captured complementary components of phytoplankton biodiversity. We therefore propose an integrated monitoring framework that combines the sensitivity of eDNA metabarcoding with morphological verification of ecologically important taxa, supporting scalable and data-driven biodiversity monitoring in urban freshwater ecosystems. • eDNA detected 6–8× more phytoplankton species than morphology across urban waterbodies. • eDNA showed higher technical reproducibility and sample representativeness than net-tow DNA. • Significant community divergence was observed between eDNA and net-tow barcoding datasets. • False negatives (e.g., key harmful bloom taxa, Raphidiopsis raciborskii ) caused by primer selection and false positives from rare taxa constrain the specificity of eDNA.
Du et al. (Sun,) studied this question.