Variability in Fish Environmental DNA Concentration in Coastal Ecosystems at Hierarchical Levels: Focusing on the...

Understanding variability in environmental DNA (eDNA) concentration is essential for improving the precision of quantitative eDNA analyses. While previous laboratory and field studies suggest that technical variation arising from sampling and PCR processes is relatively small, the hierarchical structure of this variability and its environmental dependence remain poorly understood. In this study, we conducted spatiotemporally replicated seawater sampling from a coastal ecosystem, quantified Japanese jack mackerel (Trachurus japonicus) eDNA concentrations using quantitative real-time PCR (qPCR), and assessed the magnitude, structure, and environmental drivers of eDNA variability across multiple levels. Variance component analysis revealed that more than 90% of the total variance in eDNA concentration was explained by differences among sampling sites and time points, reflecting differences in organismal distribution and dynamics, whereas sampling and PCR steps together contributed less than 10%. Using Taylor's Power Law, we demonstrated that the relationship between mean eDNA concentration and variance differed across hierarchical levels, with stronger mean-variance scaling at larger spatial and temporal scales. Notably, the relative contribution of the PCR level to the total variance increased substantially under low-concentration conditions, indicating that stochastic measurement error becomes dominant when eDNA is scarce. We also demonstrated that the PCR-level variability (Coefficient of variance; CV) increased with chlorophyll-α and decreased with pH, whereas no significant environmental effects were detected at the sample level. These results suggest that different mechanisms govern variability at each hierarchical level, including ecological processes, physical heterogeneity, and biochemical constraints. Our findings highlight that optimal sampling and replication strategies should be tailored to expected eDNA concentrations and environmental conditions in the field, helping provide a framework for maximizing signal-to-noise ratios in quantitative eDNA studies.