Impact of Rainfall on Aquatic and Terrestrial Species Monitoring Using Riverine Environmental DNA

Impact of Rainfall on Aquatic and Terrestrial Species Monitoring Using Riverine Environmental DNA | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Impact of Rainfall on Aquatic and Terrestrial Species Monitoring Using Riverine Environmental DNA Xu Chen, Nukazawa Kei This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7806039/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Water-based eDNA has a potential to simultaneously detect aquatic and terrestrial species in watershed-scale eDNA biodiversity monitoring. However, the detection of terrestrial species remains difficult or undetected due to complex environmental effects. This study targeted vertebrate eDNA transported from land to rivers through surface runoff during rainfall in three rivers within the Kiyotake River watershed in Japan. We compared the detection rates of different taxonomic groups under rainfall and non-rainfall conditions and determined how environmental factors (e.g., discharge, rainfall amount) influence sample-level alpha and beta diversity. The vertebrate detection rate after rainfall exceeded that during non-rainfall. Birds, mammals, and amphibians were detected more frequently after rainfall, whereas fish were detected more frequently under non-rainfall conditions. Generalized linear model analyses revealed that river discharge and pH had significant negative effects on the detection number, while filtration volume had a significant positive effect. Furthermore, Permutational Multivariate Analysis of Variance indicated that filtration volume and rainfall occurrence significantly affected beta diversity. Our findings underscore the importance of sampling design after rainfall events, for eDNA-based biomonitoring, offering practical guidance for enhancing biodiversity assessments at the watershed scale. Biological sciences/Ecology Earth and environmental sciences/Ecology Earth and environmental sciences/Environmental sciences Environmental DNA metabarcoding Vertebrate Rainfall General Liner Model Permutational Multivariate Analysis of Variance Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Global climate change and human activity have accelerated biodiversity loss, leading to the significant degradation of ecosystem services 1 . Environmental DNA (eDNA) enables the development and implementation of effective monitoring strategies for timely, evidence-based conservation. eDNA, comprising genetic material (e.g., skin and feces) shed by organisms and retained in environmental substrates (e.g., water, soil, and air) 2–4 . It has many advantages, including low cost, rapidity, and non-invasiveness 5–7 . In addition, eDNA metabarcoding, which combines next-generation sequencing and universal primers, facilitates large-scale biodiversity surveying 8–10 . Vertebrate eDNA research has mainly focused on aquatic species 11–14 . For terrestrial species, various sampling strategies and substrates (e.g., air, soil, mineral licks, feeding traces, and hair) have been employed. For instance, feces have been used to detect carnivores 15 , while DNA extracted from wildflowers has been used to identify arthropods and birds 16 . Tailoring eDNA sampling protocols to specific taxonomic groups or ecological contexts may improve the comprehensiveness and reliability of biodiversity assessments. However, identifying suitable environmental substrates for terrestrial vertebrate eDNA is time-consuming and resource-intensive 17,18 . These limitations may restrict the range of species detected 19 . Moreover, considerable uncertainty remains regarding the ecology of terrestrially derived eDNA, including its origin, physicochemical properties, transport mechanisms, and environmental fate 20,21 . Terrestrial species have been reported to release DNA into water via activities such as drinking and swimming 22,23 . This provides a theoretical basis for water-based eDNA approaches for the simultaneous and rapid detection of terrestrial and aquatic species. Building on this, the approach has been applied across various ecosystems, such as forests and urban areas 24–26 , and validated in watersheds of different spatial scales, ranging from small streams to the Amazon River 27,28 . However, in water-based eDNA biodiversity monitoring within watersheds, the detection rates of terrestrial species are generally lower than those of aquatic species 29,30 . This discrepancy largely arises from the continuous DNA shedding by aquatic organisms (e.g., fish) into their environments, which differs from the spatially and temporally intermittent interactions of terrestrial species with water bodies, resulting in limited eDNA deposition 31 . This indicates that target species may not be detected, even if they are present near sampling locations 32 . Furthermore, the detectability of eDNA in aquatic systems is influenced by environmental factors such as pH, water temperature, and flow distance 33–36 . This highlights the need to use water-based approaches to elucidate the effect of environmental factors on the detection of aquatic and terrestrial species. Detection in terrestrial substrates and in aquatic environments has inherent limitations 37 . Currently, water is the most commonly used eDNA substrate for assessing biodiversity 19 . This may be because of the unique advantage of water in connecting terrestrial and aquatic habitats as well as upstream and downstream areas. Such connectivity facilitates the use of water-based eDNA approaches to evaluate species diversity across broader spatial scales 4,38 . In our previous study on a specific terrestrial mammal, we confirmed that eDNA could be transported into rivers via surface runoff during rainfall events 39 . This will help improve the detection efficiency for large-scale assessments of terrestrial mammal diversity. In addition to terrestrial mammals, rainwater collected after precipitation events in forested ecosystems has been shown to contain a high diversity of invertebrate and microbial taxa 40,41 . In riverine ecosystems, post-rainfall river water samples were found to have increased microbial diversity 42 . However, studies on vertebrates in such settings are limited. A previous study found that, compared to eDNA samples collected during non-rainfall periods, the number of fish species detected in river water decreased after rainfall, whereas the number of mammal species detected increased 43 . This indicated that rainfall had distinct effects on the detection of vertebrate species with different ecological habits. Moreover, different rainfall patterns (e.g., intensity and duration) may affect species detection. For instance, heavy-rainfall periods are less suitable for fish detection, compared to light rainfall. Meanwhile, compared to peak rainfall, rainfall duration is more conducive to the detection of terrestrial mammals 39,44,45 . In conclusion, it is necessary to further investigate the effects of rainfall patterns on the detection of vertebrates, including fish and mammals. This study developed a method for detecting aquatic and terrestrial biological communities in riverine stormwater collected after rainfall events in the Kiyotake River catchment, Miyazaki Prefecture, Japan. Furthermore, we evaluated the effects of different rainfall patterns on species richness and community composition across sub-catchments with contrasting land use types, providing new insights into how hydrological and landscape factors influence eDNA-based biodiversity assessments. Results eDNA metabarcoding A total of 1,952,639 sequences were obtained from the 33 eDNA samples. One sample collected from the Oka River in June 2023 showed no detection. Among the sequences, 93.15% had a quality score >30. After removing the chimeric and noisy sequences, 812,677 reads remained. Following the removal of non-vertebrate species, humans, and environmental contaminants, 670,036 reads remained and were classified as fish, birds, amphibians, or mammals. Fish accounted for the largest proportion (65.9%), whereas birds represented the smallest proportion (4.02%). In total, 64 taxa were detected, with 26 taxa identified at the species level, representing 43.1% of the total. Figure 1 shows the detection results for each river (for the complete dataset, see Supplementary 1). In the Mizunashi River, 25 taxa were detected, comprising six fish species, seven amphibians, six birds, and six mammals. In the Oka River, 30 taxa, including 14 fish, seven amphibians, three birds, and six mammals, were identified. The Tagami River had the highest taxonomic richness, with 38 taxa, specifically 19 fish, 10 amphibians, five birds, and four mammals. The Oka River and Tagami River showed a higher dominance of fish species. Most of the detected mammalian taxa were domesticated or human-associated species (e.g., cattle, pigs, and dogs), particularly in the Tagami River, where no wild mammals were detected. The wild mammals identified in this study included Meles anakuma , Mogera wogura , Urotrichus talpoides , Lepus spp. (wild rabbits), Canidae spp. (domestic or raccoon dogs), and Sus scrofa (wild boars or domestic pigs). These were all small-bodied terrestrial mammals detected only after rainfall events. Amphibian taxa were also predominantly detected after rainfall events. All belonged to the order Anura, with the exception of Paramesotriton sp., a member of the order Caudata. Detection rate Figure 2 shows the overall detection probability for each sample. For the Mizunashi River, the detection rates ranged from 5.26% to 57.89%, with the lowest rate observed in June 2023. The two sampling events without rainfall yielded detection rates of 21.05% and 36.84%. In the Oka River, the detection rates ranged from 0% to 48.15%, with the lowest recorded in June 2023. The two non-rainfall sampling events showed detection rates of 48.15% and 33.33%. For the Tagami River, the detection rates varied between 15.15% and 54.55%, with the lowest rate occurring in November 2022. The two non-rainfall sampling events had detection rates of 21.05% and 36.36%. Regarding different taxonomic groups, the detection rates in the Mizunashi River ranged from 0% to 100% for fish, 0% to 57.14% for amphibians, and 0% to 50% for mammals. In the Oka River, the detection rates ranged from 0% to 85.71% for fish, 0% to 57.14% for amphibians, and 0% to 66.67% for mammals. Meanwhile, for the Tagami River, the detection rates were 10.53–63.16% for fish, 0–60% for amphibians, and 0–100% for mammals. Relationship between detection number and environmental factors Table 2 presents the results of the GLMs. The GLMs revealed that filtration volume had a significant positive effect on the number of detections, whereas discharge and pH had significant negative effects on the detection number. Community structure Figure 3 shows the βsor, βsim, and βsne results, calculated based on Sørensen dissimilarity, for the three sampling sites. The average of βsor, βsim, and βsne values for each river were as follows: the Mizunashi River: βsor = 0.56, βsim = 0.35, and βsne = 0.21; the Oka River (without no detection sample): βsor = 0.56, βsim = 0.42, and βsne = 0.14; the Tagami River: βsor = 0.55, βsim = 0.39, and βsne = 0.16. Figure 4 shows the result of PCoA. PCoA based on the Sørensen index showed that the first two axes explained 40.78% of the variation in community composition in the Mizunashi River. The two groups of samples did not form distinct clusters in the PCoA plot. Envfit analysis indicated that river discharge (r² = 0.60, p = 0.002), filtration volume (r² = 0.52, p = 0.019), and rainfall (r² = 0.60, p = 0.020) were significant drivers of community structure. In the Oka River, the first two PCoA axes explained 43.74% of the variation, and similarly, no distinct clustering was observed between the two groups. Envfit analysis also revealed that filtration volume (r² = 0.47, p = 0.029) and rainfall (r² = 0.81, p = 0.017) were significant drivers of community structure. In contrast to the other two rivers, the Tagami River showed a different pattern. The first two PCoA axes explained 42.04% of the variation, and the two groups of samples formed a distinct separation in the PCoA plot. Envfit analysis identified discharge (r² = 0.55, p = 0.029), filtration volume (r² = 0.88, p = 0.030), and rainfall (r² = 0.51, p = 0.041) as significant drivers of community structure. Table 3 presents the results of the PERMANOVA analysis, which supports the findings obtained from PCoA. In the Mizunashi River, filtration volume had a significant effect on βsor (R² = 0.14, p = 0.047). In the Oka River, filtration volume significantly affected βsor (R² = 0.15, p = 0.038). In the Tagami River, filtration volume had a significant influence on both βsor (R² = 0.22, p =0.004) and βsne (R² = 0.27, p = 0.002). Contrastingly, the effects of rainfall were relatively weak, with a significant impact observed only for βsor in the Tagami River (R² = 0.15, p = 0.048). When the filtration volume was standardized, rainfall still had a significant effect (R² = 0.23, p = 0.041). Discussion Using eDNA metabarcoding, this study investigated the effects of rainfall on the detection of different vertebrate taxonomic groups, by collecting river water samples after rainfall events with varying patterns. A total of 64 taxa, including fishes, mammals, birds, and amphibians, were detected in the 33 samples collected from the three rivers. The mammals detected were primarily livestock and small wild species, while the amphibians were mainly frogs. For wild boars ( S. scrofa ), the 12S-V primer used in this study could not distinguish between wild and domestic pigs 46 . Although both types could exist in the study area, this study focused on the overall ecological patterns; therefore, the related sequences were conservatively assigned to the species level ( S. scrofa ). Future studies should consider using specific primers to differentiate between wild boar and domestic pigs. In the Tagami River, one Amplicon Sequence Variant (ASV) was identified as Rana zhenhaiensis based on our analytical pipeline. However, the likelihood of this species occurring locally is extremely low, as it is distributed only in China, and the sequence was more likely derived from Rana japonica . Nevertheless, we retained this result because of its potential relevance to invasive species monitoring. We included two replicate samples for each individual sampling event; however, the number and composition of the detected taxa varied between replicates. Regarding the number of detections, discrepancies among replicates with differing filtration volumes ranged from 3 to 7 taxa in the Mizunashi River, 1 to 3 taxa in the Oka River, and 0 to 1 taxon in the Tagami River. These volume-related inconsistencies were most pronounced in the Mizunashi River, which is characterized by a low species richness and a relatively large catchment area. Larger filtration volumes tended to yield a greater number of detected taxa. In contrast, when filtration volumes were consistent across replicates, the number of taxa detected differed by at most one per site. Although rainfall may disrupt the spatial homogeneity of the eDNA distribution in river water 47 , increasing the filtration volume can enhance the comprehensiveness of biodiversity assessments 48,49 . Two sampling events were conducted on specific dates. Although rainfall persisted, turbidity generally decreased during the second sampling period owing to a reduced rainfall intensity, allowing for higher filtration volumes. The number of fish taxa decreased during the second sampling event in the larger catchments of the Mizunashi River and the Oka River, whereas it increased in the Tagami River. This may be attributed to stronger dilution caused by rainfall in larger catchments 50 . Additionally, under the conditions of prolonged heavy rainfall, an increased detection of fish eDNA may be related to the resuspension of eDNA from riverbed sediments 51 . Most of the taxa newly detected during the second sampling period in the Tagami River belonged to the Gobiidae family, which is primarily benthic. Previous studies have suggested that sediments may be more effective than water for detecting eDNA from benthic fish species 52 . However, because of the limited taxonomic resolution of the 12S rRNA primers used in this study, it remains unclear whether the newly detected taxa were benthic, midwater, or surface-dwelling species. Moreover, we did not apply metabarcoding to riverbed sediments, and this conclusion requires further validation. In May 2022, continuous rainfall led to an increase in the number of amphibian and mammal taxa detected during the second sampling period in the Mizunashi River, whereas the opposite trend was observed in the Oka River. In the Mizunashi River, most of the mammals detected were livestock, and eDNA sources were abundant near the sampling sites. Continuous rainfall likely transported their terrestrial eDNA into the river, resulting in increased mammal detection during the second sampling period. By contrast, the first sampling period in the Oka River included wild animals, which have limited eDNA sources. The dilution effect of continuous rainfall outweighed the replenishment of eDNA from surface runoff, leading to a decrease in detection during the second sampling period. Detection rate The lowest detection rate was observed after the rainfall period, which was attributed to the reduced detection rate for fish taxa after rainfall. A previous study demonstrated that rainfall may dilute the eDNA of aquatic species, leading to reduced detection 53 . In addition, increased turbidity resulting from rainfall hindered the filtration of adequate water volumes (<400 mL) and would promote an inhibition of PCR. Thus, such effects after rainfall obscured fish eDNA, and the overall detection rates across broad taxonomic groups decreased. Nevertheless, amphibians and mammals exhibited higher detection rates after rainfall events. Although rainfall dilutes the eDNA of terrestrial species in water, the transportation of terrestrial eDNA through surface runoff helps replenish the diluted portions 44 . In addition, considering the behavioral characteristics of amphibians, the increase in detection rates may not be entirely attributed to surface runoff. Amphibians tend to be more active after rainfall, thereby releasing more eDNA 54 , explaining why they are almost exclusively detected after rainfall events. A comparison of the detection rate for birds, performed using samples collected after the same period (non-rainfall: November 28, 2022; rainfall: November 29), revealed that, despite the lower filtration volume after a rainfall event (350–690 mL), the number of bird detections was still higher than that during the non-rainfall period (1,000 mL). This is particularly evident because only one species of Anatidae, a group of waterbirds, was detected on non-rainy days, whereas both waterbirds ( Anatidae.sp, Galkinula chloropus ) and non-waterbirds ( Passeriformes.sp, Gallus gallus) were detected on rainy days. It is possible that bird DNA present in the air was introduced into the river via rainfall 55 and that bird droppings were transported to the river via runoff after rainfall 56 . In this short-term comparison, bird species also showed higher detection rates on rainfall days than on non-rainfall days, which is consistent with the detection trends observed for mammals and amphibians. This further suggests that rainfall could enhance eDNA inputs and detection for certain terrestrial taxa. Thus, designing sampling strategies that consider species-specific life histories is necessary. For example, fish were more suitably surveyed during non-rainfall periods, whereas amphibians were better targeted in post-rainfall surveys. While no sampling method can achieve a 100% detection rate for all organisms 57 , our results demonstrated that vertebrate surveys based on water eDNA after rainfall can improve the detection of mammals, birds, and amphibians. Relationship between detection number and environmental factors River discharge and pH had significant negative effects on detection number. This was likely because increased discharge during rainfall dilutes eDNA concentrations 53 , thereby reducing the detection number. Although many studies have reported that acidic conditions accelerate eDNA degradation 35,36 , our results showed a negative correlation between pH and detection number. The reason for this pattern is unclear; however, a low pH may enhance DNA adsorption onto suspended particles, increasing its recoverability 33 . Given that our study was conducted in turbid waters, this may be an explanation; however, further research is required to determine whether adsorption can offset the effects of acidity on DNA degradation. Contrastingly, filtration volume had a significant positive effect on detection number, as filtering larger volumes of water captures more eDNA 58 , leading to more detection counts. Rainfall did not significantly affect the number of detections. This was because we could not determine when rainfall began to influence detection, and we, therefore, used the total rainfall within 3 h prior to sampling as the environmental variable, although the actual effective time may not be 3 h. Community structure Among the samples collected during the same period, the highest Sørensen dissimilarity indices were consistently observed in the Mizunashi River and Oka River in June 2023, and in the Tagami River in November 2022. These high dissimilarities may not have been directly caused by rainfall, but rather by filter clogging induced by rainfall, which resulted in the lowest filtration volumes for these sampling events. Compared to samples with larger filtration volumes, those with lower filtration volumes exhibited greater variability in terms of species composition 59 , and the Sørensen dissimilarity index could be affected by low DNA concentrations or the non-detection of certain taxa in some samples 60 . Under identical filtration volume conditions, βsor in the Mizunashi River and the Oka River were nearly identical between rainfall and non-rainfall periods. In the Mizunashi River, when replicate samples exhibited differences in filtration volume, βsor was primarily driven by nestedness, which was likely caused by the decrease in recovery due to a reduction in filtration volume. In contrast, in the Tagami River, the βsor during rainfall events was lower than that under non-rainfall conditions and was mainly driven by species turnover. The results of the PCoA and the PERMANOVA analyses indicated that, in the Mizunashi River and the Oka River, rainfall did not significantly affect βsor, whereas filtration volume had a significant effect. Although filtration volume is an important factor influencing βsor 61 , excessively large differences in filtration volume in these rivers may obscure the effect of rainfall. This is supported by the results obtained from the Tagami River, where the filtration volume was relatively consistent, indicating that both filtration volume and rainfall had significant effects. These findings suggest that, within the same river system, the influence of filtration volume may outweigh that of rainfall, highlighting the importance of filtration volume in the sampling design for species detection. Furthermore, when filtration volumes were equivalent, evaluating only the effect of rainfall revealed that the species composition tended to be more homogeneous 44 . This could be attributed to the lower water temperatures and reduced ultraviolet radiation following rainfall events, which favor the preservation of eDNA in water columns 36 . Owing to the limited number of samples, further studies with larger sample sizes and standardized filtration volumes are required to verify the statistical significance and generality of this trend. However, in studies involving turbid water, the rapid filtering of large volumes remains a challenge 62 . Comparisons across different watersheds further highlighted the importance of rainfall and filtration volume. During the first sampling event in May 2022, the βsor values were similar between the Mizunashi River and the Oka River, despite considerable variation in the filtration volumes among the Oka River samples. In September 2022, when a complete 1-L volume was filtered in both the Oka River and the Tagami River, the βsor values were almost identical. In November 2022, the βsor in the Tagami River was much higher than that in the Mizunashi River owing to insufficient filtration volume. Conclusion This study successfully achieved the simultaneous detection of aquatic and terrestrial species by conducting surveys after rainfall events and evaluated the differences in detection patterns in relation to rainfall and watershed characteristics. The results suggested that rainfall influenced the detectability of different taxonomic groups by altering multiple processes such as eDNA input, dilution, and resuspension. Although rainfall events could reduce the overall detection efficiency, particularly for fish, they could also enhance the detectability of terrestrial and semi-aquatic species. Detection outcomes were further shaped by factors such as filtration volume and species-specific ecological traits. Therefore, optimizing the sampling design based on the ecological characteristics of the target species is crucial for improving the detection efficiency in species-specific eDNA studies. In addition, our findings highlight the importance of implementing multi-timepoint sampling strategies across varying hydrological conditions and climatic events to achieve a more comprehensive assessment of watershed biodiversity. Materials and Methods Study area The study area is located in Miyazaki Prefecture, southwestern Japan, and belongs to the tributaries of Kiyotake River (i.e., the Mizunashi River, Oka River, and Tagami River; Figure 5- a ). The Kiyotake River is 28.8 km long, and its catchment covers an area of 166.40 km 2 . The catchment areas of the Mizunashi River, Oka River, and Tagami River are 16.70 km², 16.90 km², and 3.30 km², respectively (catchment area data available: https://data.stat.pref.miyazaki.lg.jp/dataset/nenkan-26). The climate in the catchment is characterized by an annual average temperature of approximately 18.00°C and annual precipitation of approximately 3,000 mm. We obtained data for 12 different land use types through the Advanced Land Observing Satellite (10-m accuracy, data available: https://earth.jaxa.jp/ja/data). Fig. 5- b shows the land uses in the three catchments. The major land uses in the studied rivers were evergreen needleleaf trees, in the Mizunashi River (51.80%); evergreen needleleaf trees and fields, in the Oka River (28.20% and 30.20%, respectively); and fields, in the Tagami River (39.50%). The geographic coordinates of the Mizunashi River, Oka River, and Tagami River are 31.829925°N, 131.362669°E; 31.85162°N, 131.386959°E; and 31.838143°N, 131.417732°E, respectively. Sampling and eDNA extraction All water samples were collected from publicly accessible rivers, and no permits were required for sampling. The sampling events included rainfall and non-rainfall events. Sampling for the rainfall events were conducted on May 12, September 14, and November 29, 2022, as well as on June 2, 2023. To understand the temporal pattern of eDNA detection, sampling was conducted twice daily (sampling 1 and 2) on May 12, 2022, and June 2, 2023. The time interval between sampling events was approximately 2 h. Sampling for the non-rainfall events occurred on November 28, 2022, and August 8, 2024. Before water sampling, 1-L and 0.25-L plastic bottles, tethered buckets, and cooler boxes were sterilized using 10% bleach for at least 30 min to prevent contamination by DNA and microorganisms other than the target species. During rainfall events, two 1-L replicates of river water were collected, whereas during non-rainfall events, only a single 1-L sample was collected. Another 0.25 L of river water was collected for basic measurements of environmental parameters, namely, pH, electrical conductivity (EC), and turbidity, during all events. For the 0.25-L samples, a bar-shaped mercury thermometer was used to measure water temperature at the sampling sites, EC and pH were measured using a portable multiwater quality meter (HORIBA TOA DKK), and turbidity was measured using a turbidity meter (PT-200_TROAM, Asahi Science) in a laboratory. The environmental parameters and EC, in particular, were not measured on November 28, 2022, and June 2, 2023, respectively, owing to technical problems. Rainfall data were obtained from the nearest meteorological station (Akae Observatory, data available: https://www.data.jma.go.jp/obd/stats/etrn/index). Detailed methods for water filtration and eDNA extraction are available in our previous studies 39 . In summary, water samples were filtered using 0.7-µm glass fiber filters. When the filter paper became clogged, it was replaced with a new one. If the second filter also became clogged, the remaining water was discarded. After filtration, all filter papers were frozen at −20 °C until DNA extraction. eDNA was extracted using the DNeasy PowerSoil Kit (Qiagen, Hilden, Germany), following the manufacturer’s protocol, after which it was eluted in 100-µL volumes. After extraction, the DNA template solutions were stored at −80 °C for subsequent polymerase chain reaction (PCR) amplification. Samples collected on August 8, 2024, were extracted using the DNeasy PowerSoil Pro Kit, the successor of the DNeasy PowerSoil Kit, since the latter had been discontinued. eDNA metabarcoding Using the DNeasy PowerClean Pro Cleanup Kit (Qiagen, Hilden, Germany), the template was purified to remove the PCR inhibitors. The concentration of the DNA solution was determined using the Synergy LX (Agilent Technologies) in combination with the QuantiFluor dsDNA System (Promega). Subsequently, vertebrate DNA was comprehensively detected via high-throughput sequencing using a two-step tailed PCR approach targeting the mitochondrial 12S rRNA gene region. The forward primer YAGAACAGGCTCCTCTAG and reverse primer TTAGATACCCCACTATGC were used, both of which are suitable for vertebrate detection 63 . The 12S rRNA region was selected owing to its low species-level primer bias, high specificity, and reproducibility 64 . Moreover, previous studies have reported that replacing the initial thymine (T) in the forward primer with pyrimidine (Y) enables broader binding across vertebrate taxa (Kocher et al., 2017). The first-round PCR (1 st PCR) was performed in eight replicates, with each 10-μL PCR mixture containing 1.0 μL of 10× Ex Buffer, 0.8 μL of deoxynucleotide triphosphates (dNTPs, 2.5 mM each), 0.5 μL each of 5-μM forward (5′-ACACTCTTTCCCTACACGACGCTCTTCCGATCT YAGAACAGGCTCCTCTAG-3′) and reverse (5′-GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT TTAGATACCCCACTATGC-3′) primer, 2.0 μL of template DNA, 0.5 μL each of 10-μM forward and reverse primer, 0.075 μL of ExTaq HS (5 U/μL; TaKaRa), and 5.125 μL of nuclease-free water. The thermal cycling conditions were as follows: initial denaturation at 94°C for 2 min; 30 or 35 cycles at 94°C for 30 s, 50°C for 30 s, and 72°C for 30 s; and a final extension at 72°C for 5 min. The PCR products were purified using VAHTS DNA Clean Beads (Vazyme) at a 1.5× ratio to the reaction volume. The purified products were used as templates for the second round of PCR. A negative control was prepared using nuclease-free water as the template. The 2 nd PCR process was performed as a single reaction to attach adapter sequences. Each 10-μL reaction mixture included 5.0 μL of 2× PCR Buffer for KOD FX Neo, 2.0 μL of dNTPs (2 mM each), 0.5 μL each of 5-μM forward (5′-AATGATACGGCGACCACCGAGATCTACACXXXXXXXXACTCTTTCCCTACACGACGC-3′) and reverse (5′-CAAGCAGAAGACGGCATACGAGATXXXXXXXXGTGACTGGAGTTCAGACGTGTG-3′) primer, 1.0 μL of pooled 1 st -PCR products, 0.2 μL of KOD FX Neo (1.0 U/μL; TOYOBO), and 0.8 μL of nuclease-free water. The PCR program was as follows: 94°C for 2 min; 10 or 12 cycles at 98°C for 30 s, 60°C for 30 s, and 68°C for 30 s; and 68°C for 5 min. The 2 nd PCR products were also purified using VAHTS DNA Clean Beads (1.5× volume ratio), completing library preparation. The concentration of the final libraries was measured using Synergy H1 (Agilent Technologies) with the QuantiFluor dsDNA System, and the library quality was assessed using a Fragment Analyzer with the dsDNA 915 Reagent Kit (Agilent Technologies). Finally, sequencing was conducted using the MiSeq system with the MiSeq Reagent Kit v3 (Illumina), under 2×300-bp paired-end conditions. Reads obtained using the fastx barcode splitter tool of the FASTX-Toolkit (ver. 0.0.14) were filtered to retain only those that perfectly matched the primer sequences used at the start of the reads. Using the fastx_trimmer tool of FASTX-Toolkit, the primer sequences were removed from the extracted reads. Subsequently, low-quality sequences with quality scores below 20 were discarded using Sickle (ver. 1.33), and sequences shorter than 40 bases, along with their paired reads, were removed. Paired-end reads were merged using the FLASH script (ver. 1.2.11), with a minimum overlap of 10 bases. For species identification, the dada2 plugin in Qiime2 (ver. 2024.2) was used to remove chimeric and noisy sequences. Representative sequences and an ASV table were then generated, and the representative sequences were subjected to a Basic Local Alignment Search Tool (BLAST) process against the nucleotide database (National Center for Biotechnology Information) to identify species. The identification criterion was a coverage >90%, and only the result with the highest identity was used. The identity thresholds for species-level identification were set as follows: >98% for species, >95% for genera, >90% for families, and >85% for orders 65 . When the identity index was the same, the optimal result was selected as the reference index, based on the E-value and coverage, in that order. In cases where there was incongruence between BLAST hits, the ASV(s) would be assigned to a higher rank (i.e., genus or class) to provide a consensus assignment 66 . We applied a cut-off based on a fixed-read-counts threshold; when an ASV was present in a sample with the number of reads <100, it was removed from the sample 67,68 . Since the samples were collected from freshwater rivers, based on the fish database (https://www.fishbase.se), fish species that exclusively inhabited marine environments were excluded from the analysis with human ASVs as environmental contaminants. One ASV was assigned to Homo sapiens although it may have originated from Mus musculus . Although Mus is often regarded as a laboratory contaminant 69 , its widespread presence in natural environments warranted the retention of this ASV in the analysis. Rainfall and river discharge Rainfall data for the sampling day were obtained from the Akae Meteorological Station. Given the relatively small sizes of the three watersheds, we used the cumulative rainfall during the 180 min immediately preceding the start of sampling as an approximate measure of rainfall. River discharge after rainfall was estimated using a distributed hydrological model 70 , which divided the watershed into grid cells and incorporated watershed characteristics to accurately reproduce the discharge from upstream to downstream. For sampling sites where discharge could not be directly calculated, discharge data from the nearest grid cells were used. In the case of the Tagami River, a direct estimation of discharge was not possible; therefore, discharge was calculated for the Kiyotake River and subsequently converted using the watershed area ratio between the Tagami and Kiyotake basins. These parameters, together with the water quality variables, are listed in Table 1. Data analysis Detection rate To evaluate the impact of rainfall occurrence on species detection, we calculated the overall detection rate for each sample, as well as the detection rates for individual taxonomic groups. Because a comprehensive species inventory for the study area is currently lacking, the total number of species detected across all samples was used as a proxy for the regional species pool. Based on this, the detection rate of a single sample within a sub-watershed was defined as the proportion of species detected in that sample relative to the total number of species detected within the same sub-watershed. Meanwhile, the detection rate for each taxonomic group was defined as the proportion of species detected within that group in the sample relative to the total number of species detected in the corresponding group across the sub-watershed. This study spanned a full year and encompassed all four seasons, making it possible that seasonal variations may confound the effects of rainfall on species detection. Birds exhibit strong seasonal migratory behavior, and their detection frequency may be more influenced by seasonal patterns than by rainfall. Moreover, because of the limited taxonomic resolution of the primers used in this study, most bird sequences could not be reliably identified to the species level, which made it difficult to determine whether the detected taxa were migratory. To minimize the influence of seasonal factors and enhance the robustness of the analysis, birds were excluded from the calculations of the detection rate and community structure. Although fish may migrate for spawning, they remain in freshwater rivers during certain life stages; therefore, seasonal variations did not preclude their detection. Relationship between detection number and environmental factors Generalized linear models (GLMs) with a Poisson distribution were used to assess the relationships between the number of species detected in the samples (detection numbers) and environmental factors, including turbidity, pH, water temperature, filtration volume, and discharge. Because EC was not recorded in June 2023, it was excluded from the model to maintain a sufficient sample size. The cumulative rainfall during the 3 h prior to sampling was used as an explanatory variable. Variance inflation factors (VIFs) were calculated using the R “car” package to verify multicollinearity. Variables with VIFs >5 were iteratively removed until all variables had VIF values less than 5. Turbidity was excluded during this process, and final models were constructed using the R “glmmTMB” package 71 . All analyses were conducted using R (ver. 4.3.2, R Core Team, 2023). Community structure analyses To evaluate the effect of rainfall conditions on community composition, we calculated β-diversity based on species presence/absence data, using the Sørensen dissimilarity index 72 . Specifically, used the betapart package to compute total dissimilarity (βsor), turnover (βsim), and nestedness (βsne), where βsor = βsim + βsne. Heatmaps were generated using the Origin (2025) software (OriginLab, USA) to visualize the diversity patterns. To investigate the influence of environmental variables on the directional variation in community structure, principal coordinate analysis (PCoA) was conducted based on the Sørensen dissimilarity index. Using the “envfit()” function (permutations = 999,) in the vegan package, with a fixed random seed to ensure reproducibility, we fitted the three environmental variables that were fully recorded (i.e., discharge, filtration volume, and rainfall) onto the ordination plots, and only those with significant effects were displayed, thus allowing for a clear visualization of the direction and strength of their influence on community composition. To further assess the effects of rainfall occurrence (categorical variable: presence/absence) and filtration volume (continuous variable) on community dissimilarity (βsor, βsim, and βsne), we used Permutational Multivariate Analysis of Variance (PERMANOVA). This analysis was performed using 999 permutations with the “permute” and “vegan” packages, a fixed random seed was set to ensure reproducibility. The analysis was implemented as a marginal-effects model (i.e., by = "margin" in the “adonis2()” function), enabling the independent evaluation of each explanatory variable. As βsor is a non-Euclidean distance measure, it was square-root-transformed prior to analysis; βsim and βsne were adjusted using the Lingoes correction to meet the assumptions of PERMANOVA. All analyses were conducted using R (ver. 4.3.2, R Core Team, 2023). Declarations Acknowledgments We would like to thank Mr. Yamaguchi, Mr. Inoue, Mr. Higuchi, and Mr. Nanri (University of Miyazaki) for their help with the experiments and field observations. Funding Declaration This work was partially supported by the Ministry of Education, Science, Sports and Culture through a Grant-in-Aid for Scientific Research (grant numbers 24H00329, 25K00041, and 25H01201). Author Contributions CX performed the experiment, analyzed data, and led the writing of manuscript. KN conceived the ideas, supervised the research and acquired the funds. KN and CX performed the sampling and edited and approved the manuscript. Next-generation sequencing was performed by Bioengineering Lab through external outsourcing. Data Accessibility and Benefit-Sharing Data accessibility Raw reads generated during the current study are available in the DDBJ Sequence Read Archive (DRA) under the accession numbers DRA023808. The R code used for data analysis can be found in Supplementary files. Benefit sharing This study did not involve direct contact with animals in the survey area, particularly wild or endangered species. We are committed to investigating and detecting biodiversity without direct interaction with natural fauna. Conflict of Interest The authors declared no conflict of interest in this article. References Cardinale, B. J. et al. Biodiversity loss and its impact on humanity. Nature 486 , 59–67 (2012). Frøslev, T. G. et al. The biodiversity effect of reduced tillage on soil microbiota. Ambio 51 , 1022–1033 (2022). Redondo, M. A., Berlin, A., Boberg, J. & Oliva, J. Vegetation type determines spore deposition within a forest–agricultural mosaic landscape. FEMS Microbiology Ecology 96 , fiaa082 (2020). Yang, J., Zhang, L., Mu, Y. & Zhang, X. Small changes make big progress: A more efficient eDNA monitoring method for freshwater fish. Environmental DNA 5 , 363–374 (2023). Dyson, K. et al. Coupling remote sensing and eDNA to monitor environmental impact: A pilot to quantify the environmental benefits of sustainable agriculture in the Brazilian Amazon. PLoS ONE 19 , e0289437 (2024). Fediajevaite, J., Priestley, V., Arnold, R. & Savolainen, V. Meta‐analysis shows that environmental DNA outperforms traditional surveys, but warrants better reporting standards. Ecology and Evolution 11 , 4803–4815 (2021). Suren, A. M., Burdon, F. J. & Wilkinson, S. P. eDNA Is a Useful Environmental Monitoring Tool for Assessing Stream Ecological Health. Environmental DNA 6 , e596 (2024). Afzali, S. F. et al. Comparing environmental metabarcoding and trawling survey of demersal fish communities in the Gulf of St. Lawrence, Canada. Environmental DNA 3 , 22–42 (2021). Taberlet, P., Coissac, E., Pompanon, F., Brochmann, C. & Willerslev, E. Towards next‐generation biodiversity assessment using DNA metabarcoding. Molecular Ecology 21 , 2045–2050 (2012). Valentini, A. et al. Next‐generation monitoring of aquatic biodiversity using environmental DNA metabarcoding. Molecular Ecology 25 , 929–942 (2016). Huang, S., Yoshitake, K., Watabe, S. & Asakawa, S. Environmental DNA study on aquatic ecosystem monitoring and management: Recent advances and prospects. Journal of Environmental Management 323 , 116310 (2022). Ruppert, K. M., Kline, R. J. & Rahman, M. S. Past, present, and future perspectives of environmental DNA (eDNA) metabarcoding: A systematic review in methods, monitoring, and applications of global eDNA. Global Ecology and Conservation 17 , e00547 (2019). Sahu, A., Kumar, N., Pal Singh, C. & Singh, M. Environmental DNA (eDNA): Powerful technique for biodiversity conservation. Journal for Nature Conservation 71 , 126325 (2023). Takahashi, M. et al. Aquatic environmental DNA: A review of the macro-organismal biomonitoring revolution. Science of The Total Environment 873 , 162322 (2023). Meyer, J. M., Leempoel, K., Losapio, G. & Hadly, E. A. Molecular Ecological Network Analyses: An Effective Conservation Tool for the Assessment of Biodiversity, Trophic Interactions, and Community Structure. Front. Ecol. Evol. 8 , 588430 (2020). Thomsen, P. F. & Sigsgaard, E. E. Environmental DNA metabarcoding of wild flowers reveals diverse communities of terrestrial arthropods. Ecology and Evolution 9 , 1665–1679 (2019). Gogarten, J. F. et al. Fly‐derived DNA and camera traps are complementary tools for assessing mammalian biodiversity. Environmental DNA 2 , 63–76 (2020). Seeber, P. A. & Epp, L. S. Environmental DNA and metagenomics of terrestrial mammals as keystone taxa of recent and past ecosystems. Mammal Review 52 , 538–553 (2022). Newton, J. P., Allentoft, M. E., Bateman, P. W., Van Der Heyde, M. & Nevill, P. Targeting Terrestrial Vertebrates With eDNA : Trends, Perspectives, and Considerations for Sampling. Environmental DNA 7 , e70056 (2025). Guthrie, A. M. et al. A quantitative analysis of vertebrate environmental DNA degradation in soil in response to time, UV light, and temperature. Environmental DNA 6 , e581 (2024). Hassan, S. et al. Environmental DNA Metabarcoding: A Novel Contrivance for Documenting Terrestrial Biodiversity. Biology 11 , 1297 (2022). Rodgers, T. W. & Mock, K. E. Drinking water as a source of environmental DNA for the detection of terrestrial wildlife species. Conservation Genet Resour 7 , 693–696 (2015). Ushio, M. et al. Environmental DNA enables detection of terrestrial mammals from forest pond water. Molecular Ecology Resources 17 , e63-75 (2017). Martinelli Marín, D., Lasso, C. A. & Caballero Gaitan, S. J. eDNA metabarcoding: an effective tool for vertebrate diversity studies in the Colombian Amazon and Orinoco basins. Front. Ecol. Evol. 12 , 1409296 (2024). Reji Chacko, M. et al. Catchment-based sampling of river eDNA integrates terrestrial and aquatic biodiversity of alpine landscapes. Oecologia 202 , 699–713 (2023). Saenz‐Agudelo, P. et al. Monitoring vertebrate biodiversity of a protected coastal wetland using eDNA metabarcoding. Environmental DNA 4 , 77–92 (2022). Coutant, O. et al. Amazonian mammal monitoring using aquatic environmental DNA. Molecular Ecology Resources 21 , 1875–1888 (2021). Macher, T.-H. et al. Beyond fish eDNA metabarcoding: Field replicates disproportionately improve the detection of stream associated vertebrate species. MBMG 5 , e66557 (2021). Ryan, E., Bateman, P., Fernandes, K., Van Der Heyde, M. & Nevill, P. eDNA metabarcoding of log hollow sediments and soils highlights the importance of substrate type, frequency of sampling and animal size, for vertebrate species detection. Environmental DNA 4 , 940–953 (2022). Zhang, S., Zhao, J. & Yao, M. Urban landscape-level biodiversity assessments of aquatic and terrestrial vertebrates by environmental DNA metabarcoding. Journal of Environmental Management 340 , 117971 (2023). De Souza, L. S., Godwin, J. C., Renshaw, M. A. & Larson, E. Environmental DNA (eDNA) Detection Probability Is Influenced by Seasonal Activity of Organisms. PLoS ONE 11 , e0165273 (2016). Harper, L. R. et al. Environmental DNA (eDNA) metabarcoding of pond water as a tool to survey conservation and management priority mammals. Biological Conservation 238 , 108225 (2019). Mauvisseau, Q. et al. The Multiple States of Environmental DNA and What Is Known about Their Persistence in Aquatic Environments. Environ. Sci. Technol. 56 , 5322–5333 (2022). Nukazawa, K., Hamasuna, Y. & Suzuki, Y. Simulating the Advection and Degradation of the Environmental DNA of Common Carp along a River. Environ. Sci. Technol. 52 , 10562–10570 (2018). Seymour, M. et al. Acidity promotes degradation of multi-species environmental DNA in lotic mesocosms. Commun Biol 1 , 4 (2018). Strickler, K. M., Fremier, A. K. & Goldberg, C. S. Quantifying effects of UV-B, temperature, and pH on eDNA degradation in aquatic microcosms. Biological Conservation 183 , 85–92 (2015). Cowgill, C., Gilbert, J. D. J., Convery, I. & Lawson Handley, L. Monitoring terrestrial rewilding with environmental DNA metabarcoding: a systematic review of current trends and recommendations. Front. Conserv. Sci. 5 , 1473957 (2025). Deiner, K., Fronhofer, E. A., Mächler, E., Walser, J.-C. & Altermatt, F. Environmental DNA reveals that rivers are conveyer belts of biodiversity information. Nat Commun 7 , 12544 (2016). Xu, C. & Nukazawa, K. Detection of terrestrial species using environmental DNA during heavy rainfall events and associated influencing factors. Preprint at https://doi.org/10.22541/au.174312906.60655689/v1 (2025). Ladin, Z. S. et al. Assessing the efficacy of eDNA metabarcoding for measuring microbial biodiversity within forest ecosystems. Sci Rep 11 , 1629 (2021). Macher, T., Schütz, R., Hörren, T., Beermann, A. J. & Leese, F. It’s raining species: Rainwash eDNA metabarcoding as a minimally invasive method to assess tree canopy invertebrate diversity. Environmental DNA 5 , 3–11 (2023). Bae, M.-J. et al. Seasonal Dynamics and Heavy Rain Effects on the Diversity of Microeukaryome in the Nakdonggang River. Water 14 , 3407 (2022). Staley, Z. R. et al. Fecal source tracking and eDNA profiling in an urban creek following an extreme rain event. Sci Rep 8 , 14390 (2018). Condachou, C. et al. Tropical Rainfall and eDNA Washout Impact Estimations of Amazonian Biodiversity Patterns From Environmental Samples. Environmental DNA 7 , e70141 (2025). Osathanunkul, M. & Suwannapoom, C. A comparative study on eDNA-based detection of Siamese bat catfish (Oreoglanis siamensis) in wet and dry conditions. Sci Rep 14 , 8885 (2024). Harper, L. R. et al. Generating and testing ecological hypotheses at the pondscape with environmental DNA metabarcoding: A case study on a threatened amphibian. Environmental DNA 2 , 184–199 (2020). Urabe, H., Mizumoto, H., Tsuda-Yamaguchi, F. & Araki, H. Spatial heterogeneity of eDNA concentration as a predictor of small biomass of fish in a mountain stream. Limnology 26 , 223–233 (2025). Hunter, M. E., Ferrante, J. A., Meigs-Friend, G. & Ulmer, A. Improving eDNA yield and inhibitor reduction through increased water volumes and multi-filter isolation techniques. Sci Rep 9 , 5259 (2019). Sepulveda, A. J. et al. Improved detection of rare, endangered and invasive trout in using a new large‐volume sampling method for eDNA capture. Environmental DNA 1 , 227–237 (2019). Melendy, S. A. & Olson, J. R. Identifying Key Environmental Drivers of Reach‐Scale Salmonid eDNA Recovery With Random Forest. Environmental DNA 6 , e70001 (2024). Shogren, A. J. et al. Controls on eDNA movement in streams: Transport, Retention, and Resuspension. Sci Rep 7 , 5065 (2017). Sakata, M. K. et al. Determining an effective sampling method for eDNA metabarcoding: a case study for fish biodiversity monitoring in a small, natural river. Limnology 22 , 221–235 (2021). Curtis, A. N., Tiemann, J. S., Douglass, S. A., Davis, M. A. & Larson, E. R. High stream flows dilute environmental DNA (eDNA) concentrations and reduce detectability. Diversity and Distributions 27 , 1918–1931 (2021). Robinson, P. S., Davis, D. R., Collins, S. M. & Kline, R. J. Defining the current distribution of the imperiled Black-spotted Newt across south Texas, USA. Global Ecology and Conservation 36 , e02131 (2022). Lynggaard, C., Frøslev, T. G., Johnson, M. S., Olsen, M. T. & Bohmann, K. Airborne environmental DNA captures terrestrial vertebrate diversity in nature. Molecular Ecology Resources 24 , e13840 (2024). Ragot, R. & Villemur, R. eDNA profiling of mammals, birds, and fish of surface waters by mitochondrial metagenomics: application for source tracking of fecal contamination in surface waters. Environ Monit Assess 194 , 72 (2022). Broadhurst, H. A. et al. From Water to Land: A Review on the Applications of Environmental DNA and Invertebrate‐Derived DNA for Monitoring Terrestrial and Semi‐Aquatic Mammals. Mammal Review e70006 (2025) doi:10.1111/mam.70006. Ruan, H.-T. et al. Effects of sampling strategies and DNA extraction methods on eDNA metabarcoding: A case study of estuarine fish diversity monitoring. Zoological Research 43 , 192–204 (2022). Wang, C. et al. Effects of Filtration Volumes on Bacterial Diversity and Community Structure in Freshwater Lakes. Ecology and Evolution 15 , e71445 (2025). Chase, J. M., Kraft, N. J. B., Smith, K. G., Vellend, M. & Inouye, B. D. Using null models to disentangle variation in community dissimilarity from variation in α-diversity. Ecosphere 2 , art24 (2011). Kawakami, T. et al. Evaluating the sampling effort for the metabarcoding‐based detection of fish environmental DNA in the open ocean. Ecology and Evolution 13 , e9921 (2023). Baudry, T., Vasselon, V., Delaunay, C., Sweet, M. & Grandjean, F. When Methodological Innovation Changes the Game: A 10‐Year Review of Environmental DNA (eDNA) Applied to Crayfish. Aquatic Conservation 34 , e4245 (2024). Riaz, T. et al. ecoPrimers: inference of new DNA barcode markers from whole genome sequence analysis. Nucleic Acids Research 39 , e145–e145 (2011). Collins, R. A. et al. Non‐specific amplification compromises environmental DNA metabarcoding with COI. Methods Ecol Evol 10 , 1985–2001 (2019). Veron, P. et al. Environmental DNA complements scientific trawling in surveys of marine fish biodiversity. ICES Journal of Marine Science 80 , 2150–2165 (2023). Chung, A., Kam, Y. C., Shea, S. K. H. & Schunter, C. Detecting Fish Diversity in Urban‐Impacted Ecosystems: A Comparative Approach of eDNA Metabarcoding and UVC. Environmental DNA 6 , e70048 (2024). Holman, L. E., Chng, Y. & Rius, M. How does eDNA decay affect metabarcoding experiments? Environmental DNA 4 , 108–116 (2022). Marinchel, N. et al. Mock community experiments can inform on the reliability of eDNA metabarcoding data: a case study on marine phytoplankton. Sci Rep 13 , 20164 (2023). Allen, M. C. et al. Sampling environmental DNA from trees and soil to detect cryptic arboreal mammals. Sci Rep 13 , 180 (2023). Mineda, H., Nukazawa, K. & Suzuki, Y. Assessing environmental flow alterations induced by dams and climate change using a distributed hydrological model at catchment scale. Journal of Water and Climate Change 14 , 4821–4836 (2023). Brooks, M., E. et al. glmmTMB Balances Speed and Flexibility Among Packages for Zero-inflated Generalized Linear Mixed Modeling. The R Journal 9 , 378 (2017). Baselga, A. Partitioning the turnover and nestedness components of beta diversity. Global Ecology and Biogeography 19 , 134–143 (2010). Additional Declarations No competing interests reported. Supplementary Files Supplementary.zip Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7806039","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":556386215,"identity":"d7fda3ad-2335-4efe-8b74-59592989eef2","order_by":0,"name":"Xu Chen","email":"","orcid":"","institution":"University of Miyazaki","correspondingAuthor":false,"prefix":"","firstName":"Xu","middleName":"","lastName":"Chen","suffix":""},{"id":556386218,"identity":"2439767c-f023-47f6-9fca-1416898ab282","order_by":1,"name":"Nukazawa Kei","email":"data:image/png;base64,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","orcid":"","institution":"University of Miyazaki","correspondingAuthor":true,"prefix":"","firstName":"Nukazawa","middleName":"","lastName":"Kei","suffix":""}],"badges":[],"createdAt":"2025-10-08 09:08:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7806039/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7806039/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97736417,"identity":"c3654205-80d8-4dba-beb8-8665ea82e7ef","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"jpg","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1512408,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/28934e0ec8270fab8db5b865.jpg"},{"id":97736421,"identity":"35aed1a1-1f29-4797-a3d6-6e4c67862d14","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":160233,"visible":true,"origin":"","legend":"","description":"","filename":"ManuscriptXu.docx","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/57102d64bc144010f2ce980b.docx"},{"id":97895579,"identity":"3e1e4cc5-a06a-415e-bffc-a020a8f33488","added_by":"auto","created_at":"2025-12-10 15:34:29","extension":"jpg","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":202450,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/3e0f742190dd47c2812d9475.jpg"},{"id":97894492,"identity":"0216102a-4e81-469b-9452-469041d0ccf0","added_by":"auto","created_at":"2025-12-10 15:32:37","extension":"jpg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5689365,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/5e47aa313dc10babd9365f0f.jpg"},{"id":97895836,"identity":"06b1e7a0-4bcd-45b2-9f8e-bbb7075d16df","added_by":"auto","created_at":"2025-12-10 15:35:10","extension":"jpg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":614363,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/4d28218a5d83fffe3156879d.jpg"},{"id":97736424,"identity":"5c92f289-7f87-4808-be5c-65b4374b3688","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"jpg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1492523,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/5bea44b76115fb4e6fb3e931.jpg"},{"id":97894554,"identity":"94c0d614-2aba-48f5-a67f-f11e1c8e3801","added_by":"auto","created_at":"2025-12-10 15:32:42","extension":"json","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4655,"visible":true,"origin":"","legend":"","description":"","filename":"09b0e1d3cb314378ba6bfbc9a70d8504.json","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/38ab9868d3a1534ccf5a6ce9.json"},{"id":97736432,"identity":"344d6c6d-30d8-470f-9c66-21c835f941a0","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"zip","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1082445,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementary.zip","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/86744655cf2e4cb6fc86628f.zip"},{"id":97896526,"identity":"389881b6-06f8-40a0-9215-fe4ef97dbc54","added_by":"auto","created_at":"2025-12-10 15:36:44","extension":"xml","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":172716,"visible":true,"origin":"","legend":"","description":"","filename":"09b0e1d3cb314378ba6bfbc9a70d85041enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/c736e7ec5df8d7ebdc2472d7.xml"},{"id":97895406,"identity":"9e192810-c9aa-4e38-b66a-c31466b88142","added_by":"auto","created_at":"2025-12-10 15:34:08","extension":"jpg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1512408,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/8c2b28d30bea3ee843cd32a1.jpg"},{"id":97895463,"identity":"2f59cd0d-d5e7-4b3c-bb49-ae5f5c350cbb","added_by":"auto","created_at":"2025-12-10 15:34:15","extension":"jpg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":202450,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/d9e5d3ea03c742ab8f4d0f59.jpg"},{"id":97736443,"identity":"bcfd1bf0-fc82-4672-9214-a54df8ae5a8f","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"jpg","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5689365,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/80f5f3625d8f68f19a377c99.jpg"},{"id":97736436,"identity":"78e3467e-25e5-4acc-81a6-dca4ea2a3ad3","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"jpg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":614363,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/947473ecfc4148fcab4561da.jpg"},{"id":97736438,"identity":"1efdcf03-a27d-4623-9d3f-1e208a6a2449","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"jpg","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1492523,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/b2cf89ed8d52fdcffb7dd46e.jpg"},{"id":97736441,"identity":"d0d4c72f-6c21-4e2b-bb0e-9118f4a1a04a","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":428290,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig.1.png","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/7bede68b9d9a653a7172db59.png"},{"id":97895565,"identity":"236caceb-1f92-4504-af16-8018fe5a74e9","added_by":"auto","created_at":"2025-12-10 15:34:27","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":45648,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig.2.png","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/57f702bb6eb555102bd585ec.png"},{"id":97894269,"identity":"e39dc2cb-6c13-4ddc-b19a-e13ca14127a2","added_by":"auto","created_at":"2025-12-10 15:32:09","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1128180,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig.3.png","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/635218657caa1ebb774dc56f.png"},{"id":97896090,"identity":"887ff74c-c2e0-4f37-b887-552d58c9af22","added_by":"auto","created_at":"2025-12-10 15:35:50","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":140687,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig.4.png","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/893406a418dedc30cd5a7631.png"},{"id":97895582,"identity":"66a6c7e1-18a2-4239-bfa1-bcde0b114834","added_by":"auto","created_at":"2025-12-10 15:34:29","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":535716,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig.5.png","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/f7ade9b7a89a5e537911e2b2.png"},{"id":97895361,"identity":"d81f5ff3-e2ad-411e-9ec6-08283889641c","added_by":"auto","created_at":"2025-12-10 15:34:04","extension":"xml","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":171466,"visible":true,"origin":"","legend":"","description":"","filename":"09b0e1d3cb314378ba6bfbc9a70d85041structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/94adccaa84f04ecec187a2fc.xml"},{"id":97896674,"identity":"f084a71c-8f59-4125-9563-e40935e67145","added_by":"auto","created_at":"2025-12-10 15:36:52","extension":"html","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":182485,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/4ea14402d98f40f16dbe58ab.html"},{"id":97736419,"identity":"981baae8-c3be-4dde-b825-5500273aa671","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1512408,"visible":true,"origin":"","legend":"\u003cp\u003eDetection results for each river. Circles indicate presence, with red, yellow, blue, and green representing the detections of fish, amphibians, birds, and mammals, respectively. “S1” and “S2” indicate the first and second sampling sessions on each sampling day, respectively. “R1” and “R2” represent sample replicates 1 and 2. “Normal” denotes sampling conducted under non-rainfall conditions. The samplings in May, September, and November were conducted in 2022, the samplings in June in 2023, and the samplings in August in 2024. The non-rainfall samplings in November were conducted on the day before the rainfall event. Filled stars indicate waterbirds.\u003c/p\u003e","description":"","filename":"Fig.1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/b0c5ef9cb2fb74c27a7859d0.jpg"},{"id":97736416,"identity":"d19d7a0a-b18f-4e55-94eb-8d13972bdecb","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":202450,"visible":true,"origin":"","legend":"\u003cp\u003eDetection rates of each taxonomic group and overall detection rates at each sampling site. “S1” and “S2” indicate the first and second sampling sessions on each sampling day, respectively. “R1” and “R2” represent sample replicates 1 and 2, respectively. “Normal” denotes sampling conducted under non-rainfall conditions. The samplings in May, September, and November were conducted in 2022, the samplings in June in 2023, and the samplings in August in 2024. The non-rainfall samplings in November were conducted on the day before the rainfall event. “Total” indicates the overall detection rates for all taxonomic groups. The results within the rectangle black lines represent data collected during non-rainfall periods.\u003c/p\u003e","description":"","filename":"Fig.2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/3ae61acf36be6f33bf126f93.jpg"},{"id":97894353,"identity":"adf5462e-6434-4537-8c76-62c03f017535","added_by":"auto","created_at":"2025-12-10 15:32:23","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":5689365,"visible":true,"origin":"","legend":"\u003cp\u003eResults of βsor, βsim, and βsne, calculated based on the Sørensen dissimilarity index, for three sampling sites. “S1” and “S2” indicate the first and second sampling sessions on each sampling day, respectively. “R1” and “R2” represent sample replicates 1 and 2, respectively. “Normal” denotes sampling conducted under non-rainfall conditions. The samplings in May, September, and November were conducted in 2022, the samplings in June in 2023, and the samplings in August in 2024. The non-rainfall samplings in November were conducted on the day before the rainfall event.\u003c/p\u003e","description":"","filename":"Fig.3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/aa51d638ccef794f90e0066a.jpg"},{"id":97894475,"identity":"7ba63d4b-a462-4841-9c38-741ea39ecc16","added_by":"auto","created_at":"2025-12-10 15:32:35","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":614363,"visible":true,"origin":"","legend":"\u003cp\u003ePrincipal coordinates analysis of eDNA community composition based on the Sørensen dissimilarity index. Points represent individual samples, with red indicating no-rainfall samples and blue indicating rainfall samples. Environmental vectors fitted using “envfit” are shown as arrows; arrow direction indicates the gradient of the corresponding environmental factor, and arrow length reflects its correlation strength with community variation. Only significant factors (p \u0026lt;0.05) are shown.\u003c/p\u003e","description":"","filename":"Fig.4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/36e3c847f0873efd7929f45d.jpg"},{"id":97895303,"identity":"4a17a426-2750-465f-aa7a-9dbf1d857cbd","added_by":"auto","created_at":"2025-12-10 15:33:57","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1492523,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Sampling sites, a weather station in the Kiyotake River Basin, located in southwestern Japan, and (b) a land use distribution map. The maps were processed using ArcGIS, with geographic layers obtained from the Geospatial Information Authority of Japan (https://nlftp.mlit.go.jp/ksj/). Sampling sites were plotted based on GPS coordinates. The watershed boundary was delineated using ArcGIS. All other graphic elements were prepared by the authors. Photos of the river after rainfall were taken by the authors.\u003c/p\u003e","description":"","filename":"Fig.5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/ac6f8350807c04023145591e.jpg"},{"id":104402961,"identity":"8f282f14-2e06-4306-a689-9a5bca9d716f","added_by":"auto","created_at":"2026-03-11 12:17:00","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":10309259,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/f27e936f-8405-4b86-9282-5cb1f38fd2c8.pdf"},{"id":97736427,"identity":"dbf61d17-e707-4d4d-a760-6489e7c0303c","added_by":"auto","created_at":"2025-12-08 19:40:05","extension":"zip","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1082445,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementary.zip","url":"https://assets-eu.researchsquare.com/files/rs-7806039/v1/0172a70b1b50c850967acce1.zip"}],"financialInterests":"No competing interests reported.","formattedTitle":"Impact of Rainfall on Aquatic and Terrestrial Species Monitoring Using Riverine Environmental DNA","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGlobal climate change and human activity have accelerated biodiversity loss, leading to the significant degradation of ecosystem services\u0026nbsp;\u003csup\u003e1\u003c/sup\u003e. Environmental DNA (eDNA) enables the development and implementation of effective monitoring strategies for timely, evidence-based conservation. eDNA, comprising genetic material (e.g., skin and feces) shed by organisms and retained in environmental substrates (e.g., water, soil, and air)\u0026nbsp;\u003csup\u003e2\u0026ndash;4\u003c/sup\u003e. It has many advantages, including low cost, rapidity, and non-invasiveness\u0026nbsp;\u003csup\u003e5\u0026ndash;7\u003c/sup\u003e. In addition,\u0026nbsp;eDNA metabarcoding, which combines next-generation sequencing and universal primers,\u0026nbsp;facilitates large-scale biodiversity surveying\u0026nbsp;\u003csup\u003e8\u0026ndash;10\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eVertebrate eDNA research has mainly focused on aquatic species\u0026nbsp;\u003csup\u003e11\u0026ndash;14\u003c/sup\u003e. For terrestrial species, various sampling strategies and substrates (e.g., air, soil, mineral licks, feeding traces, and hair) have been employed. For instance, feces have been used to detect carnivores\u0026nbsp;\u003csup\u003e15\u003c/sup\u003e,\u0026nbsp;while DNA extracted from wildflowers has been used to identify arthropods\u0026nbsp;and birds\u0026nbsp;\u003csup\u003e16\u003c/sup\u003e. Tailoring eDNA sampling protocols to specific taxonomic groups or ecological contexts may improve the comprehensiveness and reliability of biodiversity assessments. However, identifying suitable environmental substrates for terrestrial vertebrate eDNA is time-consuming and resource-intensive\u0026nbsp;\u003csup\u003e17,18\u003c/sup\u003e. These limitations may restrict the range of species detected \u003csup\u003e19\u003c/sup\u003e. Moreover, considerable uncertainty remains regarding the ecology of terrestrially derived eDNA, including its origin, physicochemical properties, transport mechanisms, and environmental fate\u0026nbsp;\u003csup\u003e20,21\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eTerrestrial species have been reported to release DNA into water via activities such as drinking and swimming\u0026nbsp;\u003csup\u003e22,23\u003c/sup\u003e.\u0026nbsp;This provides a theoretical basis for water-based eDNA approaches for the simultaneous and rapid detection of terrestrial and aquatic species. Building on this, the approach has been applied across various ecosystems, such as forests and urban areas\u0026nbsp;\u003csup\u003e24\u0026ndash;26\u003c/sup\u003e, and validated in watersheds of different spatial scales, ranging from small streams to the Amazon River\u0026nbsp;\u003csup\u003e27,28\u003c/sup\u003e. However, in water-based eDNA biodiversity monitoring within watersheds, the detection rates of terrestrial species are generally lower than those of aquatic species\u0026nbsp;\u003csup\u003e29,30\u003c/sup\u003e. This discrepancy largely arises from the continuous DNA shedding by aquatic organisms (e.g., fish) into their environments, which differs from the spatially and temporally intermittent interactions of terrestrial species with water bodies, resulting in limited eDNA deposition\u0026nbsp;\u003csup\u003e31\u003c/sup\u003e. This indicates that target species may not be detected,\u0026nbsp;even if\u0026nbsp;they\u0026nbsp;are present\u0026nbsp;near\u0026nbsp;sampling locations\u0026nbsp;\u003csup\u003e32\u003c/sup\u003e. Furthermore, the detectability of eDNA in aquatic systems is influenced by environmental factors\u0026nbsp;such as pH, water temperature, and flow distance\u0026nbsp;\u003csup\u003e33\u0026ndash;36\u003c/sup\u003e. This highlights the need to use water-based approaches to elucidate the effect of environmental factors on the detection of aquatic and terrestrial species.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDetection in terrestrial substrates and in aquatic environments has inherent limitations\u0026nbsp;\u003csup\u003e37\u003c/sup\u003e. Currently, water is the most commonly used eDNA substrate for assessing biodiversity\u0026nbsp;\u003csup\u003e19\u003c/sup\u003e. This may be because of the unique advantage of water in connecting terrestrial and aquatic habitats as well as upstream and downstream areas. Such connectivity facilitates the use of water-based eDNA approaches to evaluate species diversity across broader spatial scales\u0026nbsp;\u003csup\u003e4,38\u003c/sup\u003e. In our previous study on a specific terrestrial mammal, we confirmed that eDNA could be transported into rivers via surface runoff during rainfall events\u0026nbsp;\u003csup\u003e39\u003c/sup\u003e. This will help improve the detection efficiency for large-scale assessments of terrestrial mammal diversity. In addition to terrestrial mammals, rainwater collected after precipitation events in forested ecosystems has been shown to contain a high diversity of invertebrate and microbial taxa\u0026nbsp;\u003csup\u003e40,41\u003c/sup\u003e. In riverine ecosystems, post-rainfall river water samples were found to have increased microbial diversity\u0026nbsp;\u003csup\u003e42\u003c/sup\u003e. However, studies on vertebrates in such settings\u0026nbsp;are\u0026nbsp;limited. A previous study found that, compared to eDNA samples collected during non-rainfall periods, the number of fish species detected in river water decreased after rainfall, whereas the number of mammal species detected increased\u0026nbsp;\u003csup\u003e43\u003c/sup\u003e. This indicated that rainfall had distinct effects on the detection of vertebrate species with different ecological habits. Moreover, different rainfall patterns (e.g., intensity and duration) may affect species detection. For instance, heavy-rainfall periods are less suitable for fish detection, compared to light rainfall. Meanwhile, compared to peak rainfall, rainfall duration is more conducive to the detection of terrestrial mammals\u0026nbsp;\u003csup\u003e39,44,45\u003c/sup\u003e. In conclusion,\u0026nbsp;it is necessary to further investigate the effects of rainfall patterns on the detection of vertebrates, including fish and mammals.\u003c/p\u003e\n\u003cp\u003eThis study developed a method for detecting aquatic and terrestrial biological communities in riverine stormwater collected after rainfall events in the Kiyotake River catchment, Miyazaki Prefecture, Japan. Furthermore, we evaluated the effects of different rainfall patterns on species richness and community composition across sub-catchments with contrasting land use types, providing new insights into how hydrological and landscape factors influence eDNA-based biodiversity assessments.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eeDNA\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003emetabarcoding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 1,952,639 sequences were obtained from the 33 eDNA samples. One sample collected from the Oka River in June 2023 showed no detection. Among the sequences, 93.15% had a quality score \u0026gt;30. After removing the chimeric and noisy sequences, 812,677 reads remained. Following the removal of non-vertebrate species, humans, and environmental contaminants, 670,036 reads remained and were classified as fish, birds, amphibians, or mammals. Fish accounted for the largest proportion (65.9%), whereas birds represented the smallest proportion (4.02%). In total, 64 taxa were detected, with 26 taxa identified at the species level, representing 43.1% of the total.\u003c/p\u003e\n\u003cp\u003eFigure 1 shows the detection results for each river (for the complete dataset, see Supplementary 1). In the Mizunashi River, 25 taxa were detected, comprising six fish species, seven amphibians, six birds, and six mammals. In the Oka River, 30 taxa, including 14 fish, seven amphibians, three birds, and six mammals, were identified. The Tagami River had the highest taxonomic richness, with 38 taxa, specifically 19 fish, 10 amphibians, five birds, and four mammals. The Oka River and Tagami River showed a higher dominance of fish species. Most of the detected mammalian taxa were domesticated or human-associated species (e.g., cattle, pigs, and dogs), particularly in the Tagami River, where no wild mammals were detected. The wild mammals identified in this study included \u003cem\u003eMeles anakuma\u003c/em\u003e, \u003cem\u003eMogera wogura\u003c/em\u003e, \u003cem\u003eUrotrichus talpoides\u003c/em\u003e, \u003cem\u003eLepus\u003c/em\u003e spp. (wild rabbits), \u003cem\u003eCanidae\u003c/em\u003e spp. (domestic or raccoon dogs), and \u003cem\u003eSus scrofa\u0026nbsp;\u003c/em\u003e(wild boars or domestic pigs). These were all small-bodied terrestrial mammals detected only after rainfall events. Amphibian taxa were also predominantly detected after rainfall events. All belonged to the order Anura, with the exception of \u003cem\u003eParamesotriton\u003c/em\u003e sp., a member of the order Caudata.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDetection rate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFigure 2 shows the overall detection probability for each sample. For the Mizunashi River, the detection rates ranged from 5.26% to 57.89%, with the lowest rate observed in June 2023. The two sampling events without rainfall yielded detection rates of 21.05% and 36.84%. In the Oka River, the detection rates ranged from 0% to 48.15%, with the lowest recorded in June 2023. The two non-rainfall sampling events showed detection rates of 48.15% and 33.33%. For the Tagami River, the detection rates varied between 15.15% and 54.55%, with the lowest rate occurring in November 2022. The two non-rainfall sampling events had detection rates of 21.05% and 36.36%. Regarding different taxonomic groups, the detection rates in the Mizunashi River ranged from 0% to 100% for fish, 0% to 57.14% for amphibians, and 0% to 50% for mammals. In the Oka River, the detection rates ranged from 0% to 85.71% for fish, 0% to 57.14% for amphibians, and 0% to 66.67% for mammals. Meanwhile, for the Tagami River, the detection rates were 10.53\u0026ndash;63.16% for fish, 0\u0026ndash;60% for amphibians, and 0\u0026ndash;100% for mammals.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRelationship\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;between detection number and environmental factors\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTable 2 presents the results of the GLMs. The GLMs revealed that filtration volume had a significant positive effect on the number of detections, whereas discharge and pH had significant negative effects on the detection number.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCommunity structure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFigure 3 shows the \u0026beta;sor, \u0026beta;sim, and \u0026beta;sne results, calculated based on S\u0026oslash;rensen dissimilarity, for the three sampling sites. The average of \u0026beta;sor, \u0026beta;sim, and \u0026beta;sne values for each river were as follows: the Mizunashi River: \u0026beta;sor = 0.56, \u0026beta;sim = 0.35, and \u0026beta;sne = 0.21; the Oka River (without no detection sample): \u0026beta;sor = 0.56, \u0026beta;sim = 0.42, and \u0026beta;sne = 0.14; the Tagami River: \u0026beta;sor = 0.55, \u0026beta;sim = 0.39, and \u0026beta;sne = 0.16.\u003c/p\u003e\n\u003cp\u003eFigure 4 shows the result of PCoA. PCoA based on the S\u0026oslash;rensen index showed that the first two axes explained 40.78% of the variation in community composition in the Mizunashi River. The two groups of samples did not form distinct clusters in the PCoA plot. Envfit analysis indicated that river discharge (r\u0026sup2; = 0.60, p = 0.002), filtration volume (r\u0026sup2; = 0.52, p = 0.019), and rainfall (r\u0026sup2; = 0.60, p = 0.020) were significant drivers of community structure. In the Oka River, the first two PCoA axes explained 43.74% of the variation, and similarly, no distinct clustering was observed between the two groups. Envfit analysis also revealed that filtration volume (r\u0026sup2; = 0.47, p = 0.029) and rainfall (r\u0026sup2; = 0.81, p = 0.017) were significant drivers of community structure. In contrast to the other two rivers, the Tagami River showed a different pattern. The first two PCoA axes explained 42.04% of the variation, and the two groups of samples formed a distinct separation in the PCoA plot. Envfit analysis identified\u0026nbsp;discharge\u0026nbsp;(r\u0026sup2; = 0.55, p = 0.029), filtration volume (r\u0026sup2; = 0.88, p = 0.030), and rainfall (r\u0026sup2; = 0.51, p = 0.041) as significant drivers of community structure.\u003c/p\u003e\n\u003cp\u003eTable 3 presents the results of the PERMANOVA analysis, which supports the findings obtained from PCoA. In the Mizunashi River, filtration volume had a significant effect on \u0026beta;sor (R\u0026sup2; = 0.14, p = 0.047). In the Oka River, filtration volume significantly affected \u0026beta;sor (R\u0026sup2; = 0.15, p = 0.038). In the Tagami River, filtration volume had a significant influence on both \u0026beta;sor (R\u0026sup2; = 0.22, p =0.004) and \u0026beta;sne (R\u0026sup2; = 0.27, p = 0.002). Contrastingly, the effects of rainfall were relatively weak, with a significant impact observed only for \u0026beta;sor in the Tagami River (R\u0026sup2; = 0.15, p = 0.048). When the filtration volume was standardized, rainfall still had a significant effect (R\u0026sup2; = 0.23, p = 0.041).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eUsing eDNA metabarcoding, this study investigated the effects of rainfall on the detection of different vertebrate taxonomic groups, by collecting river water samples after rainfall events with varying patterns. A total of 64 taxa, including fishes, mammals, birds, and amphibians, were detected in the 33 samples collected from the three rivers. The mammals detected were primarily livestock and small wild species, while the amphibians were mainly frogs. For wild boars (\u003cem\u003eS. scrofa\u003c/em\u003e), the 12S-V primer used in this study could not distinguish between wild and domestic pigs \u003csup\u003e46\u003c/sup\u003e. Although both types could exist in the study area, this study focused on\u0026nbsp;the overall ecological\u0026nbsp;patterns; therefore, the related sequences were conservatively assigned to the species level (\u003cem\u003eS. scrofa\u003c/em\u003e). Future studies should consider using specific primers to differentiate between wild boar and domestic pigs. In the Tagami River, one Amplicon Sequence Variant (ASV) was identified as \u003cem\u003eRana zhenhaiensis\u003c/em\u003e based on our analytical pipeline. However, the likelihood of this species occurring locally is extremely low, as it is distributed only in China, and the sequence was more likely derived from \u003cem\u003eRana japonica\u003c/em\u003e. Nevertheless, we retained this result because of its potential relevance to invasive species monitoring.\u003c/p\u003e\n\u003cp\u003eWe included two replicate samples for each individual sampling event; however, the number and composition of the detected taxa varied between replicates. Regarding the number of detections, discrepancies among replicates with differing filtration volumes ranged from 3 to 7 taxa in the Mizunashi River, 1 to 3 taxa in the Oka River, and 0 to 1 taxon in the Tagami River. These volume-related inconsistencies were most pronounced in the Mizunashi River, which is characterized by a low species richness and a relatively large catchment area. Larger filtration volumes tended to yield a greater number of detected taxa. In contrast, when filtration volumes were consistent across replicates, the number of taxa detected differed by at most one per site. Although rainfall may disrupt the spatial homogeneity of the eDNA distribution in river water \u003csup\u003e47\u003c/sup\u003e, increasing the filtration volume can enhance the comprehensiveness of biodiversity assessments \u003csup\u003e48,49\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTwo sampling events were conducted on specific dates. Although rainfall persisted, turbidity generally decreased during the second sampling period owing to a reduced rainfall intensity, allowing for higher filtration volumes. The number of fish taxa decreased during the second sampling event in the larger catchments of the Mizunashi River and the Oka River, whereas it increased in the Tagami River. This may be attributed to stronger dilution caused by rainfall in larger catchments \u003csup\u003e50\u003c/sup\u003e. Additionally, under the conditions of prolonged heavy rainfall, an increased detection of fish eDNA may be related to the resuspension of eDNA from riverbed sediments \u003csup\u003e51\u003c/sup\u003e. Most of the taxa newly detected during the second sampling period in the Tagami River belonged to the Gobiidae family, which is primarily benthic. Previous studies have suggested that sediments may be more effective than water for detecting eDNA from benthic fish species \u003csup\u003e52\u003c/sup\u003e. However, because of the limited taxonomic resolution of the 12S rRNA primers used in this study, it remains unclear whether the newly detected taxa were benthic, midwater, or surface-dwelling species. Moreover, we did not apply metabarcoding to riverbed sediments, and this conclusion requires further validation. In May 2022, continuous rainfall led to an increase in the number of amphibian and mammal taxa detected during the second sampling period in the Mizunashi River, whereas the opposite trend was observed in the Oka River. In the Mizunashi River, most of the mammals detected were livestock, and eDNA sources were abundant near the sampling sites. Continuous rainfall likely transported their terrestrial eDNA into the river, resulting in increased mammal detection during the second sampling period. By contrast, the first sampling period in the Oka River included wild animals, which have limited eDNA sources. The dilution effect of continuous rainfall outweighed the replenishment of eDNA from surface runoff, leading to a decrease in detection during the second sampling period.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDetection rate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe lowest detection rate was observed after the rainfall period, which was attributed to the reduced detection rate for fish taxa after rainfall. A previous study demonstrated that rainfall may dilute the eDNA of aquatic species, leading to reduced detection \u003csup\u003e53\u003c/sup\u003e. In addition, increased turbidity resulting from rainfall hindered the filtration of adequate water volumes (\u0026lt;400 mL) and would promote an inhibition of PCR. Thus, such effects after rainfall obscured fish eDNA, and the overall detection rates across broad taxonomic groups decreased. Nevertheless, amphibians and mammals exhibited higher detection rates after rainfall events. Although rainfall dilutes the eDNA of terrestrial species in water, the transportation of terrestrial eDNA through surface runoff helps replenish the diluted portions \u003csup\u003e44\u003c/sup\u003e. In addition, considering the behavioral characteristics of amphibians, the increase in detection rates may not be entirely attributed to surface runoff. Amphibians tend to be more active after rainfall, thereby releasing more eDNA \u003csup\u003e54\u003c/sup\u003e, explaining why they are almost exclusively detected after rainfall events.\u003c/p\u003e\n\u003cp\u003eA comparison of the detection rate for birds, performed using samples collected after the same period (non-rainfall: November 28, 2022; rainfall: November 29), revealed that, despite the lower filtration volume after a rainfall event (350\u0026ndash;690 mL), the number of bird detections was still higher than that during the non-rainfall period (1,000 mL). This is particularly evident because only one species of Anatidae, a group of waterbirds, was detected on non-rainy days, whereas both waterbirds (\u003cem\u003eAnatidae.sp, Galkinula chloropus\u003c/em\u003e) and non-waterbirds (\u003cem\u003ePasseriformes.sp, Gallus gallus)\u003c/em\u003e were detected on rainy days. It is possible that bird DNA present in the air was introduced into the river via rainfall \u003csup\u003e55\u003c/sup\u003e and that bird droppings were transported to the river via runoff after rainfall \u003csup\u003e56\u003c/sup\u003e. In this short-term comparison, bird species also showed higher detection rates on rainfall days than on non-rainfall days, which is consistent with the detection trends observed for mammals and amphibians. This further suggests that rainfall could enhance eDNA inputs and detection for certain terrestrial taxa. Thus, designing sampling strategies that consider species-specific life histories is necessary. For example, fish were more suitably surveyed during non-rainfall periods, whereas amphibians were better targeted in post-rainfall surveys. While no sampling method can achieve a 100% detection rate for all organisms \u003csup\u003e57\u003c/sup\u003e, our results demonstrated that vertebrate surveys based on water eDNA after rainfall can improve the detection of mammals, birds, and amphibians.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRelationship between detection number and environmental factors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRiver discharge and pH had significant negative effects on detection number. This was likely because increased discharge during rainfall dilutes eDNA concentrations \u003csup\u003e53\u003c/sup\u003e, thereby reducing the detection number. Although many studies have reported that acidic conditions accelerate eDNA degradation \u003csup\u003e35,36\u003c/sup\u003e, our results showed a negative correlation between pH and detection number. The reason for this pattern is unclear; however, a low pH may enhance DNA adsorption onto suspended particles, increasing its recoverability \u003csup\u003e33\u003c/sup\u003e. Given that our study was conducted in turbid waters, this may be an explanation; however, further research is required to determine whether adsorption can offset the effects of acidity on DNA degradation. Contrastingly, filtration volume had a significant positive effect on detection number, as filtering larger volumes of water captures more eDNA \u003csup\u003e58\u003c/sup\u003e, leading to more detection counts. Rainfall did not significantly affect the number of detections. This was because we could not determine when rainfall began to influence detection, and we, therefore, used the total rainfall within 3 h prior to sampling as the environmental variable, although the actual effective time may not be 3 h.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCommunity structure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAmong the samples collected during the same period, the highest S\u0026oslash;rensen dissimilarity indices were consistently observed in the Mizunashi River and Oka River in June 2023, and in the Tagami River in November 2022. These high dissimilarities may not have been directly caused by rainfall, but rather by filter clogging induced by rainfall, which resulted in the lowest filtration volumes for these sampling events. Compared to samples with larger filtration volumes, those with lower filtration volumes exhibited greater variability in terms of species composition \u003csup\u003e59\u003c/sup\u003e, and the S\u0026oslash;rensen dissimilarity index could be affected by low DNA concentrations or the non-detection of certain taxa in some samples \u003csup\u003e60\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eUnder identical filtration volume conditions, \u0026beta;sor in the Mizunashi River and the Oka River were nearly identical between rainfall and non-rainfall periods. In the Mizunashi River, when replicate samples exhibited differences in filtration volume, \u0026beta;sor was primarily driven by nestedness, which was likely caused by the decrease in recovery due to a reduction in filtration volume. In contrast, in the Tagami River, the \u0026beta;sor during rainfall events was lower than that under non-rainfall conditions and was mainly driven by species turnover. The results of the PCoA and the PERMANOVA analyses indicated that, in the Mizunashi River and the Oka River, rainfall did not significantly affect \u0026beta;sor, whereas filtration volume had a significant effect. Although filtration volume is an important factor influencing \u0026beta;sor \u003csup\u003e61\u003c/sup\u003e, excessively large differences in filtration volume in these rivers may obscure the effect of rainfall. This is supported by the results obtained from the Tagami River, where the filtration volume was relatively consistent, indicating that both filtration volume and rainfall had significant effects. These findings suggest that, within the same river system, the influence of filtration volume may outweigh that of rainfall, highlighting the importance of filtration volume in the sampling design for species detection. Furthermore, when filtration volumes were equivalent, evaluating only the effect of rainfall revealed that the species composition tended to be more homogeneous \u003csup\u003e44\u003c/sup\u003e. This could be attributed to the lower water temperatures and reduced ultraviolet radiation following rainfall events, which favor the preservation of eDNA in water columns \u003csup\u003e36\u003c/sup\u003e. Owing to the limited number of samples, further studies with larger sample sizes and standardized filtration volumes are required to verify the statistical significance and generality of this trend. However, in studies involving turbid water, the rapid filtering of large volumes remains a challenge \u003csup\u003e62\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eComparisons across different watersheds further highlighted the importance of rainfall and filtration volume. During the first sampling event in May 2022, the \u0026beta;sor values were similar between the Mizunashi River and the Oka River, despite considerable variation in the filtration volumes among the Oka River samples. In September 2022, when a complete 1-L volume was filtered in both the Oka River and the Tagami River, the \u0026beta;sor values were almost identical. In November 2022, the \u0026beta;sor in the Tagami River was much higher than that in the Mizunashi River owing to insufficient filtration volume.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study successfully achieved the simultaneous detection of aquatic and terrestrial species by conducting surveys after rainfall events and evaluated the differences in detection patterns in relation to rainfall and watershed characteristics. The results suggested that rainfall influenced the detectability of different taxonomic groups by altering multiple processes such as eDNA input, dilution, and resuspension. Although rainfall events could reduce the overall detection efficiency, particularly for fish, they could also enhance the detectability of terrestrial and semi-aquatic species. Detection outcomes were further shaped by factors such as filtration volume and species-specific ecological traits. Therefore, optimizing the sampling design based on the ecological characteristics of the target species is crucial for improving the detection efficiency in species-specific eDNA studies. In addition, our findings highlight the importance of implementing multi-timepoint sampling strategies across varying hydrological conditions and climatic events to achieve a more comprehensive assessment of watershed biodiversity.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cstrong\u003eStudy area\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study area is located in Miyazaki Prefecture, southwestern Japan, and belongs to the tributaries of Kiyotake River (i.e., the Mizunashi River, Oka River, and Tagami River; Figure 5-\u003cstrong\u003ea\u003c/strong\u003e). The Kiyotake River is 28.8 km long, and its catchment covers an area of 166.40 km\u003csup\u003e2\u003c/sup\u003e. The catchment areas of the Mizunashi River, Oka River, and Tagami River are 16.70 km\u0026sup2;, 16.90 km\u0026sup2;, and 3.30 km\u0026sup2;, respectively (catchment area data available: https://data.stat.pref.miyazaki.lg.jp/dataset/nenkan-26). The climate in the catchment is characterized by an annual average temperature of approximately 18.00\u0026deg;C and annual precipitation of approximately 3,000 mm. We obtained data for 12 different land use types through the Advanced Land Observing Satellite (10-m accuracy, data available: https://earth.jaxa.jp/ja/data). Fig. 5-\u003cstrong\u003eb\u003c/strong\u003e shows the land uses in the three catchments. The major land uses in the studied rivers were evergreen needleleaf trees, in the Mizunashi River (51.80%); evergreen needleleaf trees and fields, in the Oka River (28.20% and 30.20%, respectively); and fields, in the Tagami River (39.50%). The geographic coordinates of the Mizunashi River, Oka River, and Tagami River are 31.829925\u0026deg;N, 131.362669\u0026deg;E; 31.85162\u0026deg;N, 131.386959\u0026deg;E; and 31.838143\u0026deg;N, 131.417732\u0026deg;E, respectively.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSampling\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;and eDNA extraction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll water samples were collected from publicly accessible rivers, and no permits were required for sampling. The sampling events included rainfall and non-rainfall events. Sampling for the rainfall events were conducted on May 12, September 14, and November 29, 2022, as well as on June 2, 2023. To understand the temporal pattern of eDNA detection, sampling was conducted twice daily (sampling 1 and 2) on May 12, 2022, and June 2, 2023. The time interval between sampling events was approximately 2 h. Sampling for the non-rainfall events occurred on November 28, 2022, and August 8, 2024. Before water sampling, 1-L and 0.25-L plastic bottles, tethered buckets, and cooler boxes were sterilized using 10% bleach for at least 30 min to prevent contamination by DNA and microorganisms other than the target species.\u0026nbsp;During rainfall events, two 1-L replicates of river water were collected, whereas during non-rainfall events, only a single 1-L sample was collected. Another 0.25 L of river water was collected\u0026nbsp;for basic measurements of environmental parameters,\u0026nbsp;namely, pH, electrical conductivity (EC), and turbidity, during all events.\u0026nbsp;For the\u0026nbsp;0.25-L samples,\u0026nbsp;a bar-shaped mercury thermometer was used to measure\u0026nbsp;water temperature at\u0026nbsp;the sampling sites, EC and pH were measured using a portable\u0026nbsp;multiwater quality meter (HORIBA TOA DKK), and turbidity was measured using a turbidity meter (PT-200_TROAM, Asahi Science) in a laboratory.\u0026nbsp;The environmental parameters and EC, in particular, were\u0026nbsp;not measured on\u0026nbsp;November\u0026nbsp;28, 2022, and\u0026nbsp;June 2,\u0026nbsp;2023, respectively, owing\u0026nbsp;to technical problems. Rainfall data\u0026nbsp;were obtained from the nearest meteorological station (Akae Observatory, data available: https://www.data.jma.go.jp/obd/stats/etrn/index).\u003c/p\u003e\n\u003cp\u003eDetailed methods for water filtration and eDNA extraction are available in our previous studies\u003csup\u003e39\u003c/sup\u003e. In summary, water samples were filtered using 0.7-\u0026micro;m glass fiber filters. When the filter paper became clogged, it was replaced with a new one. If the second filter also became clogged, the remaining water was discarded. After filtration, all filter papers were frozen at \u0026minus;20 \u0026deg;C until DNA extraction. eDNA was extracted using the DNeasy PowerSoil Kit (Qiagen, Hilden, Germany), following the manufacturer\u0026rsquo;s protocol, after which it was eluted in 100-\u0026micro;L volumes. After extraction, the DNA template solutions were stored at \u0026minus;80 \u0026deg;C for subsequent polymerase chain reaction (PCR) amplification. Samples collected on August 8, 2024, were extracted using the DNeasy PowerSoil Pro Kit, the successor of the DNeasy PowerSoil Kit, since the latter had been discontinued.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eeDNA\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003emetabarcoding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUsing the DNeasy PowerClean Pro Cleanup Kit (Qiagen, Hilden, Germany), the template was purified to remove the PCR inhibitors. The concentration of the DNA solution was determined using the Synergy LX (Agilent Technologies) in combination with the QuantiFluor dsDNA System (Promega). Subsequently, vertebrate DNA was comprehensively detected via high-throughput sequencing using a two-step tailed PCR approach targeting the mitochondrial 12S rRNA gene region. The forward primer YAGAACAGGCTCCTCTAG and reverse primer TTAGATACCCCACTATGC were used, both of which are suitable for vertebrate detection \u003csup\u003e63\u003c/sup\u003e. The 12S rRNA region was selected owing to its low species-level primer bias, high specificity, and reproducibility \u003csup\u003e64\u003c/sup\u003e. Moreover, previous studies have reported that replacing the initial thymine (T) in the forward primer with pyrimidine (Y) enables broader binding across vertebrate taxa\u0026nbsp;(Kocher et al., 2017).\u0026nbsp;The first-round PCR (1\u003csup\u003est\u003c/sup\u003e PCR) was performed in eight replicates, with each 10-\u0026mu;L PCR mixture containing 1.0 \u0026mu;L of 10\u0026times; Ex Buffer, 0.8 \u0026mu;L of deoxynucleotide triphosphates (dNTPs, 2.5 mM each), 0.5 \u0026mu;L each of 5-\u0026mu;M forward (5\u0026prime;-ACACTCTTTCCCTACACGACGCTCTTCCGATCT YAGAACAGGCTCCTCTAG-3\u0026prime;) and reverse (5\u0026prime;-GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT TTAGATACCCCACTATGC-3\u0026prime;) primer, 2.0 \u0026mu;L of template DNA, 0.5 \u0026mu;L each of 10-\u0026mu;M forward and reverse primer, 0.075 \u0026mu;L of ExTaq HS (5 U/\u0026mu;L; TaKaRa), and 5.125 \u0026mu;L of nuclease-free water. The thermal cycling conditions were as follows: initial denaturation at 94\u0026deg;C for 2 min; 30 or 35 cycles at 94\u0026deg;C for 30 s, 50\u0026deg;C for 30 s, and 72\u0026deg;C for 30 s; and a final extension at 72\u0026deg;C for 5 min. The PCR products were purified using VAHTS DNA Clean Beads (Vazyme) at a 1.5\u0026times; ratio to the reaction volume. The purified products were used as templates for the second round of PCR. A negative control was prepared using nuclease-free water as the template. The 2\u003csup\u003end\u003c/sup\u003e PCR process was performed as a single reaction to attach adapter sequences. Each 10-\u0026mu;L reaction mixture included 5.0 \u0026mu;L of 2\u0026times; PCR Buffer for KOD FX Neo, 2.0 \u0026mu;L of dNTPs (2 mM each), 0.5 \u0026mu;L each of 5-\u0026mu;M forward (5\u0026prime;-AATGATACGGCGACCACCGAGATCTACACXXXXXXXXACTCTTTCCCTACACGACGC-3\u0026prime;) and reverse (5\u0026prime;-CAAGCAGAAGACGGCATACGAGATXXXXXXXXGTGACTGGAGTTCAGACGTGTG-3\u0026prime;) primer, 1.0 \u0026mu;L of pooled 1\u003csup\u003est\u003c/sup\u003e-PCR products, 0.2 \u0026mu;L of KOD FX Neo (1.0 U/\u0026mu;L; TOYOBO), and 0.8 \u0026mu;L of nuclease-free water. The PCR program was as follows: 94\u0026deg;C for 2 min; 10 or 12 cycles at 98\u0026deg;C for 30 s, 60\u0026deg;C for 30 s, and 68\u0026deg;C for 30 s; and 68\u0026deg;C for 5 min. The 2\u003csup\u003end\u003c/sup\u003e PCR products were also purified using VAHTS DNA Clean Beads (1.5\u0026times; volume ratio), completing library preparation. The concentration of the final libraries was measured using Synergy H1 (Agilent Technologies) with the QuantiFluor dsDNA System, and the library quality was assessed using a Fragment Analyzer with the dsDNA 915 Reagent Kit (Agilent Technologies). Finally, sequencing was conducted using the MiSeq system with the MiSeq Reagent Kit v3 (Illumina), under 2\u0026times;300-bp paired-end conditions.\u003c/p\u003e\n\u003cp\u003eReads obtained using the fastx barcode splitter tool of the FASTX-Toolkit (ver. 0.0.14) were filtered to retain only those that perfectly matched the primer sequences used at the start of the reads. Using the fastx_trimmer tool of FASTX-Toolkit, the primer sequences were removed from the extracted reads. Subsequently, low-quality sequences with quality scores below 20 were discarded using Sickle (ver. 1.33), and sequences shorter than 40 bases, along with their paired reads, were removed. Paired-end reads were merged using the FLASH script (ver. 1.2.11), with a minimum overlap of 10 bases. For species identification, the dada2 plugin in Qiime2 (ver. 2024.2) was used to remove chimeric and noisy sequences. Representative sequences and an ASV table were then generated, and the representative sequences were subjected to a Basic Local Alignment Search Tool (BLAST) process against the nucleotide database (National Center for Biotechnology Information) to identify species. The identification criterion was a coverage \u0026gt;90%, and only the result with the highest identity was used. The identity thresholds for species-level identification were set as follows: \u0026gt;98% for species, \u0026gt;95% for genera, \u0026gt;90% for families, and \u0026gt;85% for orders \u003csup\u003e65\u003c/sup\u003e. When the identity index was the same, the optimal result was selected as the reference index, based on the E-value and coverage, in that order. In cases where there was incongruence between BLAST hits, the ASV(s) would be assigned to a higher rank (i.e., genus or class) to provide a consensus assignment \u003csup\u003e66\u003c/sup\u003e. We applied a cut-off based on a fixed-read-counts threshold; when an ASV was present in a sample with the number of reads \u0026lt;100, it was removed from the sample \u003csup\u003e67,68\u003c/sup\u003e. Since the samples were collected from freshwater rivers, based on the fish database (https://www.fishbase.se), fish species that exclusively inhabited marine environments were excluded from the analysis with human ASVs as environmental contaminants. One ASV was assigned to \u003cem\u003eHomo sapiens\u003c/em\u003e although it may have originated from \u003cem\u003eMus musculus\u003c/em\u003e. Although \u003cem\u003eMus\u0026nbsp;\u003c/em\u003eis often regarded as a laboratory contaminant \u003csup\u003e69\u003c/sup\u003e, its widespread presence in natural environments warranted the retention of this ASV in the analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRainfall and river discharge\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRainfall data for the sampling day were obtained from the Akae Meteorological Station. Given the relatively small sizes of the three watersheds, we used the cumulative rainfall during the 180 min immediately preceding the start of sampling as an approximate measure of rainfall. River discharge after rainfall was estimated using a distributed hydrological model \u003csup\u003e70\u003c/sup\u003e, which divided the watershed into grid cells and incorporated watershed characteristics to accurately reproduce the discharge from upstream to downstream. For sampling sites where discharge could not be directly calculated, discharge data from the nearest grid cells were used. In the case of the Tagami River, a direct estimation of discharge was not possible; therefore, discharge was calculated for the Kiyotake River and subsequently converted using the watershed area ratio between the Tagami and Kiyotake basins. These parameters, together with the water quality variables, are listed in Table 1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDetection rate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the impact of rainfall occurrence on species detection, we calculated the overall detection rate for each sample, as well as the detection rates for individual taxonomic groups. Because a comprehensive species inventory for the study area is currently lacking, the total number of species detected across all samples was used as a proxy for the regional species pool. Based on this, the detection rate of a single sample within a sub-watershed was defined as the proportion of species detected in that sample relative to the total number of species detected within the same sub-watershed. Meanwhile, the detection rate for each taxonomic group was defined as the proportion of species detected within that group in the sample relative to the total number of species detected in the corresponding group across the sub-watershed.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis study spanned a full year and encompassed all four seasons, making it possible that seasonal variations may confound the effects of rainfall on species detection. Birds exhibit strong seasonal migratory behavior, and their detection frequency may be more influenced by seasonal patterns than by rainfall. Moreover, because of the limited taxonomic resolution of the primers used in this study, most bird sequences could not be reliably identified to the species level, which made it difficult to determine whether the detected taxa were migratory. To minimize the influence of seasonal factors and enhance the robustness of the analysis, birds were excluded from the calculations of the detection rate and community structure. Although fish may migrate for spawning, they remain in freshwater rivers during certain life stages; therefore, seasonal variations did not preclude their detection.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRelationship between detection number and environmental factors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGeneralized linear models (GLMs) with a Poisson distribution were used to assess the relationships between the number of species detected in the samples (detection numbers) and environmental factors, including turbidity, pH, water temperature, filtration volume, and discharge. Because EC was not recorded in June 2023, it was excluded from the model to maintain a sufficient sample size. The cumulative rainfall during the 3 h prior to sampling was used as an explanatory variable. Variance inflation factors (VIFs) were calculated using the R \u0026ldquo;car\u0026rdquo; package to verify multicollinearity. Variables with VIFs \u0026gt;5 were iteratively removed until all variables had VIF values less than 5. Turbidity was excluded during this process, and final models were constructed using the R \u0026ldquo;glmmTMB\u0026rdquo; package \u003csup\u003e71\u003c/sup\u003e. All analyses were conducted using R (ver. 4.3.2, R Core Team, 2023).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCommunity structure analyses\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the effect of rainfall conditions on community composition, we calculated \u0026beta;-diversity based on species presence/absence data, using the S\u0026oslash;rensen dissimilarity index\u0026nbsp;\u003csup\u003e72\u003c/sup\u003e. Specifically, used the betapart package to compute total dissimilarity (\u0026beta;sor), turnover (\u0026beta;sim), and nestedness (\u0026beta;sne), where \u0026beta;sor = \u0026beta;sim + \u0026beta;sne. Heatmaps were generated using the Origin (2025) software (OriginLab, USA) to visualize the diversity patterns.\u0026nbsp;To investigate the influence of environmental variables on the directional variation in community structure, principal coordinate analysis (PCoA) was conducted based on the S\u0026oslash;rensen dissimilarity index. Using the \u0026ldquo;envfit()\u0026rdquo; function (permutations = 999,) in the vegan package, with a fixed random seed to ensure reproducibility, we fitted the three environmental variables that were fully recorded (i.e., discharge, filtration volume, and rainfall) onto the ordination plots, and only those with significant effects were displayed, thus allowing for a clear visualization of the direction and strength of their influence on community composition.\u003c/p\u003e\n\u003cp\u003eTo further assess the effects of rainfall occurrence (categorical variable: presence/absence) and filtration volume (continuous variable) on community dissimilarity (\u0026beta;sor, \u0026beta;sim, and \u0026beta;sne), we used Permutational Multivariate Analysis of Variance (PERMANOVA). This analysis was performed using 999 permutations with the \u0026ldquo;permute\u0026rdquo; and \u0026ldquo;vegan\u0026rdquo; packages, a fixed random seed was set to ensure reproducibility. The analysis was implemented as a marginal-effects model (i.e., by = \u0026quot;margin\u0026quot; in the \u0026ldquo;adonis2()\u0026rdquo; function), enabling the independent evaluation of each explanatory variable. As \u0026beta;sor is a non-Euclidean distance measure, it was square-root-transformed prior to analysis; \u0026beta;sim and \u0026beta;sne were adjusted using the Lingoes correction to meet the assumptions of PERMANOVA. All analyses were conducted using R (ver. 4.3.2, R Core Team, 2023).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to thank Mr. Yamaguchi, Mr. Inoue, Mr. Higuchi, and Mr. Nanri (University of Miyazaki) for their help with the experiments and field observations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding Declaration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was partially supported by the Ministry of Education, Science, Sports and Culture through a Grant-in-Aid for Scientific Research (grant numbers 24H00329, 25K00041, and 25H01201).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCX performed the experiment, analyzed data, and led the writing of manuscript. KN conceived the ideas, supervised the research and acquired the funds. KN and CX performed the sampling and edited and approved the manuscript. Next-generation sequencing was performed by Bioengineering Lab through external outsourcing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Accessibility and Benefit-Sharing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData accessibility\u003c/p\u003e\n\u003cp\u003eRaw reads generated during the current study are available in the DDBJ Sequence Read Archive (DRA) under the accession numbers DRA023808.\u003c/p\u003e\n\u003cp\u003eThe R code used for data analysis can be found in Supplementary files.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBenefit sharing\u003c/p\u003e\n\u003cp\u003eThis study did not involve direct contact with animals in the survey area, particularly wild or endangered species. We are committed to investigating and detecting biodiversity without direct interaction with natural fauna.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declared no conflict of interest in this article.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eCardinale, B. J. \u003cem\u003eet al.\u003c/em\u003e Biodiversity loss and its impact on humanity. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e486\u003c/strong\u003e, 59\u0026ndash;67 (2012).\u003c/li\u003e\n\u003cli\u003eFr\u0026oslash;slev, T. G. \u003cem\u003eet al.\u003c/em\u003e The biodiversity effect of reduced tillage on soil microbiota. \u003cem\u003eAmbio\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, 1022\u0026ndash;1033 (2022).\u003c/li\u003e\n\u003cli\u003eRedondo, M. A., Berlin, A., Boberg, J. \u0026amp; Oliva, J. Vegetation type determines spore deposition within a forest\u0026ndash;agricultural mosaic landscape. \u003cem\u003eFEMS Microbiology Ecology\u003c/em\u003e \u003cstrong\u003e96\u003c/strong\u003e, fiaa082 (2020).\u003c/li\u003e\n\u003cli\u003eYang, J., Zhang, L., Mu, Y. \u0026amp; Zhang, X. Small changes make big progress: A more efficient eDNA monitoring method for freshwater fish. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 363\u0026ndash;374 (2023).\u003c/li\u003e\n\u003cli\u003eDyson, K. \u003cem\u003eet al.\u003c/em\u003e Coupling remote sensing and eDNA to monitor environmental impact: A pilot to quantify the environmental benefits of sustainable agriculture in the Brazilian Amazon. \u003cem\u003ePLoS ONE\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, e0289437 (2024).\u003c/li\u003e\n\u003cli\u003eFediajevaite, J., Priestley, V., Arnold, R. \u0026amp; Savolainen, V. Meta‐analysis shows that environmental DNA outperforms traditional surveys, but warrants better reporting standards. \u003cem\u003eEcology and Evolution\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 4803\u0026ndash;4815 (2021).\u003c/li\u003e\n\u003cli\u003eSuren, A. M., Burdon, F. J. \u0026amp; Wilkinson, S. P. eDNA Is a Useful Environmental Monitoring Tool for Assessing Stream Ecological Health. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, e596 (2024).\u003c/li\u003e\n\u003cli\u003eAfzali, S. F. \u003cem\u003eet al.\u003c/em\u003e Comparing environmental metabarcoding and trawling survey of demersal fish communities in the Gulf of St. Lawrence, Canada. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 22\u0026ndash;42 (2021).\u003c/li\u003e\n\u003cli\u003eTaberlet, P., Coissac, E., Pompanon, F., Brochmann, C. \u0026amp; Willerslev, E. Towards next‐generation biodiversity assessment using DNA metabarcoding. \u003cem\u003eMolecular Ecology\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 2045\u0026ndash;2050 (2012).\u003c/li\u003e\n\u003cli\u003eValentini, A. \u003cem\u003eet al.\u003c/em\u003e Next‐generation monitoring of aquatic biodiversity using environmental DNA metabarcoding. \u003cem\u003eMolecular Ecology\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 929\u0026ndash;942 (2016).\u003c/li\u003e\n\u003cli\u003eHuang, S., Yoshitake, K., Watabe, S. \u0026amp; Asakawa, S. Environmental DNA study on aquatic ecosystem monitoring and management: Recent advances and prospects. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e \u003cstrong\u003e323\u003c/strong\u003e, 116310 (2022).\u003c/li\u003e\n\u003cli\u003eRuppert, K. M., Kline, R. J. \u0026amp; Rahman, M. S. Past, present, and future perspectives of environmental DNA (eDNA) metabarcoding: A systematic review in methods, monitoring, and applications of global eDNA. \u003cem\u003eGlobal Ecology and Conservation\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, e00547 (2019).\u003c/li\u003e\n\u003cli\u003eSahu, A., Kumar, N., Pal Singh, C. \u0026amp; Singh, M. Environmental DNA (eDNA): Powerful technique for biodiversity conservation. \u003cem\u003eJournal for Nature Conservation\u003c/em\u003e \u003cstrong\u003e71\u003c/strong\u003e, 126325 (2023).\u003c/li\u003e\n\u003cli\u003eTakahashi, M. \u003cem\u003eet al.\u003c/em\u003e Aquatic environmental DNA: A review of the macro-organismal biomonitoring revolution. \u003cem\u003eScience of The Total Environment\u003c/em\u003e \u003cstrong\u003e873\u003c/strong\u003e, 162322 (2023).\u003c/li\u003e\n\u003cli\u003eMeyer, J. M., Leempoel, K., Losapio, G. \u0026amp; Hadly, E. A. Molecular Ecological Network Analyses: An Effective Conservation Tool for the Assessment of Biodiversity, Trophic Interactions, and Community Structure. \u003cem\u003eFront. Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 588430 (2020).\u003c/li\u003e\n\u003cli\u003eThomsen, P. F. \u0026amp; Sigsgaard, E. E. Environmental DNA metabarcoding of wild flowers reveals diverse communities of terrestrial arthropods. \u003cem\u003eEcology and Evolution\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1665\u0026ndash;1679 (2019).\u003c/li\u003e\n\u003cli\u003eGogarten, J. F. \u003cem\u003eet al.\u003c/em\u003e Fly‐derived DNA and camera traps are complementary tools for assessing mammalian biodiversity. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 63\u0026ndash;76 (2020).\u003c/li\u003e\n\u003cli\u003eSeeber, P. A. \u0026amp; Epp, L. S. Environmental DNA and metagenomics of terrestrial mammals as keystone taxa of recent and past ecosystems. \u003cem\u003eMammal Review\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, 538\u0026ndash;553 (2022).\u003c/li\u003e\n\u003cli\u003eNewton, J. P., Allentoft, M. E., Bateman, P. W., Van Der Heyde, M. \u0026amp; Nevill, P. Targeting Terrestrial Vertebrates With eDNA : Trends, Perspectives, and Considerations for Sampling. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, e70056 (2025).\u003c/li\u003e\n\u003cli\u003eGuthrie, A. M. \u003cem\u003eet al.\u003c/em\u003e A quantitative analysis of vertebrate environmental DNA degradation in soil in response to time, UV light, and temperature. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, e581 (2024).\u003c/li\u003e\n\u003cli\u003eHassan, S. \u003cem\u003eet al.\u003c/em\u003e Environmental DNA Metabarcoding: A Novel Contrivance for Documenting Terrestrial Biodiversity. \u003cem\u003eBiology\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1297 (2022).\u003c/li\u003e\n\u003cli\u003eRodgers, T. W. \u0026amp; Mock, K. E. Drinking water as a source of environmental DNA for the detection of terrestrial wildlife species. \u003cem\u003eConservation Genet Resour\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 693\u0026ndash;696 (2015).\u003c/li\u003e\n\u003cli\u003eUshio, M. \u003cem\u003eet al.\u003c/em\u003e Environmental DNA enables detection of terrestrial mammals from forest pond water. \u003cem\u003eMolecular Ecology Resources\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, e63-75 (2017).\u003c/li\u003e\n\u003cli\u003eMartinelli Mar\u0026iacute;n, D., Lasso, C. A. \u0026amp; Caballero Gaitan, S. J. eDNA metabarcoding: an effective tool for vertebrate diversity studies in the Colombian Amazon and Orinoco basins. \u003cem\u003eFront. Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1409296 (2024).\u003c/li\u003e\n\u003cli\u003eReji Chacko, M. \u003cem\u003eet al.\u003c/em\u003e Catchment-based sampling of river eDNA integrates terrestrial and aquatic biodiversity of alpine landscapes. \u003cem\u003eOecologia\u003c/em\u003e \u003cstrong\u003e202\u003c/strong\u003e, 699\u0026ndash;713 (2023).\u003c/li\u003e\n\u003cli\u003eSaenz‐Agudelo, P. \u003cem\u003eet al.\u003c/em\u003e Monitoring vertebrate biodiversity of a protected coastal wetland using eDNA metabarcoding. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 77\u0026ndash;92 (2022).\u003c/li\u003e\n\u003cli\u003eCoutant, O. \u003cem\u003eet al.\u003c/em\u003e Amazonian mammal monitoring using aquatic environmental DNA. \u003cem\u003eMolecular Ecology Resources\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 1875\u0026ndash;1888 (2021).\u003c/li\u003e\n\u003cli\u003eMacher, T.-H. \u003cem\u003eet al.\u003c/em\u003e Beyond fish eDNA metabarcoding: Field replicates disproportionately improve the detection of stream associated vertebrate species. \u003cem\u003eMBMG\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, e66557 (2021).\u003c/li\u003e\n\u003cli\u003eRyan, E., Bateman, P., Fernandes, K., Van Der Heyde, M. \u0026amp; Nevill, P. eDNA metabarcoding of log hollow sediments and soils highlights the importance of substrate type, frequency of sampling and animal size, for vertebrate species detection. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 940\u0026ndash;953 (2022).\u003c/li\u003e\n\u003cli\u003eZhang, S., Zhao, J. \u0026amp; Yao, M. Urban landscape-level biodiversity assessments of aquatic and terrestrial vertebrates by environmental DNA metabarcoding. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e \u003cstrong\u003e340\u003c/strong\u003e, 117971 (2023).\u003c/li\u003e\n\u003cli\u003eDe Souza, L. S., Godwin, J. C., Renshaw, M. A. \u0026amp; Larson, E. Environmental DNA (eDNA) Detection Probability Is Influenced by Seasonal Activity of Organisms. \u003cem\u003ePLoS ONE\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, e0165273 (2016).\u003c/li\u003e\n\u003cli\u003eHarper, L. R. \u003cem\u003eet al.\u003c/em\u003e Environmental DNA (eDNA) metabarcoding of pond water as a tool to survey conservation and management priority mammals. \u003cem\u003eBiological Conservation\u003c/em\u003e \u003cstrong\u003e238\u003c/strong\u003e, 108225 (2019).\u003c/li\u003e\n\u003cli\u003eMauvisseau, Q. \u003cem\u003eet al.\u003c/em\u003e The Multiple States of Environmental DNA and What Is Known about Their Persistence in Aquatic Environments. \u003cem\u003eEnviron. Sci. Technol.\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, 5322\u0026ndash;5333 (2022).\u003c/li\u003e\n\u003cli\u003eNukazawa, K., Hamasuna, Y. \u0026amp; Suzuki, Y. Simulating the Advection and Degradation of the Environmental DNA of Common Carp along a River. \u003cem\u003eEnviron. Sci. Technol.\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, 10562\u0026ndash;10570 (2018).\u003c/li\u003e\n\u003cli\u003eSeymour, M. \u003cem\u003eet al.\u003c/em\u003e Acidity promotes degradation of multi-species environmental DNA in lotic mesocosms. \u003cem\u003eCommun Biol\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, 4 (2018).\u003c/li\u003e\n\u003cli\u003eStrickler, K. M., Fremier, A. K. \u0026amp; Goldberg, C. S. Quantifying effects of UV-B, temperature, and pH on eDNA degradation in aquatic microcosms. \u003cem\u003eBiological Conservation\u003c/em\u003e \u003cstrong\u003e183\u003c/strong\u003e, 85\u0026ndash;92 (2015).\u003c/li\u003e\n\u003cli\u003eCowgill, C., Gilbert, J. D. J., Convery, I. \u0026amp; Lawson Handley, L. Monitoring terrestrial rewilding with environmental DNA metabarcoding: a systematic review of current trends and recommendations. \u003cem\u003eFront. Conserv. Sci.\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 1473957 (2025).\u003c/li\u003e\n\u003cli\u003eDeiner, K., Fronhofer, E. A., M\u0026auml;chler, E., Walser, J.-C. \u0026amp; Altermatt, F. Environmental DNA reveals that rivers are conveyer belts of biodiversity information. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 12544 (2016).\u003c/li\u003e\n\u003cli\u003eXu, C. \u0026amp; Nukazawa, K. Detection of terrestrial species using environmental DNA during heavy rainfall events and associated influencing factors. Preprint at https://doi.org/10.22541/au.174312906.60655689/v1 (2025).\u003c/li\u003e\n\u003cli\u003eLadin, Z. S. \u003cem\u003eet al.\u003c/em\u003e Assessing the efficacy of eDNA metabarcoding for measuring microbial biodiversity within forest ecosystems. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1629 (2021).\u003c/li\u003e\n\u003cli\u003eMacher, T., Sch\u0026uuml;tz, R., H\u0026ouml;rren, T., Beermann, A. J. \u0026amp; Leese, F. It\u0026rsquo;s raining species: Rainwash eDNA metabarcoding as a minimally invasive method to assess tree canopy invertebrate diversity. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 3\u0026ndash;11 (2023).\u003c/li\u003e\n\u003cli\u003eBae, M.-J. \u003cem\u003eet al.\u003c/em\u003e Seasonal Dynamics and Heavy Rain Effects on the Diversity of Microeukaryome in the Nakdonggang River. \u003cem\u003eWater\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 3407 (2022).\u003c/li\u003e\n\u003cli\u003eStaley, Z. R. \u003cem\u003eet al.\u003c/em\u003e Fecal source tracking and eDNA profiling in an urban creek following an extreme rain event. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 14390 (2018).\u003c/li\u003e\n\u003cli\u003eCondachou, C. \u003cem\u003eet al.\u003c/em\u003e Tropical Rainfall and eDNA Washout Impact Estimations of Amazonian Biodiversity Patterns From Environmental Samples. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, e70141 (2025).\u003c/li\u003e\n\u003cli\u003eOsathanunkul, M. \u0026amp; Suwannapoom, C. A comparative study on eDNA-based detection of Siamese bat catfish (Oreoglanis siamensis) in wet and dry conditions. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 8885 (2024).\u003c/li\u003e\n\u003cli\u003eHarper, L. R. \u003cem\u003eet al.\u003c/em\u003e Generating and testing ecological hypotheses at the pondscape with environmental DNA metabarcoding: A case study on a threatened amphibian. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 184\u0026ndash;199 (2020).\u003c/li\u003e\n\u003cli\u003eUrabe, H., Mizumoto, H., Tsuda-Yamaguchi, F. \u0026amp; Araki, H. Spatial heterogeneity of eDNA concentration as a predictor of small biomass of fish in a mountain stream. \u003cem\u003eLimnology\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 223\u0026ndash;233 (2025).\u003c/li\u003e\n\u003cli\u003eHunter, M. E., Ferrante, J. A., Meigs-Friend, G. \u0026amp; Ulmer, A. Improving eDNA yield and inhibitor reduction through increased water volumes and multi-filter isolation techniques. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 5259 (2019).\u003c/li\u003e\n\u003cli\u003eSepulveda, A. J. \u003cem\u003eet al.\u003c/em\u003e Improved detection of rare, endangered and invasive trout in using a new large‐volume sampling method for eDNA capture. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, 227\u0026ndash;237 (2019).\u003c/li\u003e\n\u003cli\u003eMelendy, S. A. \u0026amp; Olson, J. R. Identifying Key Environmental Drivers of Reach‐Scale Salmonid eDNA Recovery With Random Forest. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, e70001 (2024).\u003c/li\u003e\n\u003cli\u003eShogren, A. J. \u003cem\u003eet al.\u003c/em\u003e Controls on eDNA movement in streams: Transport, Retention, and Resuspension. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 5065 (2017).\u003c/li\u003e\n\u003cli\u003eSakata, M. K. \u003cem\u003eet al.\u003c/em\u003e Determining an effective sampling method for eDNA metabarcoding: a case study for fish biodiversity monitoring in a small, natural river. \u003cem\u003eLimnology\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 221\u0026ndash;235 (2021).\u003c/li\u003e\n\u003cli\u003eCurtis, A. N., Tiemann, J. S., Douglass, S. A., Davis, M. A. \u0026amp; Larson, E. R. High stream flows dilute environmental DNA (eDNA) concentrations and reduce detectability. \u003cem\u003eDiversity and Distributions\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1918\u0026ndash;1931 (2021).\u003c/li\u003e\n\u003cli\u003eRobinson, P. S., Davis, D. R., Collins, S. M. \u0026amp; Kline, R. J. Defining the current distribution of the imperiled Black-spotted Newt across south Texas, USA. \u003cem\u003eGlobal Ecology and Conservation\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, e02131 (2022).\u003c/li\u003e\n\u003cli\u003eLynggaard, C., Fr\u0026oslash;slev, T. G., Johnson, M. S., Olsen, M. T. \u0026amp; Bohmann, K. Airborne environmental DNA captures terrestrial vertebrate diversity in nature. \u003cem\u003eMolecular Ecology Resources\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, e13840 (2024).\u003c/li\u003e\n\u003cli\u003eRagot, R. \u0026amp; Villemur, R. eDNA profiling of mammals, birds, and fish of surface waters by mitochondrial metagenomics: application for source tracking of fecal contamination in surface waters. \u003cem\u003eEnviron Monit Assess\u003c/em\u003e \u003cstrong\u003e194\u003c/strong\u003e, 72 (2022).\u003c/li\u003e\n\u003cli\u003eBroadhurst, H. A. \u003cem\u003eet al.\u003c/em\u003e From Water to Land: A Review on the Applications of Environmental DNA and Invertebrate‐Derived DNA for Monitoring Terrestrial and Semi‐Aquatic Mammals. \u003cem\u003eMammal Review\u003c/em\u003e e70006 (2025) doi:10.1111/mam.70006.\u003c/li\u003e\n\u003cli\u003eRuan, H.-T. \u003cem\u003eet al.\u003c/em\u003e Effects of sampling strategies and DNA extraction methods on eDNA metabarcoding: A case study of estuarine fish diversity monitoring. \u003cem\u003eZoological Research\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, 192\u0026ndash;204 (2022).\u003c/li\u003e\n\u003cli\u003eWang, C. \u003cem\u003eet al.\u003c/em\u003e Effects of Filtration Volumes on Bacterial Diversity and Community Structure in Freshwater Lakes. \u003cem\u003eEcology and Evolution\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, e71445 (2025).\u003c/li\u003e\n\u003cli\u003eChase, J. M., Kraft, N. J. B., Smith, K. G., Vellend, M. \u0026amp; Inouye, B. D. Using null models to disentangle variation in community dissimilarity from variation in \u0026alpha;-diversity. \u003cem\u003eEcosphere\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, art24 (2011).\u003c/li\u003e\n\u003cli\u003eKawakami, T. \u003cem\u003eet al.\u003c/em\u003e Evaluating the sampling effort for the metabarcoding‐based detection of fish environmental DNA in the open ocean. \u003cem\u003eEcology and Evolution\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, e9921 (2023).\u003c/li\u003e\n\u003cli\u003eBaudry, T., Vasselon, V., Delaunay, C., Sweet, M. \u0026amp; Grandjean, F. When Methodological Innovation Changes the Game: A 10‐Year Review of Environmental DNA (eDNA) Applied to Crayfish. \u003cem\u003eAquatic Conservation\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, e4245 (2024).\u003c/li\u003e\n\u003cli\u003eRiaz, T. \u003cem\u003eet al.\u003c/em\u003e ecoPrimers: inference of new DNA barcode markers from whole genome sequence analysis. \u003cem\u003eNucleic Acids Research\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, e145\u0026ndash;e145 (2011).\u003c/li\u003e\n\u003cli\u003eCollins, R. A. \u003cem\u003eet al.\u003c/em\u003e Non‐specific amplification compromises environmental DNA metabarcoding with COI. \u003cem\u003eMethods Ecol Evol\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 1985\u0026ndash;2001 (2019).\u003c/li\u003e\n\u003cli\u003eVeron, P. \u003cem\u003eet al.\u003c/em\u003e Environmental DNA complements scientific trawling in surveys of marine fish biodiversity. \u003cem\u003eICES Journal of Marine Science\u003c/em\u003e \u003cstrong\u003e80\u003c/strong\u003e, 2150\u0026ndash;2165 (2023).\u003c/li\u003e\n\u003cli\u003eChung, A., Kam, Y. C., Shea, S. K. H. \u0026amp; Schunter, C. Detecting Fish Diversity in Urban‐Impacted Ecosystems: A Comparative Approach of eDNA Metabarcoding and UVC. \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, e70048 (2024).\u003c/li\u003e\n\u003cli\u003eHolman, L. E., Chng, Y. \u0026amp; Rius, M. How does eDNA decay affect metabarcoding experiments? \u003cem\u003eEnvironmental DNA\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 108\u0026ndash;116 (2022).\u003c/li\u003e\n\u003cli\u003eMarinchel, N. \u003cem\u003eet al.\u003c/em\u003e Mock community experiments can inform on the reliability of eDNA metabarcoding data: a case study on marine phytoplankton. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 20164 (2023).\u003c/li\u003e\n\u003cli\u003eAllen, M. C. \u003cem\u003eet al.\u003c/em\u003e Sampling environmental DNA from trees and soil to detect cryptic arboreal mammals. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 180 (2023).\u003c/li\u003e\n\u003cli\u003eMineda, H., Nukazawa, K. \u0026amp; Suzuki, Y. Assessing environmental flow alterations induced by dams and climate change using a distributed hydrological model at catchment scale. \u003cem\u003eJournal of Water and Climate Change\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 4821\u0026ndash;4836 (2023).\u003c/li\u003e\n\u003cli\u003eBrooks, M., E. \u003cem\u003eet al.\u003c/em\u003e glmmTMB Balances Speed and Flexibility Among Packages for Zero-inflated Generalized Linear Mixed Modeling. \u003cem\u003eThe R Journal\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 378 (2017).\u003c/li\u003e\n\u003cli\u003eBaselga, A. Partitioning the turnover and nestedness components of beta diversity. \u003cem\u003eGlobal Ecology and Biogeography\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 134\u0026ndash;143 (2010).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Environmental DNA metabarcoding, Vertebrate, Rainfall, General Liner Model, Permutational Multivariate Analysis of Variance","lastPublishedDoi":"10.21203/rs.3.rs-7806039/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7806039/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Water-based eDNA has a potential to simultaneously detect aquatic and terrestrial species in watershed-scale eDNA biodiversity monitoring. However, the detection of terrestrial species remains difficult or undetected due to complex environmental effects. This study targeted vertebrate eDNA transported from land to rivers through surface runoff during rainfall in three rivers within the Kiyotake River watershed in Japan. We compared the detection rates of different taxonomic groups under rainfall and non-rainfall conditions and determined how environmental factors (e.g., discharge, rainfall amount) influence sample-level alpha and beta diversity. The vertebrate detection rate after rainfall exceeded that during non-rainfall. Birds, mammals, and amphibians were detected more frequently after rainfall, whereas fish were detected more frequently under non-rainfall conditions. Generalized linear model analyses revealed that river discharge and pH had significant negative effects on the detection number, while filtration volume had a significant positive effect. Furthermore, Permutational Multivariate Analysis of Variance indicated that filtration volume and rainfall occurrence significantly affected beta diversity. Our findings underscore the importance of sampling design after rainfall events, for eDNA-based biomonitoring, offering practical guidance for enhancing biodiversity assessments at the watershed scale.","manuscriptTitle":"Impact of Rainfall on Aquatic and Terrestrial Species Monitoring Using Riverine Environmental DNA","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-08 19:40:00","doi":"10.21203/rs.3.rs-7806039/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"bf936e69-6ac8-4619-a18b-e2af5848fa00","owner":[],"postedDate":"December 8th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":59218750,"name":"Biological sciences/Ecology"},{"id":59218751,"name":"Earth and environmental sciences/Ecology"},{"id":59218752,"name":"Earth and environmental sciences/Environmental sciences"}],"tags":[],"updatedAt":"2026-03-06T18:54:21+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-08 19:40:00","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7806039","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7806039","identity":"rs-7806039","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}