1We propose a general framework of abundance estimation based on spatially replicated quan-2 titative measurements of environmental DNA in which production, transport, and degradation of 3 DNA are explicitly accounted for. Application to a Japanese jack mackerel (Trachurus japonicus) 4 population in Maizuru Bay revealed that the method gives an estimate of population abundance 5 comparable to that of a quantitative echo sounder method. These findings indicate the ability of 6 environmental DNA to reliably reflect population abundance of aquatic macroorganisms and may 7 offer a new avenue for population monitoring based on the fast, cost-effective, and non-invasive 8 sampling of genetic information. 9 10 Knowledge on the distribution and abundance of species is crucial for ecology and related 11 applied fields such as wildlife management and fisheries. The detection and quantification of 12 environmental DNA (eDNA) is an emerging methodology for ecological studies and could enhance 13 the ability of investigators to infer occurrence and abundance of species. This approach has been 14 applied, especially but not limited to, to aquatic species such as fish and amphibians and has been 15 identified as a powerful and yet cost-effective tool for species detection (Bohmann et al. 2014, Rees 16 et al. 2014, Thomsen & Willerslev 2015, Goldberg et al. 2016, Deiner et al. 2017 2018). Challenges remain, however, in quantitative applications of eDNA. Since earlier studies 18 revealed positive correlations between species abundance and eDNA concentration (Takahara et al. 19 2012, Thomsen et al. 2012, Goldberg et al. 2013, Pilliod et al. 2013, Eichmiller et al. 2014, it has 20 been expected that local population abundance may be inferred by measuring the concentration of 21 eDNA at a given locality. Indeed, an analytical framework proposed recently for eDNA-based 22 between eDNA concentration and the underlying population size (Chambert et al. 2018).
24Nonetheless, such a definite relation may not always be present, possibly depending on e.g. the 25 shedding rate, transport, and exogenous input of eDNA (Pilliod et al. 2013, Eichmiller et al. 2014 Lacoursière- Roussel et al. 2016, Yamamoto et al. 2016, Jo et al. 2017. 27 The fundamental factors that underlie such context dependency are the 'ecology of eDNA': 28 the distribution of eDNA in space and time stems from processes governing the origin, state, 29 transport, and fate of eDNA particles (Barnes & Turner 2016). Thus, in applications of the eDNA 30 methodology, detailed information about such processes may be critical. Without relevant knowledge 31 of these processes, for example, the spatial and temporal scales of information provided by eDNA 32 remain largely uncertain (Thomsen & Willerslev 2015, Goldberg et al. 2016, Hansen et al. 2018.
33Therefore, here, our purpose was to develop a general approach to eDNA-based abundance 34 estimation that can fully account for the ecology of eDNA, i.e. the rate of production and 35 degradation of eDNA as well as the transpor...