Chen Xu, Kei Nukazawa
Water-based environmental DNA (eDNA) can simultaneously detect aquatic and terrestrial species during watershed-scale eDNA biodiversity monitoring. However, detecting terrestrial species remains a challenge because of their complex environmental effects. This study targeted vertebrate eDNA transported from land to rivers through surface runoff during rainfalls in three rivers within the Kiyotake River watershed, Japan. We compared the relative detection indices of different taxonomic groups under rainfall and non-rainfall conditions and determined how environmental factors (e.g., discharge and filtration volume) influence sample-level alpha and beta diversity. The results showed that mammals and amphibians were detected more frequently after rainfall, whereas fish showed lower detection frequencies due to reduced filtration volumes caused by filter clogging. Generalized linear mixed model analyses revealed that river discharge had significant negative effects on the detection number, whereas filtration volume had a significant positive effect. Furthermore, Permutational Multivariate ANOVA indicated that filtration volume and rainfall occurrence significantly affected beta diversity. Our findings highlight the importance of filtration volume for detecting alpha and beta diversity and demonstrate that eDNA-based biomonitoring remains effective during rainfall events, offering practical guidance for enhancing biodiversity assessments at the watershed scale. Such approaches may provide an efficient tool for detecting not only aquatic species, but also terrestrial vertebrates that are otherwise difficult to survey using conventional field methods.