水体环境DNA监测流域生态系统生物多样性的有效度：以环境微生物为例

Environmental DNA (eDNA) is DNA extracted from environmental sample (e.g. water, soil, air, etc.). It has emerged as a reliable tool for biodiversity investigation, fisheries resources survey, endangered species monitoring, invasion species detection, parasite detection et al in aquatic ecosystem. Sequencing and analyzing aquatic eDNA sampled from a downstream sampling site could reveal the biological assemblages in upstream drainage system. Sequencing and analyzing aquatic eDNA sampled from a waterbody sampling site could reveal the biological assemblages in its catchment (including riparian zone and upland). However, its effectiveness or detection rate is unknown. To investigate the effectiveness of monitoring the watershed biodiversity using aquatic eDNA, we introduced the concept of watershed biological information flow (WBIF), and then analyzed these two types of detection in the framework of two types of WBIF, WBIF from upstream to downstream and WBIF from upland to stream. The effectiveness of two types of detection depends on the efficiency of two types of WBIF. Here, we had a case study on the efficiency of two types of WBIF and the effectiveness of two types of detection in the Shaliu River basin according analyzing the environmental microbial assemblages in water and soil. Results showed that (1) the effectiveness of detecting upstream biological assemblages by sequencing and analyzing eDNA sampled from downstream sampling site (the efficiency of WBIF from upstream to downstream) was impacted by temperature and runoff flux. The effectiveness is lower in spring than in summer and autumn. In spring, only 76% of the bacterial OTU could be detected in 1 km downstream. In summer and autumn, more than 96% of the bacterial OTU could be detected in 1 km downstream. (2) The effectiveness of detecting riparian zone biological assemblages by sequencing and analyzing eDNA sampled from adjacent waterbody sampling site (the efficiency of WBIF from riparian zone to stream) was impacted by season and rainfall. The effectiveness is lower in spring than in summer and autumn. In spring, only 17% of the bacterial OTU in riparian soil could be detected in adjacent river water. In summer and autumn, more than 62% of the bacterial OTU in riparian soil could be detected in adjacent river water. Conclusion based on these results, (1) summer and autumn were the most cost-effective season to monitor upstream biological assemblages using downstream eDNA sample. The sampling sites should be added when the monitoring work was carried out in spring. (2) The rainy days in summer and autumn were the most cost-effective time to monitor riparian zone and upland biological assemblages using aquatic eDNA sample. Moreover, we need pay attention on its effectiveness of detection for reevaluating each monitoring result. "