Introduction
Organisms release their DNA molecules into their surroundings, which are termed as environmental DNA (eDNA) (Levy-Booth et al., 2007; Nielsen et al., 2007; Taberlet et al., 2012). The analysis of eDNA has recently been applied to monitor the abundance and composition of macro-organisms, such as fish and amphibians (Ficetola et al., 2008; Minamoto et al., 2012; Bohmann et al., 2014; Deiner et al., 2017; Jo et al., 2020a). Detection of eDNA in water samples does not involve any damage to the target species and their habitats, thus enabling non-invasive and cost-effective monitoring of species in aquatic environments, contrary to traditional monitoring methods such as capturing and observing (Darling & Mahon, 2011). However, the characteristics and dynamics of eDNA are not yet completely understood, and thus, the spatiotemporal scale of eDNA signals at a given sampling time and location is not certain, which can result in false-positive or false-negative detection of eDNA in natural environments (Darling & Mahon, 2011; Hansen et al., 2018; Beng & Corlett, 2020).
To determine the spatiotemporal scale of eDNA signals and accurately estimate species presence/absence and abundance in the environment, understanding the processes of eDNA persistence and degradation is important. Aqueous eDNA is detectable from days to weeks (Barnes & Turner, 2016; Collins et al., 2018), depending on various environmental factors. For example, moderately high temperature (Strickler et al., 2015; Eichmiler et al., 2016; Lance et al., 2017; Jo et al., 2020b) and low pH (Strickler et al., 2015; Lance et al, 2017; Seymour et al., 2018) accelerate eDNA degradation. In addition, eDNA decay rates are higher in environments with higher species biomass density (Bylemans et al., 2018; Jo et al., 2019a). These abiotic and biotic factors contribute to the increase in microbial activities and abundance in water, thus indirectly affecting eDNA degradation (Strickler et al., 2015). Moreover, eDNA decay rates were found to be different between the trophic states of studied lakes, and were negatively correlated with the dissolved organic carbon (DOC) concentrations (Eichmiller et al., 2016). This may be attributed to the binding of DNA molecules to humic substances, protecting eDNA from enzymatic degradation.
However, apart from the effects of such environmental conditions, little is known about the influence of the physiochemical and molecular states of eDNA on its persistence and degradation. Fish eDNA has been detected at various size fractions (<0.2 µm to >180 µm in diameter; Turner et al., 2014; Jo et al., 2019b) in water, suggesting that eDNA is present as various states and cellular structures, from larger-sized and intra-cellular DNA (e.g., cell and tissue fragments) to smaller-sized and extra-cellular DNA (e.g., organelles and dissolved DNA). Enzymatic and chemical degradation of DNA molecules in the environment depends on the presence of cellular membranes around the DNA molecules, and thus, the persistence of eDNA is likely to be linked to its state. In addition, eDNA persistence may be different depending on the target genetic regions. Recent studies have suggested that eDNA decay rates may vary between mitochondrial and nuclear DNA (Bylemans et al., 2018; Moushomi et al., 2019; Jo et al., 2020b). Moreover, studies comparing eDNA degradation between different target DNA fragment lengths (i.e. PCR amplification length) have yielded inconsistent conclusions; Jo et al. (2017) and Wei et al. (2018) reported higher eDNA decay rates for longer DNA fragments, whereas Bylemans et al. (2018) did not observe any difference in the eDNA decay rates of different DNA fragment sizes. Notably, Jo et al. (2020c) reported that selective collection of larger-sized eDNA using a larger pore size filter increased the ratio of long to short eDNA concentrations and altered the ratio of nuclear to mitochondrial eDNA concentrations; however, such reports linking eDNA state to its persistence are scarce.
Although our understanding of the relationship between eDNA state and persistence is currently limited, this relationship can be inferred by integrating previous findings of eDNA persistence and degradation. Here, we used meta-analyses to examine the relationship between eDNA states and persistence. We extracted data on filter pore size, DNA fragment size, target gene, and environmental parameters from previous studies estimating first-order eDNA decay rate constants, and investigated the influence of these factors on eDNA degradation. By assembling and integrating the results of previous eDNA studies, our meta-analyses revealed the hitherto unknown relationships between eDNA state and persistence, which could not have been observed in the individual studies. Furthermore, we assessed the validity of the findings of the meta-analyses by re-analysing the dataset from a previous tank experiment (Jo et al., 2019b).