Key points are not available for this paper at this time.
Water pollution and its ecological consequences have become central concerns in watershed governance and aquatic ecosystem conservation. Conventional ecotoxicological research on water pollution has long relied on physicochemical monitoring, laboratory-based single-species exposure tests, and morphology-based biological surveys. Although these approaches have provided essential support for pollutant identification, toxicity characterization, and environmental standard setting, they remain insufficient for resolving community-level responses, food-web perturbations, and ecosystem degradation under multiple-stressor conditions. Environmental DNA (eDNA) has emerged as a promising molecular tool because it is non-invasive, highly sensitive, high-throughput, and capable of detecting multiple taxa simultaneously. In aquatic systems, eDNA applications have expanded from biodiversity detection to pollution diagnosis, ecological health assessment, restoration monitoring, and early warning of ecological risk, while increasingly being integrated with eRNA, multi-omics approaches, machine learning, hydrological modeling, and ecological network analysis. However, several challenges still constrain its broader application, including incomplete methodological standardization, false-positive and false-negative detections, insufficient reference databases, limited quantitative capacity, scale mismatches caused by transport and mixing, and difficulties in causal attribution. This review synthesizes recent progress in the use of eDNA for water-pollution research, with emphasis on its technical workflow, major application domains, integrative analytical frameworks, and methodological boundaries. More specifically, three main points are highlighted: (1) eDNA is shifting water-pollution research from single-species toxicity characterization toward community- and ecosystem-level ecological interpretation; (2) its greatest value lies in its integrative role at the interface of biodiversity monitoring, ecological risk assessment, and management-oriented decision support; and (3) future progress will depend on improvements in standardization, quantitative inference, regional reference databases, and multi-source data integration. Overall, this review clarifies how eDNA can contribute to more robust, ecologically meaningful, and management-relevant assessment of water pollution.
Zhang et al. (Wed,) studied this question.