Can eDNA replace trawl surveys for estuarine species distribution modeling: Insights from Collichthys lucidus in the Yangtze River Estuary

Can eDNA replace trawl surveys for estuarine species distribution modeling: Insights from Collichthys lucidus in the Yangtze River Estuary | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL Ecology and Evolution This is a preprint and has not been peer reviewed. Data may be preliminary. 13 March 2025 V1 Latest version Share on Can eDNA replace trawl surveys for estuarine species distribution modeling: Insights from Collichthys lucidus in the Yangtze River Estuary Authors : Xiaoyu Geng , Wei Tang , Jianhui Wu , Chunxia Gao , and Xuefang Wang 0000-0001-7955-8776 [email protected] Authors Info & Affiliations https://doi.org/10.22541/au.174184964.44198355/v1 Published Ecology and Evolution Version of record Peer review timeline 397 views 265 downloads Contents Abstract Supplementary Material Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Species occurrence data form the basis of the establishing Species Distribution Models (SDMs). As a new technology, environmental DNA (eDNA) has been widely used for species monitoring and species diversity assessment, but it is still unclear whether it can replace or supply traditional trawl surveys to provide occurrence data for SDMs. This study took a bifurcated estuary with seawater flow backward phenomenon - the Yangtze River estuary as the research area, and the typical marine benthos - Collichthys lucidus as the research object, we carried out a comparative survey during two trips in August and November 2021, aiming to use trawl-survey data as the control group to assess the results of habitat modeling using eDNA data alone or in combination. The results showed that the trawl surveys indicated that C. lucidus inhabited almost exclusively the nearshore waters of estuaries with high salinity, which is consistent with the traditional perception that it does not inhabit freshwater (i.e. marine species). However, eDNA sampling suggested that this species was also widely distributed in the freshwater rivers of the south branch of the Yangtze River, which is likely a result of seawater intrusion from the north branch. Consequently, the Maximum Entropy Model (MaxEnt) predictions based on eDNA expanded the suitable habitat range of C. lucidus in the estuarine area. Moreover, imprecise spatial information affects the identification of key environmental requirements, underestimating the importance of salinity in habitat selection for C. lucidus. Our study highlights that in highly dynamic and open water environments like estuaries, cautious evaluation is necessary when using eDNA as species occurrence data for prediction with spatially explicit models requiring precise spatial information. 1. Introduction Species distribution models (SDMs) are mathematical models that combine species distribution data with environmental data to predict past, present, or future species distributions (Elith et al., 2009; Pecchi et al., 2019). As a common tool and model for ecological and biogeographic research (Peterson et al., 2015), SDMs have been widely applied to terrestrial, marine, and freshwater environments, including for protected area planning, climate change impact assessment, biodiversity evaluation, and species habitat modeling (Melo-Merino et al., 2020). Species occurrence data form the basis for SDMs and can be collected from literature, specimens, or the field (Melo-Merino et al., 2020). The accuracy of the SDMs depends not only on the sampling effort (number of occurrence data) used to generate the model (Aizpurua et al., 2015), but also on the accuracy of occurrence data, which includes uncertainty in species identification, bias in the selection of sample locations, and incomplete spatial coverage of the true distribution of species (Kramer-Schadt et al., 2013; Guisan et al., 2017). In traditional approaches for the modeling of marine organisms, the most common method of collecting species occurrence data is through surveys and monitoring using various gears (e.g. trawl, gillnet, trap etc.), which has the advantage of being real and reliable (Pennino et al., 2016), but has the disadvantage of being invasive, labor-intensive, time-consuming, and expensive (Thomsen et al., 2015), and hence, does not fully satisfy the needs of exploring the oceans. This situation calls for the development of new and emerging species monitoring methods to collect species occurrence data (Rondinini et al., 2006). Environmental DNA (eDNA) technology has been widely used in recent years for research on species diversity detection, rare species monitoring, invasive species detection, and habitat modeling (Aizpurua et al., 2015). Environmental DNA technology is used to qualitatively or quantitatively analyze species, community structure, and habitats of organisms in the environment by collecting DNA extracted from the environment (Thomsen et al., 2015). Compared to traditional trawl surveys, eDNA technology has the characteristics of high sensitivity, biological friendliness, low operational requirements, and high efficiency, which are conducive to quickly obtaining various types of biological information in the ocean (Thomsen et al., 2012). However, eDNA technology also has problems supporting habitat modeling, including a certain amount of mobility (Andruszkiewicz Allan et al., 2021) and false-positive and false-negative results (Jerde 2021), which may cause modeling bias. Based on these characteristics, it is worth evaluating whether eDNA technology can replace or supplement the results of traditional gear surveys to provide species occurrence information for habitat modeling. For example, Shelton et al. (2022) conducted traditional acoustic trawl surveys on the Pacific Coast, while conducting eDNA sampling. Although there are certain differences between the two methods, both provide valuable information when using spatial statistical models to study the large-scale spatial distribution and abundance of species. Neto et al. (2020) used species occurrence data from multiple databases to construct the maximum entropy model (MaxEnt) to simulate suitable habitats for salamanders followed by sampling using eDNA and validated the assessment using occupancy models with good results. Estuaries, as open areas where oceans and rivers meet, provide spawning grounds, nurseries and migratory corridors for many marine species, and play a critical role in fish life history processes and in shaping the spatial patterns of community compositio(McLusky et al., 2004). In temperate estuaries, fish richness continuously declines from the ocean toward the river (Selleslagh et al., 2008; Whitfield et al., 2011). Compared to freshwater species that spawn in estuaries and primarily inhabit riverine areas, marine and estuarine species dominate the high-salinity regions near the ocean, resulting in pronounced spatial variation in fish communities (Nicolas et al., 2010). Therefore, accurate monitoring of fish distribution has significant practical applications in the management and conservation of estuarine ecosystems. Recent studies have demonstrated the effectiveness of eDNA technology in monitoring estuarine fish diversity and rare species (Chen et al., 2024; Gibson et al., 2023), making its application to species distribution modeling and habitat preference identification increasingly realistic. The Yangtze River Estuary is the largest estuary in the western Pacific Ocean and is an ecologically important area (Yuan et al., 2002) with a unique topography that divides it into north branch and south branch, where more than 140 species of fish feed, reproduce, and inhabit (Zhuang et al., 2006). Traditionally, bottom-trawl survey data have been used for fish distribution and habitat modeling in this area (Pan et al., 2021a; Meng et al., 2021). However, in recent years, in the context of the ten-year fishing ban imposed by the Chinese government on the Yangtze River, there has been an explosive growth in the application of eDNA-based technology for fish resource monitoring (Jia et al., 2020; Whitaker., et al., 2021), which has provided new sources of data on species occurrence and distribution and has led to the practice of evaluating the effectiveness of modeling different sources of species occurrence data. Collichthys lucidus , a warm-water and nearshore benthic fish, is a common catch in traditional fisheries of the Yangtze River Estuary and its adjacent waters (Zhang et al., 2010). This study uses occurrence data of C. lucidus obtained through both net surveys and eDNA to perform habitat suitability modeling. The objective of the study is to explore whether eDNA can effectively replace or complement bottom trawl surveys in estuarine areas, providing species occurrence information to accurately identify suitable habitats and key environmental requirements for marine fish. This could offer a stronger foundation for improving fish spatiotemporal monitoring and spatial planning of conservation areas in the Yangtze River Estuary. 2. Materials and methods 2.1. Collection of occurrence data Two trips of bottom-trawl surveys combined with eDNA sampling were conducted during August and November 2021 to evaluate aquatic resources in the Yangtze River Estuary under Shanghai’s jurisdiction (121.4° E–122.2° E, 31.3° N–31.7 °N). A total of 16 fixed stations were set up for monitoring fishery resources and the environment. The distribution of the stations is shown in Fig. 1. Fig. 1. Distribution map of fish monitoring and survey stations in the Yangtze River Estuary in 2021 (the blue arrows in the figure represent the path of saltwater intrusion in the north branch of the estuary, and the black arrows represent the river path map). 2.1.1. Data from bottom-trawl surveys The bottom trawl used for sampling had a mouth width of 6 m, cod-end mesh size of 2 cm, and height of 2 m. One tow was performed at each station at a speed of 2 NM/h for 30 min. All catches obtained at each site were brought back to the laboratory for species identification and enumeration, and the spatial location of the station at which C. lucidus occurred was recorded. 2.1.2. Data from eDNA surveys Water samples (1000 mL each) were collected from each site, and on-site vacuum filtration was conducted using 47 mm diameter mixed-fiber filter membranes with a 0.45-μm pore size. Filtration equipment was pre-sterilized in 75% ethanol. The filtered samples were then stored at −80 ℃ in 1.5 mL brown centrifuge tubes and promptly transported to the laboratory. Filtration blanks and negative controls were coextracted alongside the samples and were subjected to the same protocol as the samples. DNA extraction was carried out using the DNeasy Blood and Tissue kit, following the protocols of Ficetola et al. (2008), Dejean et al. (2011), and Renshaw et al. (2015). Filter membranes were cut, ground, and soaked in 300 µL of buffer ATL and 20 µL of proteinase K. Incubation at 56 ℃ for 1.5 h was followed by kit instructions, including washing and elution in 40 µL of AE buffer. Extracted eDNA concentration was immediately determined using a Nanodrop 2000 and assessed on a 1.0% agarose gel, showing no data or bands for filtration blanks or negative controls. PCR amplification utilized FISH eDNA universal primers (Mifish-F [(5’-GTCGGTAAAACTCGTGCCAGC-3’] and Mifish-R [3’-GTTTGACCCTAATCTATGGGGGGTGATAC-5’]), generating fragments of 297 ± 25 bp. Three PCR replicates were performed for each sample, with environmental samples, filtration blanks, and negative controls included. The 25-μL amplification system contained 5× reaction buffer (5 µL), 5 µL of 5× GC buffer, 2 µL dNTP (2.5 mmol/L), 1 µL each of forward and reverse primers (10 µmol/L), 2 µL template DNA (20 ng/µL), 8.75 µL double-distilled water, and 0.25 µL Q5 DNA polymerase. A two-step PCR protocol involved 95 ℃ pre-denaturation for 5 min, 55 cycles of denaturation at 95 ℃ for 30 s and annealing at 60 ℃ for 30 s. Gel electrophoresis (1.5% agarose gel) confirmed no amplification in filtration blanks or negative controls. Sequencing on the Illumina MiSeq platform completed the process. Raw sequencing data were saved in FASTQ format and initially screened based on sequence quality. The raw bipartite sequencing data were spliced and deduplicated using Vsearch (v2.13.4_linux_x86_64) and cutadapt (v2.3) software and clustered into operational taxonomic units (OTUs) according to a 97% similarity level to obtain the representative sequences and OTU tables. Representative OTU sequences were compared with the reference sequence databases NCBI (https://www.ncbi.nlm. nih.gov/) and MitoFish (http://mitofish. aori. u-tokyo. ac. jp/download. html) for taxonomic annotation of crop species, and site distribution information on the occurrence of C. lucidus was recorded and organized. 2.2. Acquisition and preprocessing of predictor variables The environmental variables were selected based on previous characteristics studies of fish ecological habitats in Yangtze River Estuary (Pan et al., 2021), including temperature (T), salinity (S), dissolved oxygen (DO), pH, and chlorophyll concentration (Chl) were measured. All marine environmental data were downloaded from the COPERNICUS website (http://marine.copernicus.eu/), we used Pearson’s correlation coefficients (r), using an |r| > 0.70 to cull collinear predictors (Dormann et al., 2013). The retained environmental variables were coupled with species occurrence data and the specific range and resolution of each environment variable are shown in Table 1. ArcMap 10.4 software was used to convert the netCDF raw files into raster layers in the TIFF format. Due to the small spatial extent of this study area, raster layers were converted to ASCII format files with a spatial resolution of 0.083° × 0.083° and identical boundaries. Additionally, considering that salinity variation in estuarine areas is the best predictor for explaining fish distribution36, using a finer spatial resolution allows for a more accurate depiction of the influence of salinity gradients on the distribution of C. lucidus . Therefore, during the November survey, a multiparameter water quality analyzer (WTW-3430) was employed for field measurements of salinity at each station. Kriging interpolation was applied to generate the detailed spatial distribution of salinity across the study area, which was then compared with species presence obtained through trawling and eDNA to reveal the specific effects of salinity on the distribution of C. lucidus . Table 1 Overview of environmental variables used in this study for the MaxEnt. 2.3. Construction, prediction, and evaluation of the MaxEnt In the present study, the MaxEnt was used to construct a habitat model for C. lucidus in the Yangtze River Estuary. The MaxEnt is based on the principle of maximum entropy and uses the relationship between the presence-only data of species and environmental variables to simulate the distribution of species (Phillips et al., 2008), It possesses high prediction precision and accuracy (Hernandez et al., 2006), and has been widely used in the field of marine life modeling (Elith et al., 2011). To evaluate the predictability of the model, 70% of the occurrences data were randomly set as the training set and 30% as the test set, and the calculation was repeated 100 times to reduce error. The prediction results were imported into ArcMap 10.4 software and maps of the habitat suitability distribution of C. lucidus in the Yangtze River Estuary were drawn using model simulation (Wang et al., 2023). The habitat suitability index (HSI) was used to characterize the distribution of habitat suitability for C. lucidus , which ranges from 0 to 1, and the closer it is to 1, the higher the habitat suitability (Phillips et al., 2008). Two indicators were chosen to assess the model performance: (1) the area under the curve (AUC) value of the area of interest under the receiver operating characteristic (ROC) curve of the subjects. If this has a value of approximately 0.5, it indicates random prediction, and a value close to 1 indicates better model performance and correlation between environmental variables; 0.5 < AUC < 0.7 indicates the average predictive ability, and 0.7 < AUC < 0.9 indicates good predictive ability, and 0.9 < AUC < 1 indicates an excellent predictive ability (Swets, 1988). (2) Model sensitivity is a widely used conditional probability in SDMs and it represents the probability of the model correctly predicts that a species will be observed at a given location and is indexed between 0 and 1, with values closer to 1 indicating better modeling (Liu et al., 2016). The effects of different environmental factors on distribution of habitat suitability for C. lucidus were analyzed using the jackknife method, and the contribution of different environmental factors to habitat suitability was ranked and compared (Muposhi et al., 2016). Environmental factor response curves were plotted to reflect the habitat preferences for C. lucidus. The modeling of the MaxEnt and visualization of the output results were accomplished using MAXENT 3.4.1 software (https://biodiversityinformatics.amnh.org/open_source/maxent/). 3. Results 3.1. Species distribution obtained from different sources of species occurrence data During the August 2021 survey, there was a significant difference in the spatial distribution of occurrence sites of C. lucidus , as indicated by the two methods. The eDNA technique revealed that C. lucidus was detected at eight sites, or 50% of the total, but was almost exclusively distributed in the waters of the south branch of the Yangtze River Estuary, whereas the trawl collected the target species at six sites (37.5% of the total), which were all distributed in nearshore marine areas outside estuaries (Fig. 2A). Only one site (Z15) showed targets identified using both methods (Fig. 2A). The survey conducted during November 2021 showed that trawl-collected C. lucidus sites remained predominantly outside the estuary, with 11 eDNA detections (68.7%). However, several sites in the entrance area of the south branch continued to have only eDNA indicative of C. lucidus (Z4, Z5, and Z6, Fig. 2B), whereas multiple detections in the nearshore marine areas outside estuaries appeared to have targets investigated by both methods (Z13, Z14, Z15and Z12, Fig. 2B). Fig. 2. Schematic diagram of stations with eDNA and trawl detections of Collichthys lucidus in the Yangtze River Estuary for August and November 2021. (A) Distribution of stations detected during the August survey. (B) Distribution of stations detected during the November survey. 3.2. Model performance based on different species occurrence data Using AUC values to assess the performance of models built using different species occurrence data revealed that modeling based on trawl data produced the best results (AUC ≥ 0.74, Fig. 3), significantly outperforming the predictions of the stochastic model, and the peak of the ROC curve was closest to the point (0.0, 1.0). Modeling results using the eDNA were inferior to those of trawl data, with significantly poor performance, especially for November data (Fig. 3). Predictions of the model that combined the two types of data were the worst for the two months, and were substantially below the indicator level of trawl data. Sensitivity metrics reflected similar results, with the model constructed using trawl data yielding the highest sensitivity and the best performance (0.89 and 0.74 for August and November, respectively), and the results were close to the modeling results using eDNA (0.64 and 0.66 for August and November, respectively). The sensitivity of the model constructed by combining the two monitoring datasets was the worst (0.69 and 0.66 for August and November, respectively). Fig. 3. Receiver operating characteristic (ROC) curves for Collichthys lucidus habitat using the MaxEnt for August and November 2021 surveys using eDNA and trawl data in the Yangtze River Estuary (the top row shows the results of traditional gear, environmental DNA, and a combination of the two types of monitoring data for the August survey and the bottom row corresponds to the results for the November survey). 3.3.Validation of habitat suitability patterns based on different types of species occurrence data The habitat suitability pattern of C. lucidus showed significant differences between the models constructed using eDNA and trawl data for the August 2021 survey. The area of habitat suitability obtained by the model constructed using eDNA was mainly located in the south branch of the Yangtze River Estuary and the coastal waters outside the estuary (Fig. 4A), whereas trawl data showed that the areas of habitat suitability were mainly distributed in the offshore area outside the estuary, with higher suitability closer to the ocean (Fig. 4B). The most significant difference between the two data simulations was whether the south branch of the Yangtze River was identified as a suitable habitat area (Fig. 4A and B). In contrast, the habitat suitability patterns in November showed smaller differences between the two surveys (Fig. 4D and E). However, compared to August, the modeling using trawl surveys more prominently highlighted the difference in suitability between freshwater and brackish habitats (Fig. 4B and E). A comparison of the predicted results with the observed data reveals that trawl surveys consistently detected stably species occurrence points across different months (Fig. 4B and E), and the absence points also remained relatively stable (Fig. 5B and E). In contrast, species absence locations inferred from eDNA exhibited notable spatial variability (Fig. 5A and D), with two sampling stations in the high suitability areas predicted by the November model failing to detect the target species (Fig. 5D). Fig. 4. Spatial distribution of habitat suitability for Collichthys lucidus in the Yangtze River Estuary for the August and November 2021 surveys and presence stations (A–C: August, D–F: November, ‘●’ is presence station). Fig. 5. Spatial distribution of habitat suitability for Collichthys lucidus in the Yangtze River Estuary for the August and November 2021 surveys and absence stations (A–C: August, D–F: November, ‘ ’ is absence station). 3.4. Ranking the importance of environmental factors The ranking of environmental factors influencing the distribution of C. lucidus revealed by the three types of occurrence data also shows some differences: the modeling outputs based on trawl data are robust, with the importance of environmental factors in August and November consistently ranked as SST > SSS > DO > Chl > pH (Fig. 5B and E). However, models using eDNA data alone or in combination show that the ranking of importance varies across months and data types (Fig. 6 A, C, D, F). Moreover, the importance of salinity, a critical factor for identifying marine fish habitats, has been underestimated. Fig. 6. Importance ranking of environmental factors based on the jackknife method (the top row shows the results of traditional trawl, environmental DNA, and the combination of the two types of monitoring data for the August survey, and the bottom row corresponds to the results for the November survey). 3.5 Comparison of Environmental Factor Response Curves based on different types of species occurrence data The response curves of environmental factors derived from different data sources showed notable differences (Fig. 7). Specifically, in the results based on trawl data, the partial dependence curves for each environmental factor exhibited similar trends across different months, particularly for salinity, a key factor in defining estuarine habitat characteristics. In contrast, the eDNA based results showed consistent partial dependence patterns only for chl and DO across months (Fig. 7A, B, F, G), while SSS and SST displayed varying degrees of opposite trends (Fig. 7D, E, I, J). Fig. 7. Response curves of environmental factors based on different types of species occurrence data (A – E for August and F – J for November) 4 Discussion 4.1. Can eDNA technology provide data for modeling species distribution? Owing to its high sensitivity, high identification efficiency, independence of developmental stages, and low sample requirements, eDNA technology can rapidly identify the presence of multiple species in a short period of time with low environmental impact (Thomsen et al., 2015). Theoretically, eDNA technology can provide three types of data for species surveys: occurrence, abundance, and biological habitat data (Pennino et al., 2016). Occurrence data refer to the presence or absence of a species as determined by detected DNA and can be used to assess biodiversity and monitor species distribution (Wu et al., 2023). When obtaining species presence data, both traditional gear and eDNA technology have errors, but their mechanisms are different. In the case of correct species identification, species information captured by nets must be real (Renshaw et al., 2015), but if the amount of effort is insufficient or based on the selectivity of gear, the results of the gear survey can produce false negatives (Green et al., 1993), i.e., the fact that the location where the target species is not caught by gears does not necessarily imply that it does not occur in the sampling station. In contrast, although eDNA technology, can provide more information on the occurrence of species at the sampling site (Doi et al., 2017), owing to experimental errors, unknown identification of hybridized species DNA, and missing genetic banks, the information provided about the presence of species is not completely accurate. If samples are not properly preserved, eDNA technology cannot be used to detect all species information (Wang et al., 2023). Therefore, in this study, information on the occurrence of C. lucidus obtained from gear surveys was the most reliable for modeling purposes. Therefore, models based on presence-only data could be used as a control group to evaluate the effectiveness of species distribution modeling using eDNA data. Obtaining catch per unit of effort with gear is often used as an index of abundance for SDMs (Yang et al., 2021). In eDNA technology, abundance information is obtained by judging the relative abundance of a species in the environment based on the concentration of eDNA (Haxton et al., 2020), The quantitative relationship between eDNA concentration and species biomass is currently unclear (Mariani et al., 2019), leading to the possibility that it may be unsuitable for use as information on species occurrence required for modeling. In recent years, the MaxEnt, a modeling technique based on presence-only species data, has been applied by many ecologists to simulate terrestrial and marine habitats (Pearson et al., 2006). The MaxEnt requires only a small number of species occurrence positions to produce highly accurate results (Dejean et al., 2011). Therefore, the MaxEnt was used in the present study to compare the differences between simulations using different monitoring data to assess the impact of this new technique on habitat suitability and critical environmental requirements of the Yangtze River Estuary with respect to C. lucidus . 4.2. What causes the variation in distribution patterns of C. lucidus indicated by eDNA and trawl data? In recent years, eDNA technology has been widely used in the construction of species diversity indicators. Many of these surveys believe that eDNA technology can capture biological information in the region more comprehensively than traditional technology (Elith et al., 2011). Therefore, eDNA technology is considered a potential biodiversity monitoring tool (He et al., 2023). For example, Zhou et al. (2022) showed that eDNA identified 38.2% more fish species than bottom-trawl surveys in the Zhoushan Sea. Zou et al. (2020) showed that eDNA metabarcoding identified 32.05% more fish species than bottom-trawl surveys in the Pearl River Estuary. Thomas et al. (2023) found that eDNA detected higher species richness per 30 samples compared to bottom trawls or fyke nets. However, the assessment of species diversity within a region need not consider the precise matching of species occurrence information with the spatial location at each site and only needs a survey of the total region of interest and interpret results accordingly (DeAngelis et al., 2017). In contrast, SDMs are typically spatially explicit models in which the precise spatial location of species occurrence data is a prerequisite for matching and predicting habitat environmental requirements (Bertolino et al., 2020). A comparison of the distribution (species occurrence points) of C. lucidus in the Yangtze River Estuary revealed significant differences between trawl and eDNA surveys (Fig. 2). The most important difference was that the eDNA survey suggested that C. lucidus could be found widely in the freshwater of the river (i.e., the river-dominated south branch of the Yangtze River Estuary, Fig. 2 A and D), whereas trawl-based observations indicated that except for the Z3 site in November the species almost exclusively inhabited nearshore waters outside the estuary, where the salinity was higher (Fig. 2 B and E). Whereas Z3 is where the north saltwater intrusion meets the freshwater of the river, this site has a higher salinity than the other sites in the river, and may be the edge of the distribution of C. lucidus in the river. Clearly, the conclusions of the trawl survey are consistent with the current knowledge, as C. lucidus is considered a benthic shallow marine fish (Zhang et al., 2010). In their historical survey of the Yangtze River Estuary, Zhang and Zhang (1985) used a stow net to continuously sample the south and north branches of the Yangtze River Estuary for nine months from 1982 to 1983 and found that none of the stations in the south branch caught C. lucidus in stark contrast to the monthly collection of samples from the north branch of the Yangtze River Estuary, which experiences seawater intrusion (Fig. A1). In temperate regions, salinity influences fish distribution through physiological salinity tolerance, with spatial variations in salinity shaping fish community composition and causing significant spatial shifts along entire estuaries (Whitfield et al., 2011; Gibson et al., 2023). Studies on five estuaries in Japan have shown that the proportion of detected marine species increases with rising salinity, compared to freshwater and brackish species (Ahn et al., 2020). The difference in the distribution of C. lucidus between the north branch and south branch of the Yangtze River Estuary may be closely related to the distribution of salinity. The Yangtze River Estuary has three bifurcations and four outlets to the sea (Qiu et al., 2012; Fig. 1), and its water quality is affected by both the freshwater transported by the Yangtze River and the saline water injected into the Yellow Sea, with a huge difference in the salinities of the south branch and north branch (Song et al., 2020). This phenomenon is attributed to the distinctive mechanism wherein the south branch, serving as a conduit for the Yangtze River’s discharge into the sea, exhibits lower salinity. The substantial saltwater intrusion from tides, predominantly in the north branch, results in a notable influx of saltwater into the south branch, particularly at Chongtou (Fig. 1). Consequently, the salinity level in the north branch of the Yangtze River Estuary becomes comparable to that of the outer sea. Thus, C. lucidus , a shallow-sea fish, inhabits the north branch of the Y Yangtze River Estuary (Zhang et al., 2010). However, owing to the transport mechanism of seawater intrusion in the Yangtze River Estuary, eDNA (or carcass) can be transported over long distances with the water flow (Deiner et al., 2014), resulting in frequent detection of eDNA from the north branch in the south branch of the Yangtze River Estuary, which ultimately leads to the identification of species that should not be present in unsuitable habitats, which is in contrast to the results of trawl surveys. 4.3. What is the impact of spatially biased species occurrence information on estimates of habitat suitability and environmental requirements? MaxEnt is a SDMs that is based on the principle of maximum entropy, which compares information on environmental conditions of occurrence positions of species with those of background points, simulates species distribution by analyzing the limitations of the environmental variables at the presence points and obtains the relationship between environmental factors and species distribution (Elith et al., 2011). The habitat suitability of C. lucidus was constructed using three different occurrence datasets that were compared with the nearshore areas of habitat suitability constructed based on trawl data. Results of the simulation using only eDNA and combined eDNA defined the south branch of the Yangtze River Estuary as the area of habitat suitability, which significantly enlarged the range of suitable habitats for C. lucidus in the Yangtze River Estuary (Fig. 4). Accurate occurrence information directly affects the outputs of SDMs in terms of areas of habitat suitability (Melo-Merino et al., 2020). Traditionally, occurrence data have often included significant geographic bias, which arises mainly from inadequate sampling; for example, littoral areas are more likely to be surveyed than others because of their proximity or low-cost accessibility, whereas offshore areas are difficult to sample or have insufficient data, resulting in missed or underestimated information on species occurrences (Araújo et al., 2006), making it impossible to identify some of the areas of habitat suitability. Estuaries, as open systems, may have eDNA monitoring influenced by external systems. Their highly dynamic environments and fluid nature complicate the spatial distribution of eDNA (Dickie et al., 2018). In some cases, expand the geographic range over which species information exists, resulting in spatially biased sample data, which results in inaccurate information on species presence in some areas (Yackulic et al., 2013). Similarly, spatially biased sample data can lead to the misestimation of environmental requirements. Theoretically, suitable environmental conditions are a prerequisite for the emergence of species. In addition, environmental information on the location of the species reflects the preference of the species for its habitat (Phillips et al., 2009). However, spatial bias in eDNA can cause models to fit environmental requirements associated with a particular geographic location incorrectly or excessively, resulting in unreliable models, thereby affecting model evaluation (Yackulic et al., 2013). Specifically, for the Yangtze River Estuary, the watershed of the estuarine-dominated south branch and the marine-dominated outer edge of the estuary have significant salinity differences; therefore, the distribution and frequency of occurrences in different regions will directly affect the judgment of critical environmental requirements for the species. Environmental DNA frequently detect species presence in low-salinity areas, leading to models based solely on eDNA or incorporating eDNA underestimating the importance of salinity as a key environmental factor for marine fish habitats. 5. Conclusions and outlook This study explores the potential of using eDNA to provide species occurrence data for modeling marine fish distributions in estuarine areas. The findings on C. lucidus. indicate that while eDNA can detect more occurrence points compared to trawling, it may not accurately match the spatial locations of species occurrences. Spatial accuracy has significant implications for SDMs in identifying primary suitable habitats and critical environmental requirements. In previous studies, despite being influenced by persistence and current transport, eDNA metabarcoding has successfully detected variations in fish communities at different spatial scales (<1 km) within estuaries and between adjacent environments(Cole et al., 2022; Sellers et al., 2018). However, the spatial accuracy of species occurrence information provided by this technique may still be insufficient to meet the fine spatial scale requirements for habitat modeling in small areas. Thus, when sampling eDNA in a field with complex terrain, it is necessary to analyze hydrodynamic characteristics such as curvature, flow velocity, and topography and environmental factors such as ultraviolet light, temperature, and pH at the sampling points to determine their effects on eDNA stability and fluidity to evaluate the reliability of subsequent analyses using eDNA. Appendix Fig. A1. Distribution of Collichthys lucidus in the waters of the north and south branches of the Yangtze River Estuary based on stow net and stake-hold net surveys during 1982–1983 (based on the results of Zhang and Zhang (1985)). CRediT authorship contribution statement Xiaoyu Geng: Writing-Original Draft, Methodology, Visualization, Conceptualization. Wei Tang: Writing-Original Draft, Methodology, Visualization, Conceptualization. Jianhui Wu: Investigation, Project administration, Funding acquisition. Chunxia Gao: Project administration, Supervision. Xuefang Wang: Methodology, Formal analysis, Supervision, Writing-Review & Editing, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability The original species data and environmental data used in the MaxEnt model are included in the Supporting Information. All marine environmental data could be find from the COPERNICUS website (http://marine.copernicus.eu/). For further inquiries, the corresponding author may be contacted. Acknowledgements This work was supported by the Technology Innovation Project of the Shanghai Agricultural and Rural Committee (Shanghai Agricultural Science and Technology Innovation No. 2-1, 2022) and Capacity Building Project for Local Universities, Science and Technology Commission of Shanghai Municipality (21010502200). References Ahn, H., Kume, M., Terashima, Y., Ye, F., Kameyama, S., Miya, M., … & Kasai, A. (2020). Evaluation of fish biodiversity in estuaries using environmental DNA metabarcoding. PLoS One , 15(10), e0231127. https://doi.org/10.1371/journal.pone.0231127 Aizpurua, O., Paquet, J. Y., Brotons, L., & Titeux, N. (2015). Optimising long‐term monitoring projects for species distribution modelling: how atlas data may help. Ecography , 38 (1), 29-40. https://doi.org/10.1111/ecog.00749 Andruszkiewicz Allan, E., Zhang, W. G., C Lavery, A., & F Govindarajan, A. (2021). Environmental DNA shedding and decay rates from diverse animal forms and thermal regimes. Environmental DNA , 3 (2), 492-514. https://doi.org/10.1002/edn3.141 Araújo, M. B., & Guisan, A. (2006). Five (or so) challenges for species distribution modelling. Journal of biogeography , 33 (10), 1677-1688. https://doi.org/10.1111/j.1365-2699.2006.01584.x Bertolino, S., Sciandra, C., Bosso, L., Russo, D., Lurz, P. W., & Di Febbraro, M. (2020). Spatially explicit models as tools for implementing effective management strategies for invasive alien mammals. Mammal Review , 50 (2), 187-199. https://doi.org/10.1111/mam.12185 Chen, X., Li, S., Zhao, J., & Yao, M. (2024). Passive eDNA sampling facilitates biodiversity monitoring and rare species detection. Environment International , 187 , 108706. https://doi.org/10.1016/j.envint.2024.108706 Cole, V. J., Harasti, D., Lines, R., & Stat, M. (2022). Estuarine fishes associated with intertidal oyster reefs characterized using environmental DNA and baited remote underwater video. Environmental DNA , 4 (1), 50-62. https://doi.org/10.1002/edn3.190 DeAngelis, D. L., & Yurek, S. (2017). Spatially explicit modeling in ecology: a review. Ecosystems , 20 (2), 284-300. https://doi.org/10.1007/s10021-016-0066-z Deiner, K., & Altermatt, F. (2014). Transport distance of invertebrate environmental DNA in a natural river. PloS one , 9 (2), e88786. https://doi.org/10.1371/journal.pone.0088786 Dejean, T., Valentini, A., Duparc, A., Pellier-Cuit, S., Pompanon, F., Taberlet, P., & Miaud, C. (2011). Persistence of environmental DNA in freshwater ecosystems. PloS one , 6 (8), e23398. https://doi.org/10.1371/journal.pone.0023398 Dickie, I. A., Boyer, S., Buckley, H. L., Duncan, R. P., Gardner, P. P., Hogg, I. D., … & Weaver, L. (2018). Towards robust and repeatable sampling methods in eDNA‐based studies. Molecular Ecology Resources , 18 (5), 940-952. https://doi.org/10.1111/1755-0998.12907 Doi, H., Inui, R., Akamatsu, Y., Kanno, K., Yamanaka, H., Takahara, T., & Minamoto, T. (2017). Environmental DNA analysis for estimating the abundance and biomass of stream fish. Freshwater Biology , 62 (1), 30-39. https://doi.org/10.1111/fwb.12846 Dormann, C. F., Elith, J., Bacher, S., Buchmann, C., Carl, G., Carré, G., … & Lautenbach, S. (2013). Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography , 36 (1), 27-46. https://doi,org/10.1111/j.1600-0587.2012.07348.x Elith, J., & Leathwick, J. R. (2009). Species distribution models: ecological explanation and prediction across space and time. Annual review of ecology, evolution, and systematics , 40 (1), 677-697. https://doi.org/10.1146/annurev.ecolsys.110308.120159 Elith, J., Phillips, S. J., Hastie, T., Dudík, M., Chee, Y. E., & Yates, C. J. (2011). A statistical explanation of MaxEnt for ecologists. Diversity and distributions , 17 (1), 43-57. https://doi.org/10.1111/j.1472-4642.2010.00725.x Ficetola, G. F., Miaud, C., Pompanon, F., & Taberlet, P. (2008). Species detection using environmental DNA from water samples. Biology letters , 4 (4), 423-425. https://doi.org/10.1098/rsbl.2008.0118 Gibson, T. I., Carvalho, G., Ellison, A., Gargiulo, E., Hatton-Ellis, T., Lawson-Handley, L., … & Creer, S. (2023). Environmental DNA metabarcoding for fish diversity assessment in a macrotidal estuary: a comparison with established fish survey methods. Estuarine, Coastal and Shelf Science , 294 , 108522. https://doi.org/10.1016/j.ecss.2023.108522 Green, R. H., & Young, R. C. (1993). Sampling to detect rare species. Ecological Applications , 3 (2), 351-356. https://doi.org/10.2307/1941837 Guisan, A., Thuiller, W., & Zimmermann, N. E. (2017). Habitat suitability and distribution models: with applications in R . Cambridge University Press. Haxton, T., & Friday, M. (2020). Spatiotemporal changes in juvenile lake sturgeon abundance in a large river. Journal of Applied Ichthyology , 36 (6), 772-779. https://doi.org/10.1111/jai.14106 He, X., Jeffery, N. W., Stanley, R. R. E., Hamilton, L. C., Rubidge, E. M., & Abbott, C. L. (2023). eDNA metabarcoding enriches traditional trawl survey data for monitoring biodiversity in the marine environment. ICES Journal of Marine Science , 80 (5), 1529-1538. https://doi.org/10.1093/icesjms/fsad083 Hernandez, P. A., Graham, C. H., Master, L. L., & Albert, D. L. (2006). The effect of sample size and species characteristics on performance of different species distribution modeling methods. Ecography , 29 (5), 773-785. https://doi.org/10.1111/j.0906-7590.2006.04700.x Jerde, C. L. (2021). Can we manage fisheries with the inherent uncertainty from eDNA?. Journal of fish biology , 98 (2), 341-353. https://doi.org/10.1111/jfb.14218 Jia, H., Wang, Y., Yoshizawa, S., Iwasaki, W., Li, Y., Xian, W., & Zhang, H. (2020). Seasonal variation and assessment of fish resources in the Yangtze Estuary based on environmental DNA. Water , 12 (10), 2874. https://doi.org/10.3390/w12102874 Kramer‐Schadt, S., Niedballa, J., Pilgrim, J. D., Schröder, B., Lindenborn, J., Reinfelder, V., … & Wilting, A. (2013). The importance of correcting for sampling bias in MaxEnt species distribution models. Diversity and distributions , 19 (11), 1366-1379. https://doi.org/10.1111/ddi.12096 Liu, C., Newell, G., & White, M. (2016). On the selection of thresholds for predicting species occurrence with presence‐only data. Ecology and evolution , 6 (1), 337-348. https://doi.org/10.1002/ece3.1878 Mariani, S., Baillie, C., Colosimo, G., & Riesgo, A. (2019). Sponges as natural environmental DNA samplers. Current Biology , 29 (11), R401-R402. https://doi.org/10.1016/j.cub.2019.04.031 McLusky, D. S., & Elliott, M. (2004). The estuarine ecosystem: ecology, threats and management (Vol. 216). Oxford University Press. Melo-Merino, S. M., Reyes-Bonilla, H., & Lira-Noriega, A. (2020). Ecological niche models and species distribution models in marine environments: A literature review and spatial analysis of evidence. Ecological Modelling , 415 , 108837. https://doi.org/10.1016/j.ecolmodel.2019.108837 Meng, W., Gong, Y., Wang, X., Tong, J., Han, D., Chen, J., & Wu, J. (2021). Influence of spatial scale selection of environmental factors on the prediction of distribution of Coilia nasus in Changjiang River Estuary. Fishes , 6 (4), 48. https://doi.org/10.3390/fishes6040048 Muposhi, V. K., Gandiwa, E., Chemura, A., Bartels, P., Makuza, S. M., & Madiri, T. H. (2016). Habitat heterogeneity variably influences habitat selection by wild herbivores in a semi-arid tropical savanna ecosystem. PloS one , 11 (9), e0163084. https://doi.org/10.1371/journal.pone.0163084 Neto, J. G. D. S., Sutton, W. B., Spear, S. F., Freake, M. J., Kéry, M., & Schmidt, B. R. (2020). Integrating species distribution and occupancy modeling to study hellbender ( Cryptobranchus alleganiensis ) occurrence based on eDNA surveys. Biological Conservation , 251 , 108787. https://doi.org/10.1016/j.biocon.2020.108787 Nicolas, D., Lobry, J., Le Pape, O., & Boët, P. (2010). Functional diversity in European estuaries: relating the composition of fish assemblages to the abiotic environment. Estuarine, Coastal and Shelf Science , 88 (3), 329-338. https://doi.org/10.1016/j.ecss.2010.04.010 Pan, S., Tian, S., Wang, X., Dai, L., Gao, C., & Tong, J. (2021). Comparing different spatial interpolation methods to predict the distribution of fishes: A case study of Coilia nasus in the Changjiang River Estuary. Acta Oceanologica Sinica , 40 , 119-132. https://doi.org/10.1007/s13131-021-1789-z Pearson, R. G., Raxworthy, C. J., Nakamura, M., & Townsend Peterson, A. (2007). Predicting species distributions from small numbers of occurrence records: a test case using cryptic geckos in Madagascar. Journal of biogeography , 34 (1), 102-117. https://doi.org/10.1111/j.1365-2699.2006.01594.x Pecchi, M., Marchi, M., Burton, V., Giannetti, F., Moriondo, M., Bernetti, I., … & Chirici, G. (2019). Species distribution modelling to support forest management. A literature review. Ecological Modelling , 411 , 108817. https://doi.org/10.1016/j.ecolmodel.2019.108817 Pennino, M. G., Conesa, D., López-Quílez, A., Munoz, F., Fernández, A., & Bellido, J. M. (2016). Fishery-dependent and-independent data lead to consistent estimations of essential habitats. ICES Journal of Marine Science , 73 (9), 2302-2310. https://doi.org/10.1093/icesjms/fsw062 Peterson, A. T., Papeş, M., & Soberón, J. (2015). Mechanistic and correlative models of ecological niches. European Journal of Ecology, 1 (2): 28–38. https://doi.org/10.1515/eje-2015-0014 Phillips, S. J., & Dudík, M. (2008). Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography , 31 (2), 161-175. https://doi.org/10.1111/j.0906-7590.2008.5203.x Phillips, S. J., Dudík, M., Elith, J., Graham, C. H., Lehmann, A., Leathwick, J., & Ferrier, S. (2009). Sample selection bias and presence‐only distribution models: implications for background and pseudo‐absence data. Ecological applications , 19 (1), 181-197. https://doi.org/10.1890/07-2153.1 Qiu, C., Zhu, J., & Gu, Y. (2012). Impact of seasonal tide variation on saltwater intrusion in the Changjiang River estuary. Chinese Journal of Oceanology and Limnology , 30 (2), 342-351. https://doi.org/10.1007/s00343-012-1115-x Renshaw, M. A., Olds, B. P., Jerde, C. L., McVeigh, M. M., & Lodge, D. M. (2015). The room temperature preservation of filtered environmental DNA samples and assimilation into a phenol–chloroform–isoamyl alcohol DNA extraction. Molecular ecology resources , 15 (1), 168-176. https://doi.org/10.1111/1755-0998.12281 Rondinini, C., Wilson, K. A., Boitani, L., Grantham, H., & Possingham, H. P. (2006). Tradeoffs of different types of species occurrence data for use in systematic conservation planning. Ecology letters , 9 (10), 1136-1145. https://doi.org/10.1111/j.1461-0248.2006.00970.x Sellers, G. S., Di Muri, C., Gómez, A., & Hänfling, B. (2018). Mu-DNA: a modular universal DNA extraction method adaptable for a wide range of sample types. Metabarcoding and Metagenomics , 2 , e24556. https://doi.org/10.3897/mbmg.2.24556 Selleslagh, J., Amara, R., Laffargue, P., Lesourd, S., Lepage, M., & Girardin, M. (2009). Fish composition and assemblage structure in three Eastern English Channel macrotidal estuaries: a comparison with other French estuaries. Estuarine, Coastal and Shelf Science , 81 (2), 149-159. https://doi.org/10.1016/j.ecss.2008.10.008 Shelton, A. O., Ramón-Laca, A., Wells, A., Clemons, J., Chu, D., Feist, B. E., … & Park, L. (2022). Environmental DNA provides quantitative estimates of Pacific hake abundance and distribution in the open ocean. Proceedings of the Royal Society B , 289 (1971), 20212613. https://doi.org/10.1098/rspb.2021.2613 Song, Z., Shi, W., Zhang, J., Yuan, D., Wu, Q., & Wang, R. (2020, June). Impact of sea level rise on estuarine salt water intrusion–A numerical model study for Changjiang Estuarine. In IOP Conference Series: Earth and Environmental Science (Vol. 525, No. 1, p. 012130). IOP Publishing. https://doi.org/10.1088/1755-1315/525/1/012130 Swets, J. A. (1988). Measuring the accuracy of diagnostic systems. Science , 240 (4857), 1285-1293. https://doi.org/10.1126/science.3287615 Thomsen, P. F., & Willerslev, E. (2015). Environmental DNA–An emerging tool in conservation for monitoring past and present biodiversity. Biological conservation , 183 , 4-18. https://doi.org/10.1016/j.biocon.2014.11.019 Thomsen, P. F., Kielgast, J., Iversen, L. L., Møller, P. R., Rasmussen, M., & Willerslev, E. (2012). Detection of a diverse marine fish fauna using environmental DNA from seawater samples. https://doi.org/10.1371/journal.pone.0041732 Wang, J., & Tabeta, S. (2023). MaxEnt modeling to show patterns of coastal habitats of reef-associated fish in the South and East China Seas. Frontiers in Ecology and Evolution , 11 , 1027614. https://doi.org/10.3389/fevo.2023.1027614 Whitaker, J. M., Brower, A. L., & Janosik, A. M. (2021). Invasive lionfish detected in estuaries in the northern Gulf of Mexico using environmental DNA. Environmental Biology of Fishes , 104 (11), 1475-1485. https://doi.org/10.1007/s10641-021-01177-6 Whitfield, A. K., Elliott, M., Basset, A., Blaber, S. J. M., & West, R. J. (2012). Paradigms in estuarine ecology–a review of the Remane diagram with a suggested revised model for estuaries. Estuarine, Coastal and Shelf Science , 97 , 78-90. https://doi.org/10.1016/j.ecss.2011.11.026 Wu, F., Zou, Y., Qin, S., Li, F., & Zhang, Y. (2023). eDNA biomonitoring of macroinvertebrate communities for the bioassessment of a River’s ecological status. Water , 15 (2), 308. https://doi.org/10.3390/w15020308 Yackulic, C. B., Chandler, R., Zipkin, E. F., Royle, J. A., Nichols, J. D., Campbell Grant, E. H., & Veran, S. (2013). Presence‐only modelling using MAXENT: when can we trust the inferences? Methods in Ecology and Evolution , 4 (3), 236-243. https://doi.org/10.1111/2041-210x.12004 Yang, Z., Bai, Y., Alatalo, J. M., Huang, Z., Yang, F., Pu, X., … & Guo, X. (2021). Spatio-temporal variation in potential habitats for rare and endangered plants and habitat conservation based on the maximum entropy model. Science of the Total Environment , 784 , 147080. https://doi.org/10.1016/j.scitotenv.2021.147080 Yuan, X. Z., Liu, H., & Lu, J. J. (2002). The ecological and environmental characteristics and conservation of the wetlands in the Changjiang Estuary, China. Environmentalist , 22 , 311-318. https://doi.org/10.1023/A:1020758712639 Zhang, T., Zhuang, P., Zhang, L. Z., Hou, J. L., Liu, J. Y., Wang, Y. L., & Liu, J. (2010). Composition and diversity of fish species in the coast of the Yangtze River Estuary. Chin J Appl Environ Biol , 16 (6), 817-821. https://doi.org/ 10.3724/SP.J.1145.2010.00817 Zhang, Gx., & Zhang, Xs. (1985). A survey on the fisheries of set stow net in the Changjiang Estuary. Journal of Fisheries of China , 185-198. Zhou, S., Fan, C., Xia, H., Zhang, J., Yang, W., Ji, D., … & Liu, N. (2022). Combined use of eDNA metabarcoding and bottom trawling for the assessment of fish biodiversity in the Zhoushan Sea. Frontiers in marine science , 8 , 809703. https://doi.org/10.3389/fmars.2021.809703 Zhuang, P. (2006). Fishes of the Yangtze Estuary . Shanghai Science and Technology Press. Zou, K., Chen, J., Ruan, H., Li, Z., Guo, W., Li, M., & Liu, L. (2020). eDNA metabarcoding as a promising conservation tool for monitoring fish diversity in a coastal wetland of the Pearl River Estuary compared to bottom trawling. Science of the Total Environment , 702 , 134704. https://doi.org/10.1016/j.scitotenv.2019.134704 Table 1 Overview of environmental variables used in this study for the MaxEnt. Temperature T 16–30 °C 0.25° × 0.25° Salinity S 0–17 ‰ 0.25° × 0.25° pH pH 7–8 - 0.25° × 0.25° Dissolved oxygen DO 243–332 mmol·m −3 0.25° × 0.25° Chlorophyll concentration Chl 6–14 mg·m −3 0.25° × 0.25° Tables and Figures Captions Table 1 Overview of environmental variables used in this study for the MaxEnt. Fig. 1. Distribution map of fish monitoring and survey stations in the Yangtze River Estuary in 2021. Fig. 2. Schematic diagram of stations with eDNA and trawl detections of Collichthys lucidus in the Yangtze River Estuary for August and November 2021. Fig. 3. Receiver operating characteristic (ROC) curves for Collichthys lucidus habitat using the MaxEnt for August and November 2021 surveys using eDNA and trawl data in the Yangtze River Estuary. Fig. 4. Spatial distribution of habitat suitability for Collichthys lucidus in the Yangtze River Estuary for the August and November 2021 surveys and presence stations (A–C: August, D–F: November, ‘●’ is presence station). Fig. 5. Spatial distribution of habitat suitability for Collichthys lucidus in the Yangtze River Estuary for the August and November 2021 surveys and presence stations (A–C: August, D–F: November, ‘ ’ is absence station). Fig. 6. Importance ranking of environmental factors based on the jackknife method. Fig. 7. Response curves of environmental factors in different months (A–E for August and F–J for November). Fig. A1. Distribution of Collichthys lucidus in the waters of the north and south branches of the Yangtze River Estuary based on stow net and stake-hold net surveys during 1982–1983. Fig. 1. Distribution map of fish monitoring and survey stations in the Yangtze River Estuary in 2021 (the blue arrows in the figure represent the path of saltwater intrusion in the north branch of the estuary, and the black arrows represent the river path map). Fig. 2. Schematic diagram of stations with eDNA and trawl detections of Collichthys lucidus in the Yangtze River Estuary for August and November 2021. (A) Distribution of stations detected during the August survey. (B) Distribution of stations detected during the November survey. Fig. 3. Receiver operating characteristic (ROC) curves for Collichthys lucidus habitat using the MaxEnt for August and November 2021 surveys using eDNA and trawl data in the Yangtze River Estuary (the top row shows the results of traditional gear, environmental DNA, and a combination of the two types of monitoring data for the August survey and the bottom row corresponds to the results for the November survey). Fig. 4. Spatial distribution of habitat suitability for Collichthys lucidus in the Yangtze River Estuary for the August and November 2021 surveys and presence stations (A–C: August, D–F: November, ‘●’ is presence station). Fig. 5. Spatial distribution of habitat suitability for Collichthys lucidus in the Yangtze River Estuary for the August and November 2021 surveys and presence stations (A–C: August, D–F: November, ‘ ’ is absence station). Fig. 6. Importance ranking of environmental factors based on the jackknife method (the top row shows the results of traditional trawl, environmental DNA, and the combination of the two types of monitoring data for the August survey, and the bottom row corresponds to the results for the November survey). Fig. 7. Response curves of environmental factors in different months (A–E for August and F–J for November). Fig. A1. Distribution of Collichthys lucidus in the waters of the north and south branches of the Yangtze River Estuary based on stow net and stake-hold net surveys during 1982–1983 (based on the results of Zhang and Zhang (1985)) Supplementary Material File (image1.jpg) Download 5.04 MB File (image7.jpg) Download 12.56 MB Information & Authors Information Version history V1 Version 1 13 March 2025 Peer review timeline Published Ecology and Evolution Version of Record 21 Aug 2025 Published Copyright This work is licensed under a Non Exclusive No Reuse License. Collection Ecology and Evolution Keywords comparative ecological experiment marine none of the above vertebrate Authors Affiliations Xiaoyu Geng Shanghai Ocean University College of Marine Living Resource Sciences and Management View all articles by this author Wei Tang Shanghai Ocean University College of Marine Living Resource Sciences and Management View all articles by this author Jianhui Wu Shanghai Aquatic Wildlife Conservation Research Center View all articles by this author Chunxia Gao Shanghai Ocean University College of Marine Living Resource Sciences and Management View all articles by this author Xuefang Wang 0000-0001-7955-8776 [email protected] Shanghai Ocean University College of Marine Living Resource Sciences and Management View all articles by this author Metrics & Citations Metrics Article Usage 397 views 265 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Xiaoyu Geng, Wei Tang, Jianhui Wu, et al. Can eDNA replace trawl surveys for estuarine species distribution modeling: Insights from Collichthys lucidus in the Yangtze River Estuary. Authorea . 13 March 2025. DOI: https://doi.org/10.22541/au.174184964.44198355/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.174184964.44198355/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9fe6427f7ef158f4',t:'MTc3OTIyNjgyMw=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();