Meta-analyses of environmental DNA downstream transport and deposition in relation to hydrogeography in riverine environments

Environmental DNA (eDNA) analysis is a promising tool to monitor species distribution and abundance in aquatic ecosystems. These estimates can be biased in lotic environments by the effects of eDNA transport and deposition processes. However, our understanding of eDNA downstream transport dynamics is limited because it can fluctuate greatly depending on target species abundance and eDNA quantity in a source site and the design of aquatic sampling. Furthermore, eDNA deposition into bottom substrates also determines the availability of aqueous eDNA, but knowledge about deposition is much more limited. Consequently, there is little consensus about eDNA downstream transport and deposition, as well as their interactions with hydrogeographic conditions. In this study, we compiled previous papers concerning riverine eDNA transport and synthesised knowledge about eDNA transport and deposition. A few studies have recently applied a hydrodynamic modelling approach and calculated the average length of eDNA downstream transport (S-w) and velocity of eDNA deposition (V-dep). Referring to those studies, we manually calculated, or directly extracted, S-w and V-dep from the nine eDNA papers and conducted meta-analyses to reveal the association of these estimates with river discharges. Furthermore, we simulated the mean length of eDNA downstream transport in a channel with given hydrogeographic conditions. The estimated S-w and V-dep values ranged from 2.6 to 13,187.4 m and 0.003 to 1.900 mm/s, respectively. When log-transformed, the median S-w and V-dep values were 10(2.188) (= 154.1) m and 10(-1.387) (= 0.041) mm/s. The calculated maximum eDNA transport distances until 95% and 99% of the target eDNA particles are deposited ranged from 334.0 to 1,272.7 m and 513.5 to 1,956.4 m, respectively. Moreover, we found that river discharges were positively associated with S-w, whereas V-dep was lower for natural environments than mesocosms. This trend could be due to other environmental factors, such as bottom substrates. Our meta-analysis implied that, under ordinary hydrogeographic conditions, most eDNA particles could travel less than 2 km downstream even when considering the resuspension of eDNA from riverbeds. By contrast, our study did not consider multiple environmental factors that potentially affect eDNA transport and would accordingly introduce a moderate to high degree of heterogeneity across studies. These issues will be addressed in future studies. To our knowledge, this study is the first to quantitatively synthesise the length of eDNA downstream transport in riverine environments and infer general properties of eDNA downstream transport distance and deposition velocity. According to our results, much of the eDNA released from individuals might have little effect on the detection of target eDNA in the sampling sites more than 2 km downstream. Although further accumulation of empirical studies is necessary, our findings propose the groundwork for optimal eDNA sampling designs in riverine ecosystems to reduce the false-positive inference of species presence and false-negative eDNA detection.

Automated summary

Environmental DNA (eDNA) analysis is an effective method to monitor species distribution and abundance in aquatic ecosystems. However, our understanding of eDNA downstream transport dynamics is limited. Based on investigations and simulations, most eDNA particles are transported within 2 km downstream.