https://orcid.org/0000-0001-8064-7880
Figures
Abstract
Estuarine ecosystems globally are being restored through re-naturalization efforts, including the Nakdong River Estuary (NRE) in South Korea, which partially re-opened in 2019 after decades of disconnection. We investigated how this restoration affected the feeding ecology and fitness of Erythroculter erythropterus (Skygager), a non-native cyprinid that became dominant in the upper NRE following barrage construction. We surveyed fish populations, analyzed stomach contents using Next Generation Sequencing, and assessed morphological changes from May 2018 to June 2022. Following NRE opening, zooplankton and fish diversity indices increased, while phytoplankton and benthic invertebrate diversity decreased. E. erythropterus showed altered prey selection patterns, shifting from diverse prey items to predominantly zooplankton after NRE opening, coinciding with altered vertical migration patterns of plankton communities. While length distribution did not change significantly, the condition factor K decreased significantly, indicating reduced fitness. Body shape parameters also shifted toward more elongated forms, suggesting morphological responses to changed hydraulic conditions and feeding ecology. Our findings demonstrate that controlled estuarine reconnection can affect non-native species through cascading trophic effects, potentially serving both restoration and non-native species management objectives. This study highlights the complex ecological consequences of partial estuarine restoration and the importance of considering multiple trophic levels when evaluating restoration outcomes.
Citation: Gim J-S, Hong D, Kim D-K, Lineman MJ, Jeong K-S, Lee J-Y, et al. (2025) Changes in prey selection and fitness of non-native Erythroculter erythropterus following estuarine restoration in the Nakdong River, South Korea. PLoS One 20(9): e0328372. https://doi.org/10.1371/journal.pone.0328372
Editor: Ram Kumar, Central University of South Bihar, INDIA
Received: December 10, 2024; Accepted: July 1, 2025; Published: September 24, 2025
Copyright: © 2025 Gim et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Obtained sequences were deposited in the NCBI repository under accession number PRJNA935053.
Funding: This research was funded by the National Research Foundation of Korea (grant number RS-2024-00344522 to H.J.).
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Estuarine ecosystems represent unique environments where biological, physical, and chemical exchanges occur between marine and freshwater habitats [1]. Nutrient interactions between the two aquatic systems give rise to a diverse range of habitats within the estuarine environment [2,3]. These habitats are particularly important as breeding grounds for marine fishes, provide suitable environments for marine fish larvae, and offer protection from predators, acting as nursery and reproduction grounds, serving as refuges and feeding areas for fish, potentially optimizing survival, development, and growth [4,5]. However, over the past century, estuarine barrages have been constructed for a variety of purposes, including urbanization, tidal power generation, and traffic management [6], leading to the disappearance of many natural estuarine ecosystems [7].
With the increasing recognition of the importance of estuarine ecosystems, some governments have recently implemented policies focusing on the benefits of opening barrages to facilitate estuarine ecosystem restoration [8–10]. Restoration of estuarine ecosystem permits saltwater inflow from the marine ecosystem, which considerably alters the local biota [11–13] by impacting; (i) morphological traits (such as brook trout gills and stickleback caudal peduncles and head size, and Cyprinidae fish from pectoral fin to ventral fin; [14–16]); (ii) migration patterns (such as fish’s habitat selection, move to feeding area and hatching area selection; [17–20]); (iii) life-history characteristics (such as fish’s maximum body size and fitness of small pelagic fish; [21–24]); (iv) community structure (such as decreasing stability of fish community and impact of non-native species; [25–27]); (v) behavioural patterns (such as European sea bass diet and feeding response of Cyprinidae fish; [28,29]). Several studies have examined the effects of environmental changes on morphological traits and migration patterns, but gaps remain in understanding how these changes affect community structure, behavioural patterns, and life history characteristics [30]. Furthermore, existing studies have principally focused on examining the effects of individual factors, which limits understanding sequential results generated by cascading effects for a series of processes.
Prior to the construction of the Nakdong River Estuary (NRE) barrage in 1987, the system functioned as a dynamic natural estuary with seasonal variations in river flow and salinity gradients [31]. During the monsoon season (summer), increased freshwater input pushed the saltwater boundary downstream, while during dry seasons (winter), saltwater extended further upstream. This natural variability created diverse microhabitats supporting rich biodiversity, including important commercial fish species and benthic communities adapted to fluctuating salinity conditions [31]. The natural NRE provided essential ecological functions including nutrient cycling, sediment transport, and connectivity between riverine and marine environments. In South Korea, the construction of the NRE barrage in 1987 severely disrupted the river-sea connectivity [31]. The barrage was initially constructed to secure freshwater supplies and prevent saltwater intrusion, but it effectively transformed the estuarine ecosystem into a primarily freshwater system. This transformation was further intensified by the Four Major Rivers Project, which deepened water channels and further decreased habitat diversity [32]. These changes had profound ecological consequences, including a sharp decrease in native species diversity [31,33] and the proliferation of non-native freshwater species [34]. The altered hydrological conditions created lentic (still water) habitats favoring non-native species adapted to lake-like conditions. Recognizing these ecological impacts, the Korean government initiated a restoration effort through partial re-opening of the NRE barrage, creating a brackish transition zone to potentially recover species diversity [35]. This controlled re-opening provides a valuable opportunity to explore the ecological consequences of partial estuarine restoration [36].
Erythroculter erythropterus (skygager) is a cyprinid fish species native to China, Russia, and parts of Korea, but considered non-native to the Nakdong River basin [37]. This predatory fish typically inhabits lentic water bodies such as lakes and reservoirs, preferring slow-flowing or still water environments [38]. E. erythropterus is primarily zooplanktivorous in its juvenile stages but becomes increasingly piscivorous as it matures, feeding on small fish and large zooplankton [39]. The species has become dominant in the upper reaches of the NRE following barrage construction, likely due to the creation of reservoir-like conditions that favor its ecological requirements [33]. Recent studies have documented the spatial distribution and migration patterns of E. erythropterus in the Nakdong River basin, including its responses to water quality variations [40]. Understanding how this dominant non-native species responds to estuarine restoration efforts is crucial for managing both the species and the overall ecosystem.
Optimal Foraging Theory (OFT) provides a theoretical framework for understanding predator feeding decisions based on energy maximization principles [41,42]. According to OFT, predators should preferentially select prey that maximize their net energy intake, considering both the energy gained from consumption and the costs associated with searching, capturing, and handling prey [43]. The concept of prey profitability – defined as the ratio of energy gained to handling time – is central to OFT predictions [42]. When multiple prey types are available, predators are expected to exhibit prey switching behavior, shifting their preference toward more abundant or profitable prey types to optimize their foraging efficiency [44,45]. In aquatic ecosystems, changes in environmental conditions can alter prey availability and profitability, potentially leading to shifts in predator selectivity patterns. The re-establishment of estuarine conditions in the NRE likely altered the relative abundance and accessibility of different prey types for E. erythropterus, providing an opportunity to examine how environmental restoration affects predator-prey dynamics through the lens of optimal foraging principles.
In this study, we investigated how the partial restoration of estuarine conditions resulting from barrage opening affected the food sources and fitness (specifically condition factor K and body shape parameters) of E. erythropterus, a dominant non-native species in the upper NRE. Specifically, we (1) compared differences in potential food sources and stomach contents of E. erythropterus using Next Generation Sequencing (NGS) analysis, and (2) explored changes in the fitness and morphological characteristics of E. erythropterus before and after NRE opening. We hypothesized that the shift from predominantly freshwater to partial estuarine conditions would alter available food sources, which would subsequently affect the prey selection patterns and fitness of this non-native species. Understanding these responses may provide insights into how controlled barrage management could be used as a potential tool for managing non-native fish populations while promoting ecosystem restoration.
1.1. Study area and NRE re-opening protocol
The Nakdong River is the longest river in South Korea, with a length of 510 km and a drainage area of approximately 23,384 km2. Rainfall occurs primarily during summer due to the monsoon, but in winter it is low. Therefore, storing fresh water is important [46]. The Nakdong River barrage (35°08′20.3″N, 128°57′26.2″E) was constructed from 1983 to 1987 to serve as a sustainable water supply and to control saltwater intrusion [47]. The barrage comprising 10 sluices, has four regulatory sluices (length: 47.5 m, height: 8.3 m) and six main gates (length: 47.5 m, height: 9.2 m). The regulatory sluices release freshwater during the rainy season (typically July–August) and at low tides, thus maintaining the water level. Conversely, the main gates typically remain closed (Fig 1A, C, and S1 Fig).
Appearance after intrusion of different amounts of seawater before opening the NRE (A: Sky view; C: Cross-section view) and after intrusion of 9,300,000 tons (third opening of the NRE) of seawater (B: Sky view; D: Cross-section view). The amount of salt is represented by colours ranging from blue (0) to red (greater than 9.5).
When the opening began, only one regulatory gate (gate number 9) was utilized (Fig 1A, C, and S1 Fig). The regulatory gate was opened by dropping the lower (S1 Fig) or raising the upper (S1 Fig) halves of the gate. This gate was opened seven times between 2019 and 2022 (S1 Table in S1 File). The first and second openings served as pilot runs, during which approximately 1 million tons of saltwater flowed in less than 1 hour following the opening (S2 Table in S1 File). Beginning with the third opening of the NRE, the opening period was extended to approximately one month, during which the highest recorded saltwater flow occurred (9.3 million tons [47.8% of total amount of seawater intruded]; Fig 1B, D; S2 Table in S1 File). Therefore, our study period was divided into before and after the third opening of NRE (2020-06-04), which recorded the highest salt inflow.
2. Materials and methods
2.1. Field survey and data collection
To assess the impact of opening the NRE, we chose a study site located 3.9 km upstream, based on the extent of salinity penetration before and after the NRE opening (Fig 1A). This study site has the representativeness of monitoring salt inflow from both branches of the Nakdong River in the upper part of the NRE, which divides into two branches and flows to the sea (Fig 1A). Fish sampling was carried out at the study site from May 2018 to June 2022 (Fig 1; 10 in total; 35°07′35.00″N, 128°57′11.37″E). Fishes were collected for 24 h during each sampling period using a fixed shore net (mesh: 15 × 15 mm, height: 1 m, leader net: 15 m) to which local fishermen attached three fish traps (each mesh: 15 × 15 mm, width 1.5 m; height 1.8 m; length 3 m). Collected fish were identified and classified according to Kim and Park [37] and Nelson [48] and then released (S3 Table in S1 File). However, exotic species were not released, as this is prohibited by the act on protecting native and endemic species (Act on the Conservation and Use of Biological Diversity, Act No. 14513, Dec.27, 2016 by Ministry of the Environment). Morphological parameters (total length ± 0.1 cm, weight ± 0.1 g) of E. erythropterus, exotic species from other basins in South Korea, were measured immediately following capture, and their stomachs were removed for stomach content analysis. To avoid contamination from foreign-derived DNA, we clamped the lower oesophageal portion using forceps and dissected the stomach using medical scissors rinsed in 98% methanol and flame-sterilised. The stomach and tissue samples were preserved in 99% ethanol and separated into prey items. All samples were stored at room temperature prior to the analysis. All fish sampling and handling procedures were conducted in accordance with ethical guidelines and with approval from the Pusan National University Laboratory Safety Management Center (Institutional Animal Care and Use Committee). We obtained field research permission (No. 1380) from the National Institute of Environmental Research. Exotic species were not released back to the river as prohibited by the Act on the Conservation and Use of Biological Diversity (Act No. 14513, Dec. 27, 2016) by the Ministry of Environment. Fish were euthanized humanely using an overdose of MS-222 (tricaine methanesulfonate) prior to dissection and tissue sampling to minimize suffering.
We measured water quality parameters including water temperature, dissolved oxygen (DO), pH, conductivity, and salinity for surface water. Furthermore, we collected water samples (2 L at a depth of 0.5 m) from the fish sampling sites to analyse total nitrogen (TN) and total phosphorus (TP) levels before (20 May 2020) and after (4 June 2020) opening the NRE. A DO meter (YSI Model 58, USA) was used to measure water temperature and DO; conductivity and pH were measured using a conductivity meter (YSI model 152; Yellow Springs Instruments, Yellow Springs, OH, USA) and a pH meter (Orion Model 250A; Orion Research Inc., Boston, MA, USA), respectively. The TN and TP concentrations were determined using a spectrophotometer (Japan Fantec Research Institute, Shizuoka, Japan) as described by Wetzel and Likens [49]. To evaluate the maximum saltwater inflow after opening the NRE, we collected data on saltwater inflow quantities at 1 km intervals based on previous literature [50].
For phytoplankton, zooplankton, and benthic invertebrate analyses, we followed standardized sampling and identification protocols. Phytoplankton samples were collected using a Van Dorn water sampler at 0.5 m depth, preserved with Lugol’s solution, and concentrated by sedimentation. Identification was performed using an Olympus BX51 microscope at 400 × magnification following taxonomic keys by [51]. Cell counts were performed using a Sedgwick-Rafter counting chamber, and biomass was estimated using standardized biovolume measurements and conversion factors [52]. Zooplankton were collected by vertical hauls using a conical net (mesh size 64 μm, mouth diameter 30 cm) from near-bottom to surface. Samples were preserved in 4% formaldehyde solution and later identified to species level using identification keys [52]. Counting was conducted using a Bogorov counting chamber under a stereomicroscope (Olympus SZX16). Benthic invertebrates were collected using a Petersen grab sampler (0.025 m2) with three replicates at each sampling site. Samples were sieved through a 500 μm mesh, preserved in 70% ethanol, and sorted under a stereomicroscope. Specimens were identified to the lowest taxonomic level possible using appropriate taxonomic keys [53]. Biomass was determined as wet weight after blotting.
2.2. Potential food source before and after the NRE opening
To evaluate potential food sources for E. erythropterus, we compared the diversity indices of phytoplankton, zooplankton, benthic invertebrates and fish before and after NRE opening. The dominant and subdominant species were classified based on the species with highest relative abundance. For phytoplankton, zooplankton and benthic invertebrates, the results of the literature search were used [50], and the diversity indices of fish were calculated using the results of our field survey. The dominance [54], diversity [55], evenness [56], and richness [57] indices were calculated. These indices were computed using PAST 3.0 [58] based on data for the number of individuals. To assess vertical diel migration patterns, we conducted stratified sampling at three depths (surface: 0.5 m, middle: 6 m, and bottom: 10 m) before (10 May 2022) and after (23 May 2022) NRE opening. At each depth, water samples (2 L) were collected using a Van Dorn water sampler for phytoplankton, and zooplankton were collected using a horizontal tow of a 64 μm mesh net at each depth stratum. All samples were collected during daylight hours (10:00–14:00) to standardize potential diurnal effects. Abundance was calculated as individuals per liter for each depth, and vertical distribution patterns before and after NRE opening were compared by examining proportional changes at each depth. Due to the limited sampling occasions (one sampling event before and after opening), descriptive comparisons rather than statistical tests were used to assess changes in vertical distribution patterns (Fig 6).
2.3. Stomach contents of E. erythropterus before and after the NRE opening
To achieve high-resolution identification of prey items in fish stomach contents, we employed Next Generation Sequencing (NGS) for amplifying prey DNA, which provides superior taxonomic resolution and accuracy compared to traditional visual examination methods [59]. To analyze prey selection patterns and compare food sources before and after NRE opening, we randomly selected 24 E. erythropterus (before the NRE opening (12); and after the NRE opening (12). Stomach contents and tissue samples were extracted, stored separately in a fresh state, and rinsed in autoclaved water to avoid external and self-DNA contamination. Ethanol was completely volatilised from the samples prior to DNA extraction. The samples were then homogenised. Genomic DNA from each stomach sample was isolated using the DNeasy Blood and Tissue Kit (Qiagen) according to the manufacturer’s protocol. We selected the V9 region of the 18S rDNA gene (18S V9), primarily because of its broad range in eukaryotes [60]. NGS approaches using the 18S V9 region have recently enabled accurate metabarcoding for diet characterization across diverse taxonomic groups [61], making it an established approach for biomarker analysis [62]. The library preparation followed standard protocols with modifications for stomach content samples as detailed in the supporting Information appendix.
We sequenced library samples using NGS and identified all species taxonomically. The library was sequenced using an Illumina iSeq platform (Illumina, San Diego, CA, USA) and data processing was performed using USEARCH (v11.0.667) [63]. Demultiplexed raw sequences (FASTQ files) were merged into one sequence, allowing for a maximum of ten mismatches. Merged reads with expected errors >3.0 were discarded after filtering for quality. The remaining sequences were de-replicated and clustered into operational taxonomic units (OTUs) at a 97% OTU cutoff value, removing chimeric and singleton sequences. In total, 542,207 filtered paired-end reads from 24 samples were generated using the platform, of which 97% passed the Q30 (Phred quality score >30). Each sample yielded 10,044–29,397 paired-end reads (mean: 20,903 reads) and rarefaction curve analysis indicated that the number of OTUs were saturated for all samples (S2 Fig). Merged reads were processed and clustered into OTUs using the UCLUST [64], at 97% OTU cut-off value (153 OTUs in gamma diversity). The resulting 57 OTUs were classified into 11 genus-level taxonomic groups (S4 Table in S1 File).
The resulting out sequences were searched against the NCBI database (Release 250.0; July 2022) using BLASTn [63]. We obtained a list of the top 153 taxa for each OTU with the highest identity percentage. Considering the known characteristics and distribution of each taxon, we excluded taxa not present at the study sites. If all taxa found in the search results had low scores (query cover<80%, identity <85%) or were regarded as not present, the OTUs were deleted. Finally, the taxa with the highest identity was assigned to each OTU. OTUs with 97% or higher identity were identified at the species level, while the rest (93–97%) were identified at the gene level. OTUs identified as E. erythropterus were considered ‘self-DNA’ and excluded from the subsequent procedure. Obtained sequences were deposited in the NCBI repository under accession number PRJNA935053.
We constructed a phylogenetic tree based on known or empirical predator-prey relationships among taxonomic groups. Species traits of fish, aquatic insects, and protozoa were retrieved from the literature and database repositories [65]. All species or genera were classified into one of eight taxonomic groups: etc (Unionicolidae, worms, and protozoa), fish, fungi, amoeba, phytoplankton, zooplankton, magnoliophyta, or cnidaria. To compare opening-mediated spatiotemporal changes in the NRE, we divided prey items before the NRE opening group and after the NRE opening group. A detailed description of the phylogenetic construction protocol is provided in the methods section of the appendix.
To evaluate prey selection patterns, we compared the relative abundance of potential prey items in the environment with their occurrence in E. erythropterus stomach contents. Environmental prey abundance data were obtained from diversity analyses of phytoplankton, zooplankton, benthic invertebrates, and fish communities (Table 2), while stomach content data were derived from NGS analysis results (Fig 2). Prey selection was assessed qualitatively by comparing proportional representation between environmental availability and stomach contents before and after NRE opening, following established approaches for analyzing predator feeding selectivity in the presence of multiple prey types [43,44].
A taxonomic phylogenetic tree of prey items identified from fish stomach contents was constructed, with bar graphs showing the operational taxonomic unit (OTU) count per phylum or order. Silhouettes represent the prey organism types; when a food source appeared both before and after the NRE opening, its silhouette was placed outside the circular bar. Red and blue colors represent prey items found after and before NRE opening, respectively.
To compare patterns of preferred food sources before and after NRE opening, a self-organising map (SOM) analysis was performed. SOM is an algorithm that reduces data dimensionality [66]. The Kohonen network is a competitive network system where neurones in the Euclidean map space compete against each other. SOM training was achieved using a series of species data obtained from E. erythropterus stomach contents. A detailed description of the SOM modelling analysis is provided in the methods section of the appendix. A total of 24 data samples (before NRE opening (12); and after NRE opening (12)) were collected for each variable. No strict rules regarding the number of output neurones that should be obtained were present [67]; therefore, we trained the SOM with different map sizes to choose a suitable model. A batch learning algorithm was used in this study. After training, data were clustered according to the calculated U-matrix, which provided the dissimilarity of data samples on the map plane. Clustering mostly focused on organizing feeding patterns into clusters based on (i) different discrete series of species in the stomach and (ii) before and after NRE opening. The SOM model was developed using MATLAB 6.1 with the SOM Toolbox [68]. Model validation was performed using quantization error and topographic error metrics to ensure proper network training (S3 Fig). The final neighborhood size and learning rate were determined through iterative optimization to minimize these error values while maintaining interpretable output clusters. Complete SOM mapping results for individual species patterns are provided in S4 Fig.
2.4. Population structure and fitness of E. erythropterus before and after the NRE opening
During the study period, a total of 309 length-weight data points were obtained for E. erythropterus. The analysis was based on two groups of data: before NRE opening (n = 205, May 2018 to May 2020) and after NRE opening (n = 104, June 2020 to June 2021). Sampling effort (equipment, duration, and methodology) was identical during both periods, with the difference in sample sizes reflecting actual changes in capture rates. This difference in sample size was accounted for in statistical analyses through appropriate normalization procedures. To exclude seasonal effects from the sampling periods (pre and post opening), we conducted a field survey during the same season (May and June). We verified the significance of the distribution of total fish length before and after opening the NRE. The distribution of total E. erythropterus length was categorised, and significance was confirmed using chi-square distance analysis. In our feature selection task for the two-class problem, we computed the chi-square distance using the following equation [69]:
where n is the number of bins, xi is the value of first bin and yi is the value of the second bin. This was a simple and quick approach to ranking histogram bins and was selected because it was suitable for examining distributional changes when comparing two or more clusters [69].
Length-weight relationships (LWRs) assessing the fitness of E. erythropterus, were used to estimate the weight correspondent to a given length, whereas condition factors were employed to compare fish condition, considering that heavy fish of a given length were in good condition [70]. The LWRs of each species were estimated based on the equation:
where log a is the regression intercept and b is the regression slope [71]. Prior to regression, outliers were removed using linear regression of the log-transformed equation [70]. The condition factor (K) was estimated from the relationship K = 100 W/L3 to assess the fish conditions [70]. Additionally, the form factor (a) was used to determine whether the body shape changed significantly, with reference ranges for different body forms: eel-like (0.00099–0.00165), elongated (0.00775–0.00906), fusiform (0.01310–0.01400), and short and deep (0.01720–0.01930) [70].
3. Results
3.1. Changes in water quality and potential food sources following NRE opening
Surface water quality variables (water temperature, DO, TN, TP, pH, conductivity and salinity) showed minimal differences before and after opening the NRE (Table 1). However, a pronounced vertical stratification was observed after NRE opening, with salinity of water samples collected from ≥ 8 m depth reaching ≥ 5 psu compared to only 0.4 psu for the surface water (Fig 1D, S2 Table in S1 File). This vertical salinity gradient was not present before the NRE opening.
NRE opening did not result in a change in dominant phytoplankton species, but substantially increased the relative abundance of the dominant species Aulacoseira ambigua f. japonica from 18.0% to 67.5% (Table 2), representing a 3.75-fold increase. The dominance index more than doubled after NRE opening (from 0.14 to 0.38), while the diversity, evenness, and richness indices all decreased (Table 2), with diversity showing the most notable decline (from 2.41 to 1.73).
In zooplankton communities, NRE opening resulted in a shift in subdominant species from Mesocyclops leuckarti to Diaphanosoma brachyuru (Table 2). The dominant species remained Daphnia longispina, but its relative abundance decreased markedly from 37.8% to 9.0%, representing a 76% reduction. The zooplankton richness index decreased from 1.70 to 1.44, while the dominance, diversity, and evenness indices all showed increasing trends (Table 2).
For benthic invertebrates, the dominant species remained Limnodrilus gotoi both before and after NRE opening, with a slight increase in relative abundance (from 22.4% to 25.7%). Following re-opening, a notable shift in subdominant species occurred from the estuarine amphipod Grandidierella japonica to predominantly freshwater Chironomidae larvae, with relative abundance of the latter increasing from 11.4% to 22.7%, representing almost a two-fold increase. This shift appears contradictory to expectations as G. japonica is typically associated with estuarine conditions while Chironomidae are primarily freshwater organisms. This unexpected result may be related to local habitat modifications or short-term disturbance effects following the opening. All benthic invertebrate diversity indices (dominance, diversity, evenness, and richness) decreased following NRE opening (Table 2), with the richness index showing the most dramatic decline (from 1.30 to 0.32), a 75% reduction.
E. erythropterus (a non-native species) and L. macrochirus (another non-native species) were the dominant and subdominant fish species before NRE opening, respectively, constituting 93.5% and 5.5% of total abundance. After NRE opening, these species remained dominant, but the relative abundance of E. erythropterus decreased from 93.5% to 78.3% (Table 2), a 16% reduction indicating a potential restructuring of the fish community. The fish community dominance index decreased substantially after opening the NRE (from 0.88 to 0.49), while diversity, evenness, and richness indices all increased notably (Table 2), with the richness index showing the most dramatic improvement (from 0.57 to 1.82), representing a 3.2-fold increase. These changes in fish community structure suggest that the partial restoration of estuarine conditions may reduce the dominance of non-native freshwater species, potentially providing more opportunities for native estuarine species.
3.2. Shifts in prey diversity and feeding patterns of E. erythropterus following NRE opening
The total number of OTUs in the stomachs decreased significantly from 52 (91.2% of total identified OTUs) to 29 (50.9%) after NRE opening (χ² = 146.7, df = 114, p = 0.021), indicating a substantial reduction in prey diversity as measured by OTU richness (Fig 2, S4 Table in S1 File). Food source distribution revealed that the overall proportions of fungi, phytoplankton, and other species decreased after opening the NRE, while the overall proportion of zooplankton increased (Fig 2). In phytoplankton, a variety of species were distributed before the NRE opening; however, following NRE opening only Discostella sp. (freshwater species), Chlorophyceae sp. (marine species), and Nitzschia sp. (marine species) were present (Fig 2). For zooplankton, Daphnia spp. (freshwater species) and Lecane sp. (freshwater species) showed a significant increase in distribution proportions after NRE opening (Fig 2). With respect to fish species, since the re-opening of the NRE, the proportion of Cyprinidae increased, while the proportion of Perchidae decreased (Fig 2).
The Self-Organizing Map (SOM) analysis of E. erythropterus stomach content data revealed distinct dietary patterns before and after opening the NRE (Fig 3). The SOM model differentiated feeding patterns into two clearly distinct clusters based on prey composition. Using a dissimilarity threshold value of 1.0 as the clustering criterion, nine cells on the map plane were diagonally grouped according to the U-matrix (quantisation error, 1.502; topographic error, 0.000), resulting in clear separation between samples collected after NRE opening (lower left) and before NRE opening (upper right) segments (Fig 3A). For prey items with the highest aggregation rates, three distinct groups were identified in each period (Fig 3B and C). Before opening the NRE, prey item distribution was diverse, with relatively even representation of fish, phytoplankton, and other prey groups (Fig 3B); however, after opening the NRE, zooplankton (particularly Daphnia spp. and Lecane sp.) became notably more dominant in the diet (Fig 3C). This pronounced shift in diet composition indicates a major change in feeding behavior following the altered estuarine conditions.
(A) U-matrix clustering and dendrogram showing dissimilarity of the cells on the map plane. The patterning results for proportion of OTU numbers per phylum or order parameters (B) before and (C) after the NRE opening. Colour scale is related to distances between map units. Black colours represent large distances and white colours represent small distances. The diagonal red line separates the two main clusters (before and after NRE opening) identified by the SOM analysis.
To assess prey selection patterns, we compared the relative abundance of prey items in the environment (Table 2) with their occurrence in E. erythropterus stomach contents (Fig 2). Before NRE opening, zooplankton comprised 37.8% of environmental abundance but showed lower representation in stomach contents, suggesting selective feeding on other prey types including fish and phytoplankton. After NRE opening, despite zooplankton environmental abundance decreasing to 9.0%, they became the dominant prey in stomach contents, indicating a shift from selective feeding to opportunistic feeding on the most accessible prey. This pattern suggests that E. erythropterus prey selection was influenced more by prey accessibility and capture efficiency than by environmental abundance alone, particularly following the altered vertical distribution of zooplankton toward surface waters after NRE opening.
3.3. Effects of NRE opening on population structure and fitness of E. erythropterus
The chi-squared distance analysis revealed differences in population size structure before and after NRE opening (χ2 distance = 6.04) (Fig 4), indicating a notable change in size distribution. E. erythropterus total length was normally distributed, ranging between 150–800 mm before NRE opening and 150–600 mm after opening (Fig 4A and B). The most notable change was the complete absence of very large individuals (>600 mm) after NRE opening, despite identical sampling effort. Additionally, the proportion of medium-sized individuals (300–450 mm) decreased substantially after NRE opening, with the number dropping from 73 individuals before opening to only 18 individuals after opening, representing a 75% reduction. This shift in size structure suggests potential changes in population demographics following altered estuarine conditions. We compared the length-weight relationship (LWR) parameters and condition factor K (as indicators of fitness) before and after NRE opening (Table 3). The b value (the slope of LWR; Eq. 2) increased from 2.97 to 3.07, while the a value (intercept of LWR; Eq. 2) decreased from 0.0079 to 0.0041 after NRE opening. The non-overlapping 95% confidence intervals for the a value (0.0052 ~ 0.0120 vs. 0.0029 ~ 0.0058) suggest a meaningful change in body shape toward a more elongated form following NRE opening. Most importantly, the mean condition factor K after NRE opening decreased significantly compared to pre-opening conditions (from 0.47 to 0.43; t = 3.809, df = 292, P < 0.001; Fig 4), indicating reduced overall fitness of E. erythropterus following the re-establishment of estuarine conditions. This change occurred despite the differences in size distribution, suggesting that the reduced condition factor reflects actual changes in fish fitness rather than merely being an artifact of different size compositions. The change in body shape was further confirmed by analyzing the form factor (a), which showed a significant decrease from 0.0079 before NRE opening to 0.0041 after opening (Fig 5A). This change indicates a shift from the ‘elongated’ form category (0.00775–0.00906) towards the ‘eel-like’ form category (0.00099–0.00165). When standardized to a common slope (a3.0), the form factor still showed a decrease, though less pronounced (Fig 5B). These morphological changes are consistent with the observed changes in length-weight relationship parameters and further support the finding that E. erythropterus underwent notable physical adaptations following changes in environmental conditions.
Before, 205 individuals; After, 104 individuals.; A statistical significance of the t-test (P < 0.05) performed to deduce if there is significant difference between condition factors K before and after opening the NRE; Solid line: median; Dash line: mean; * P < 0.01.
Before, 205 individuals; After, 104 individuals. Range of form factor for body shape categories: Elongated (0.00775–0.00906), Eel-like (0.00099–0.00165). Error bars represent standard deviation.
3.4. Vertical distribution patterns of plankton before and after NRE opening
Following NRE opening, notable changes were observed in the vertical distribution of both zooplankton and phytoplankton (Fig 6). Zooplankton showed a marked shift in vertical distribution patterns. Before NRE opening, zooplankton abundance was relatively evenly distributed across depths, with slightly higher concentrations in middle waters. After opening, zooplankton abundance increased dramatically in surface (0.5 m) waters, with rotifer abundance increasing 2.6-fold (from 350 to 899 ind./L) and copepod abundance increasing 1.6-fold (from 205 to 324 ind./L). Conversely, zooplankton abundance decreased substantially in bottom waters (10 m), with rotifer abundance decreasing by 82% (from 254 to 46 ind./L) and cladoceran abundance decreasing by 58% (from 146 to 61 ind./L) (Fig 6A, B). Phytoplankton showed an opposite pattern, with higher concentrations developing in bottom waters after NRE opening (Fig 6C, D). Before opening, phytoplankton were relatively evenly distributed throughout the water column, but after opening, their abundance increased by 44% in bottom waters (from 3,542,500–5,100,000 cells/L) while decreasing by 78% in surface waters (from 4,562,500–1,001,750 cells/L). These changes in vertical distribution patterns coincided with the development of a strong salinity gradient, with higher salinity (≥ 5 psu) in bottom waters after NRE opening. The vertical redistribution of plankton communities, particularly the shift of zooplankton toward surface waters, likely increased their availability as prey for surface-feeding fish like E. erythropterus.
Sampling period: before NRE opening (10 May 2022); after NRE opening (23 May 2022).
4. Discussion
This study compared changes in food sources and fitness of the non-native cyprinid E. erythropterus before and after partial re-opening of the Nakdong River Estuary. E. erythropterus, native to parts of northeastern Asia but introduced to the Nakdong River, typically inhabits lentic (still water) environments and is predominantly zooplanktivorous when young, becoming increasingly piscivorous with age [39]. In the NRE system, this species became dominant following barrage construction which transformed the estuary into a more reservoir-like environment. Our results revealed a significant reduction in food item diversity and an altered prey selection from a diverse diet to mainly zooplankton following the re-establishment of partial estuarine conditions. E. erythropterus trophically is a high-level predator in aquatic ecosystems [39], while the food sources identified from gastric contents for fish from our study sites mainly consisted of organisms from lower-trophic-levels (Fig 2). This finding was similar to other studies, which reported that changes in low-trophic-level communities’ diversity influenced by dam construction ultimately impacted high-trophic-level fish populations [23,72]. In addition, saltwater influx appears to be a key factor affecting the fitness of high-trophic-level consumers through both direct physiological stress and indirect effects via food web alterations. The apparent reduction in fitness of E. erythropterus following NRE opening likely reflects this species’ preference for freshwater environments over brackish conditions.
From an optimal foraging perspective, the observed dietary shift in E. erythropterus following NRE opening can be interpreted as a response to changes in prey profitability rather than simply prey availability. While zooplankton abundance increased after restoration, the shift toward predominantly zooplankton consumption may reflect suboptimal foraging conditions. According to OFT principles, the profitability of prey items depends not only on their energy content but also on capture efficiency and handling time [41,43]. Fish prey typically provide higher energy returns per individual compared to zooplankton, making them more profitable despite lower abundance [44]. The observed reduction in condition factor K suggests that the increased reliance on zooplankton, while energetically feasible, was less efficient than the previous more diverse diet. This finding aligns with OFT predictions that predators forced to rely on less profitable prey due to environmental constraints will exhibit reduced fitness outcomes [42]. The vertical redistribution of plankton communities following salinity stratification likely influenced prey encounter rates and capture success, further affecting the cost-benefit ratio of different foraging strategies. The concentration of zooplankton in surface waters may have increased their detectability but also their escape probability, altering the overall profitability of zooplankton as prey.
The vertical redistribution of plankton communities [73–75] we observed after NRE opening likely influenced prey availability for E. erythropterus. The upward migration of zooplankton toward surface waters in response to bottom water salinization appears to have made them more accessible to E. erythropterus, which typically feeds in surface to mid-water depths [40,39]. This behavioral response by zooplankton to avoid increased salinity in bottom waters is consistent with previous studies documenting salinity avoidance behaviors in freshwater zooplankton [76–78]. However, despite increased availability of zooplankton in the feeding zone of E. erythropterus, their fitness decreased, suggesting that the dietary shift toward almost exclusive zooplankton consumption was suboptimal for this species. This could be due to energetic constraints, as zooplankton generally provide less energy per individual prey item compared to fish prey [79,80].
The comparison between environmental prey availability and stomach contents revealed important insights into E. erythropterus prey selection strategies. Before NRE opening, this species demonstrated selective feeding behavior, consuming fish and diverse prey types despite zooplankton being the most abundant prey in the environment (37.8% relative abundance). This selective behavior aligns with optimal foraging theory predictions, where predators choose prey that maximize energy returns relative to handling costs [43]. Fish prey, though less abundant, provide higher energy content per individual compared to zooplankton, making them more profitable prey items [81,82]. Following NRE opening, the feeding strategy shifted dramatically. Despite zooplankton environmental abundance decreasing to 9.0%, they became the predominant prey in stomach contents. This apparent contradiction can be explained by the altered spatial distribution of prey following salinity stratification. The upward migration of zooplankton to surface waters increased their encounter rates with E. erythropterus, effectively increasing their availability despite lower overall abundance [43,44]. This shift represents a transition from energy-maximizing selective feeding to accessibility-driven opportunistic feeding, consistent with optimal foraging theory when preferred prey become less accessible [45]. The observed reduction in condition factor K following this dietary shift supports the interpretation that increased reliance on zooplankton, while energetically feasible, was less efficient than the previous more selective diet. This finding demonstrates how environmental changes can force predators to adopt suboptimal foraging strategies, ultimately affecting their fitness [43,45].
The observed changes in the fish community following NRE opening, particularly the decreased relative abundance of E. erythropterus, suggest potential shifts in competitive and predatory interactions (Figs 3 and 4). The increased fish community diversity indices after NRE opening indicate that other fish species, potentially including native estuarine species, may have gained competitive advantages under the new environmental conditions [81]. Furthermore, the complete absence of larger E. erythropterus individuals (>600 mm) after NRE opening could indicate either size-specific emigration from the study area or increased predation pressure on larger individuals by piscivorous predators that may have gained access to upper reaches following reconnection [82]. While this study focused primarily on E. erythropterus as a dominant species, these community-level interactions likely played an important role in the observed responses and warrant further investigation.
Therefore, it is possible that zooplankton, based on their increased abundance in the surface water, could increase their probability of becoming a food source for E. erythropterus. The food source and mouth shape of E. erythropterus suggest that they feed on organisms in surface waters or middle depths [39,40]. Subsequently, our findings suggest that this effect could propagate from zooplankton (low-trophic-level consumers compared to fish) to E. erythropterus that prey on them. These changes in food sources could affect the fitness of fish, as observed in studies on the reduction for condition factors of salmon due to changes in feeding behaviour resulting from variations in prey availability [83]. These food sources therefore changed alter competition and predation pressure, which in turn might drive fitness of E. erythropterus [84].
The observed changes in zooplankton diversity following NRE opening may reflect complex interactions between environmental changes and predation pressure. While E. erythropterus increased its consumption of zooplankton after restoration, the zooplankton diversity indices (evenness and richness) showed increasing trends (Table 2). This apparent paradox can be explained by the reduced overall abundance and condition of E. erythropterus following NRE opening, which likely decreased total predation pressure on zooplankton communities despite increased per-capita consumption. Additionally, the altered vertical distribution of zooplankton may have created spatial refugia in deeper waters with higher salinity, reducing their vulnerability to surface-feeding predators. These findings highlight the importance of considering both direct predation effects and indirect environmental modifications when assessing top-down control in aquatic ecosystems [45]. The complex interplay between predator fitness, prey accessibility, and environmental heterogeneity demonstrates how restoration efforts can trigger cascading effects throughout the food web.
Even though our results didn’t show significant differences for the population structure of E. erythropterus (Fig 4), the condition factor K (fitness) decreased significantly after the NRE opening (Fig 4). Also, our results which show that a shift in the food source from a broad range of prey items (Fig 3B) to a narrower range, focused on zooplankton (Fig 3C), resulting in a decrease in the fitness of individuals that consumed the altered food source (Fig 4). Crowding low-trophic-level communities might result in temporary reductions in the fitness of high-trophic-level communities due to concentrated food sources [84]. Based on our results, it was found that food source diversity like benthic invertebrates and phytoplankton decreased, while the diversity of zooplankton increased after NRE opening (Fig 2, Table 2). This implied a shift in food source structure. In particular, zooplankton, the main food source for E. erythropterus, have a low carbon content, which may adversely affect E. erythropterus condition, unlike the fish food sources they had previously consumed [79,80].
Fish undergo sensory adaptations and morphological shifts in response to changes in environmental connectivity and food availability [85,86]. Our findings indicate a significant decrease in the length-weight relationship parameter a, representing a shift in body form from a deeper-bodied elongated shape toward a more eel-like form (Table 3, Fig 6). This morphological plasticity likely represents an adaptive response to altered flow regimes and prey selection behavior.
More elongated body forms are generally associated with more efficient swimming in flowing water environments [87], suggesting that E. erythropterus may be adapting to the increased water movement following NRE opening. Additionally, the shift toward zooplankton as a primary food source may favor different foraging strategies requiring morphological adjustments. While such morphological plasticity typically develops over multiple generations, the relatively rapid changes observed here suggest phenotypic plasticity within the existing population. Similar rapid morphological responses have been documented in other cyprinid species following habitat alterations [88–90]. Additionally, plastic shifts, like phenotypic plasticity due to fish life-history traits, involve reduced lifespans and maximum body sizes as well as increased reproductive efforts after impoundment [23,88–91]. We suspected that plastic shifts of E. erythropterus might be influenced by re-opening the NRE. Therefore, our study showed that NRE re-opening can produce negative effects on food source shifts and fitness for E. erythropterus.
An important consideration for interpreting these results is the potential influence of seasonal variations in physico-chemical parameters and biological communities between monsoon and non-monsoon periods [92]. While our sampling was designed to control for seasonal effects by conducting surveys during the same seasons (May and June), the NRE system likely experiences significant seasonal fluctuations that interact with the effects of barrage opening [93]. During monsoon seasons, increased freshwater input would push the saltwater boundary downstream, potentially reducing the estuarine influence even when the barrage is partially open [94]. Conversely, during dry periods, saltwater intrusion would extend further upstream, potentially increasing stress on freshwater-adapted species like E. erythropterus [95]. These seasonal dynamics may significantly influence both the short-term responses we observed and the long-term ecological trajectory of the system following restoration efforts.
The dominance of non-native species such as E. erythropterus in the Nakdong River system [33] represents a significant management challenge, as these species can reduce native biodiversity and alter ecosystem functioning [96]. Our findings suggest that partial barrage opening could serve as a potential management tool for controlling non-native freshwater species while promoting estuarine restoration. The observed reduction in E. erythropterus fitness and the concurrent increase in overall fish diversity indices indicate that controlled estuarine reconnection may help shift the ecosystem toward a more natural state. Various approaches to estuarine restoration have been implemented globally, ranging from complete barrier removal [97,98] to various degrees of managed connectivity [99–101]. While complete removal would maximize ecological connectivity, it would eliminate the water management benefits provided by the barrage. Our results suggest that a middle-ground approach using controlled, partial openings may provide ecological benefits while maintaining water management capabilities. The management implications extend beyond E. erythropterus to the broader ecosystem. The increased fish diversity following NRE opening suggests a potential pathway for native species recovery, though longer-term monitoring would be needed to confirm this trend. However, it is important to recognize that management decisions should not be based solely on the response of a single species, even if dominant. Rather, a holistic ecosystem approach considering multiple trophic levels and native species recovery should guide barrage management strategies [102,103].
5. Conclusions
Our study provided insights into the ecological consequences of partial estuarine restoration through controlled barrage opening in the Nakdong River Estuary. We observed a significant shift in the preferred food source of the non-native E. erythropterus from a diverse diet to primarily zooplankton following NRE opening, accompanied by reduced fitness as measured by condition factor K. These changes coincided with altered vertical distribution patterns of plankton communities and significant changes in community structure across multiple trophic levels. While several studies have recommended complete barrier removal to fully restore original estuarine conditions [97,98,104], our findings highlight the complex ecological responses that can occur during partial restoration. The apparent negative effects on E. erythropterus fitness, while potentially beneficial from an invasive species management perspective, illustrate how rapid environmental changes can create challenges for established populations. Our results align with previous studies showing that morphological traits and food web structures can be significantly impacted by altered habitats following restoration efforts [105,106]. Future research should investigate the long-term responses of the NRE ecosystem to different opening regimes, considering variables such as opening duration, timing relative to seasonal cycles, and the number of open sluice gates. Additionally, expanding monitoring to include a broader range of native and non-native species would provide a more comprehensive understanding of community-level responses. We emphasize that barrage management should be approached with caution, as sudden environmental changes require organisms to adapt rapidly, potentially causing temporary stress even in species that might ultimately benefit from restoration. In conclusion, while our study focused primarily on a single dominant non-native species, the observed cascading effects—from changes in plankton communities to fish fitness—demonstrate the interconnected nature of estuarine food webs. These findings suggest that controlled barrage opening could potentially serve both ecological restoration goals and non-native species management objectives, but a gradual, adaptive management approach would likely provide the best pathway toward sustainable estuarine restoration.
Supporting information
S1 Fig. Diverse regulatory gate forms.
(A) The regulatory gate remains closed, blocking freshwater and seawater mixing. (B) The underflow and (C) the overflow method for opening the regulatory gate.
https://doi.org/10.1371/journal.pone.0328372.s001
(TIF)
S2 Fig. Rarefaction curves of in Erythroculter erythropterus’s stomach contents in NGS results.
(A to X: Sample 1 to Sample 24).
https://doi.org/10.1371/journal.pone.0328372.s002
(TIF)
S3 Fig. The results of SOM clustering.
(A) Clustering of the cells on the map plane, (B) dendrogram showing dissimilarity of the cells in the map, (C) Clustering of the cells on the map plane by three groups and (D) The patterning results for before and after NRE opening on the SOM plane, B: Before NRE opening; A: After NRE opening.
https://doi.org/10.1371/journal.pone.0328372.s003
(TIF)
S4 Fig. The patterning results for stomach of Erythroculter erythropterus on the SOM plane.
Colour scale is related to distances between map units. Black colours represent large distances and white colours represent small distances.
https://doi.org/10.1371/journal.pone.0328372.s004
(TIF)
S1 File.
S1 Table. Specific information of opening experiments conducted. S2 Table. Salinity (psu) distribution at the time of the maximum seawater inflow in the NRE. S3 Table. Fish assemblage in the NRE (2018 ~ 2022). ▲, exotic species; #, introduced species; ※ endemic species. S4 Table. Identified taxa in the stomach contents of Erythroculter erythropterus before (n = 12) and after (n = 12) the opening of the NRE.
https://doi.org/10.1371/journal.pone.0328372.s005
(ZIP)
References
- 1. Pinckney JL, Paerl HW, Tester P, Richardson TL. The role of nutrient loading and eutrophication in estuarine ecology. Environ Health Perspect. 2001;109 Suppl 5(Suppl 5):699–706. pmid:11677178
- 2. Elliott M, Whitfield AK. Challenging paradigms in estuarine ecology and management. Estuarine, Coastal and Shelf Science. 2011;94(4):306–14.
- 3. Day JW, Crump BC, Kemp WM, Yáñez‐Arancibia A. Estuarine Ecology. Wiley. 2012.
- 4. Martinho F, Leitão R, Viegas I, Dolbeth M, Neto JM, Cabral HN, et al. The influence of an extreme drought event in the fish community of a southern Europe temperate estuary. Estuarine, Coastal and Shelf Science. 2007;75(4):537–46.
- 5. Gillanders BM, Elsdon TS, Halliday IA, Jenkins GP, Robins JB, Valesini FJ. Potential effects of climate change on Australian estuaries and fish utilising estuaries: a review. Mar Freshwater Res. 2011;62(9):1115.
- 6.
Burt N, Rees A. Guidelines for the assessment and planning of estuarine barrages. London: Thomas Telford. 2001.
- 7. Kim JY, Kim G. Effects of regulated dam discharge on plants and migratory waterfowl are mediated by salinity changes in estuaries. Internat Rev Hydrobiol. 2020;106(1):58–63.
- 8. Geist J, Hawkins SJ. Habitat recovery and restoration in aquatic ecosystems: current progress and future challenges. Aquatic Conservation. 2016;26(5):942–62.
- 9. Jiang Y, Wang Y, Zhou D, Ke Y, Bai J, Li W, et al. The impact assessment of hydro-biological connectivity changes on the estuary wetland through the ecological restoration project in the Yellow River Delta, China. Sci Total Environ. 2021;758:143706. pmid:33250237
- 10. Wang Y, Li Z, Xie L, Pan Y, Wang R, Zhang Z, et al. Brackish water promote the ecological restoration of estuarine wetland. Ecological Engineering. 2023;187:106843.
- 11. Young GC, Potter IC, Hyndes GA, de Lestang S. The Ichthyofauna of an Intermittently Open Estuary: Implications of Bar Breaching and Low Salinities on Faunal Composition. Estuarine, Coastal and Shelf Science. 1997;45(1):53–68.
- 12. Weng X, Jiang C, Zhang M, Yuan M, Zeng T. Numeric Study on the Influence of Sluice-Gate Operation on Salinity, Nutrients and Organisms in the Jiaojiang River Estuary, China. Water. 2020;12(7):2026.
- 13. Loveridge A, Pitt KA, Lucas CH, Warnken J. Extreme changes in salinity drive population dynamics of Catostylus mosaicus medusae in a modified estuary. Mar Environ Res. 2021;168:105306. pmid:33839400
- 14. Laurent P, Perry SF. Environmental Effects on Fish Gill Morphology. Physiological Zoology. 1991;64(1):4–25.
- 15. Seebacher F, Webster MM, James RS, Tallis J, Ward AJW. Morphological differences between habitats are associated with physiological and behavioural trade-offs in stickleback (Gasterosteus aculeatus). R Soc Open Sci. 2016;3(6):160316. pmid:27429785
- 16. Sattari M, Namin JI, Bibak M, Vajargah MF, Hedayati A, Khosravi A, et al. Morphological comparison of western and eastern populations of Caspian kutum, Rutilus kutum (Kamensky, 1901)(Cyprinidae) in the southern Caspian Sea. International Journal of Aquatic Biology. 2018;6(4):242–7.
- 17. Shimada Y, Murakami N, Tsuzaki T, Seikai T. Changes in growth, higher salinity tolerance and behavioral traits in Japanese flounder Paralichthys olivaceus after repeated cross-breeding. Fisheries Sci. 2007;73(2):364–70.
- 18. van Leeuwen CHA, Museth J, Sandlund OT, Qvenild T, Vøllestad LA. Mismatch between fishway operation and timing of fish movements: a risk for cascading effects in partial migration systems. Ecol Evol. 2016;6(8):2414–25. pmid:27110352
- 19. Haraldstad T, Haugen TO, Kroglund F, Olsen EM, Höglund E. Migratory passage structures at hydropower plants as potential physiological and behavioural selective agents. R Soc Open Sci. 2019;6(11):190989. pmid:31827840
- 20. Landsman SJ, McLellan NR, Platts J, van den Heuvel MR. Fishway Effectiveness and Upstream Residency of Three Fish Species at Four Fishways in Prince Edward Island, Canada. Northeastern Naturalist. 2020;27(1):48.
- 21. Anger K. Effects of temperature and salinity on the larval development of the Chinese mitten crab Eriocheir sinensis (Decapoda: Grapsidae). Mar Ecol Prog Ser. 1991;72:103–10.
- 22. Duponchelle F, Panfili J. Variations in age and size at maturity of female Nile tilapia, Oreochromis niloticus, populations from man-made lakes of Côte d’Ivoire. Environmental Biology of Fishes. 1998;52(4):453–65.
- 23. Mérona B de, Mol J, Vigouroux R, Chaves P de T. Phenotypic plasticity in fish life-history traits in two neotropical reservoirs: Petit-Saut Reservoir in French Guiana and Brokopondo Reservoir in Suriname. Neotrop ichthyol. 2009;7(4):683–92.
- 24. Lloret-Lloret E, Albo-Puigserver M, Giménez J, Navarro J, Pennino MG, Steenbeek J, et al. Small pelagic fish fitness relates to local environmental conditions and trophic variables. Progress in Oceanography. 2022;202:102745.
- 25. Goren M, Galil BS. A review of changes in the fish assemblages of Levantine inland and marine ecosystems following the introduction of non-native fishes. J Appl Ichthyol. 2005;21(4):364–70.
- 26. Jensen E, Brucet S, Meerhoff M, Nathansen L, Jeppesen E. Community structure and diel migration of zooplankton in shallow brackish lakes: role of salinity and predators. Hydrobiologia. 2010;646(1):215–29.
- 27. Arimoro FO, Ikomi RB, Nwadukwe FO, Eruotor OD, Edegbene AO. Fluctuating salinity levels and an increasing pollution gradient on fish community structure and trophic levels in a small creek in the Niger delta, Nigeria. Int Aquat Res. 2014;6(4):187–202.
- 28. Rubio VC, Sánchez-Vázquez FJ, Madrid JA. Effects of salinity on food intake and macronutrient selection in European sea bass. Physiol Behav. 2005;85(3):333–9. pmid:15932763
- 29. Islam M, Ahsan DA, Mandal SC, Hossain A. Effects of salinity changes on growth performance and survival of rohu fingerlings, Labeo rohita (Hamilton, 1822). Journal of Coastal Development. 2014;17:379.
- 30. Zarri LJ, Palkovacs EP, Post DM, Therkildsen NO, Flecker AS. The Evolutionary Consequences of Dams and Other Barriers for Riverine Fishes. BioScience. 2022;72(5):431–48.
- 31. Park JM, Riedel R, Ju HH, Choi HC. Fish Assemblage Structure Comparison between Freshwater and Estuarine Habitats in the Lower Nakdong River, South Korea. JMSE. 2020;8(7):496.
- 32. Normile D. Environmental restoration. Restoration or devastation?. Science. 2010;327(5973):1568–70. pmid:20339043
- 33. Yoon J-D, Jang M-H, Jo H-B, Jeong K-S, Kim G-Y, Joo G-J. Changes of fish assemblages after construction of an estuary barrage in the lower Nakdong River, South Korea. Limnology. 2016;17(2):183–97.
- 34. Jo H, Jeppesen E, Ventura M, Buchaca T, Gim J-S, Yoon J-D, et al. Responses of fish assemblage structure to large-scale weir construction in riverine ecosystems. Sci Total Environ. 2019;657:1334–42. pmid:30677900
- 35. Choi H, Kim B, Lee G, Noh SJ. Enhancing Tidal Wave Predictions for the Estuary of the Nakdong River Using a Fixed-Lag Smoother. Energies. 2022;16(1):237.
- 36. Isanta-Navarro J, Hairston NG Jr, Beninde J, Meyer A, Straile D, Möst M, et al. Reversed evolution of grazer resistance to cyanobacteria. Nat Commun. 2021;12(1):1945. pmid:33782425
- 37.
Kim IS, Park JY. Freshwater Fishes of Korea. Seoul: Kyo-Hak Publishing Co. 2002.
- 38. Heo MA, An HC, Park MS, Yang YJ, Lee W-O. Species Composition of Fishes Collected by Fyke Net and Length-Weigth Relationships of Skygager (Erythroculter erythropterus) in Asan Lake and Chungju Lake. ISK. 2021;33(4):287–96.
- 39. Gal J-K, Ock G, Park H-K, Shin K-H. The effect of summer monsoon on pelagic and littoral food webs in a large regulated reservoir (Lake Paldang, Korea): A stable isotope approach. Journal of Freshwater Ecology. 2016;31(3):327–40.
- 40. Lim JY, Mamun M, Kim JY, An K-G. Spatial distribution and migration pattern analysis of Erythroculter erythropterus utilizing a telemetry approach in relation to water quality in the Nakdong River basin. Int J Environ Sci Technol. 2024;21(10):6935–46.
- 41. Werner EE, Hall DJ. Optimal Foraging and the Size Selection of Prey by the Bluegill Sunfish (Lepomis Macrochirus). Ecology. 1974;55(5):1042–52.
- 42. Oaten A, Murdoch WW. Switching, Functional Response, and Stability in Predator-Prey Systems. The American Naturalist. 1975;109(967):299–318.
- 43. Kumar R, Ramakrishna Rao T. Predation on Mosquito Larvae by Mesocyclops thermocyclopoides (Copepoda: Cyclopoida) in the Presence of Alternate Prey. International Review of Hydrobiology. 2003;88(6):570–81.
- 44. Kumar R, Muhid P, Dahms H-U, Tseng L-C, Hwang J-S. Potential of three aquatic predators to control mosquitoes in the presence of alternative prey: a comparative experimental assessment. Mar Freshwater Res. 2008;59(9):817.
- 45. Kumar R, Muhid P, Dahms H-U, Sharma J, Hwang J-S. Biological mosquito control is affected by alternative prey. Zool Stud. 2015;54:e55. pmid:31966142
- 46. Jeong K-S, Kim D-K, Shin H-S, Yoon J-D, Kim H-W, Joo G-J. Impact of summer rainfall on the seasonal water quality variation (chlorophyll a) in the regulated Nakdong River. KSCE Journal of Civil Engineering. 2011;15(6):983–94.
- 47. Park S, Yaan HS, Lee IC, Kim HT. Correlation between meteorological factors and water discharge from the Nakdong River barrage, Korea. Journal of the Korean Society of Marine Environment & Safety. 2008;14:111–7.
- 48.
Nelson JS. Fishes of the World. 4th ed. New York: John Wiley and Sons. 2006.
- 49. Wetzel RG, Likens GE. Composition and Biomass of Phytoplankton. Limnological Analyses. Springer New York. 2000. 147–74.
- 50. Noh HK, Ryu HK, Ryu JH, Kim HY, Chun JH. Status and plan of ‘Operation rule improvement and ecological restoration plan of Nakdong estuary. In: Proceedings of the Korea Water Resources Association Conference, 2020. 21–2.
- 51. Agrawal A, Gopal K. Analysis of Phytoplankton and Zooplankton: Qualitative and Quantitative. Biomonitoring of Water and Waste Water. Springer India. 2012. 33–40.
- 52.
Slotwinski A, Coman F, Richardson AJ. Introductory Guide to Zooplankton Identification. Brisbane: Integrated Marine Observing System. 2014.
- 53. Tagliapietra D, Sigovini M. Benthic fauna: collection and identification of macrobenthic invertebrates. Terre et Environnement. 2010;88:253–61.
- 54. Simpson EH. Measurement of diversity. Nature. 1949;163:688.
- 55. Shannon CE. A mathematical theory of communication. The Bell System Technical Journal. 1948;27:379–423.
- 56. Pielou EC. Shannon’s formula as a measure of specific diversity: its use and misuse. The American Naturalist. 1966;100:463–5.
- 57. Margalef DR. Information theory in ecology. International Journal of General Systems. 1958;3:36–71.
- 58. Hammer Ø, Harper DA, Ryan PD. PAST: Paleontological statistics software package for education and data analysis. Palaeontologia Electronica. 2001;4:9.
- 59. Liu G, Zhang S, Zhao X, Li C, Gong M. Advances and Limitations of Next Generation Sequencing in Animal Diet Analysis. Genes (Basel). 2021;12(12):1854. pmid:34946803
- 60. de Vargas C, Audic S, Henry N, Decelle J, Mahé F, Logares R, et al. Ocean plankton. Eukaryotic plankton diversity in the sunlit ocean. Science. 2015;348(6237):1261605. pmid:25999516
- 61. Albaina A, Aguirre M, Abad D, Santos M, Estonba A. 18S rRNA V9 metabarcoding for diet characterization: a critical evaluation with two sympatric zooplanktivorous fish species. Ecol Evol. 2016;6(6):1809–24. pmid:27087935
- 62. Massana R, Gobet A, Audic S, Bass D, Bittner L, Boutte C, et al. Marine protist diversity in European coastal waters and sediments as revealed by high-throughput sequencing. Environ Microbiol. 2015;17(10):4035–49. pmid:26119494
- 63. Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics. 2010;26(19):2460–1. pmid:20709691
- 64. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215(3):403–10. pmid:2231712
- 65.
Froese R, Pauly D. FishBase. 2023 [cited 2023 June 15]. https://www.fishbase.org
- 66. Kohonen T. Self-Organization and Associative Memory. Springer Berlin Heidelberg. 1989.
- 67. Park Y-S, Song M-Y, Park Y-C, Oh K-H, Cho E, Chon T-S. Community patterns of benthic macroinvertebrates collected on the national scale in Korea. Ecological Modelling. 2007;203(1–2):26–33.
- 68.
Vesanto J, Himberg J, Alhoniemi E, Parhankangas J. SOM toolbox for Matlab 5. 2000;57(2).
- 69. Chung JK, Kannappan PL, Ng CT, Sahoo PK. Measures of distance between probability distributions. Journal of Mathematical Analysis and Applications. 1989;138(1):280–92.
- 70. Froese R. Cube law, condition factor and weight-length relationships: history, meta-analysis and recommendations. J Appl Ichthyol. 2006;22(4):241–53.
- 71.
Anderson RO, Gutreuter SJ. Length, weight, and associated structural indices. In: Nielsen LA, Johnson DL. Fisheries techniques. Bethesda, MD: American Fisheries Society. 1983. 283–300.
- 72. Arantes CC, Fitzgerald DB, Hoeinghaus DJ, Winemiller KO. Impacts of hydroelectric dams on fishes and fisheries in tropical rivers through the lens of functional traits. Current Opinion in Environmental Sustainability. 2019;37:28–40.
- 73. Stich H-B, Lampert W. Predator evasion as an explanation of diurnal vertical migration by zooplankton. Nature. 1981;293(5831):396–8.
- 74. Burks RL, Lodge DM, Jeppesen E, Lauridsen TL. Diel horizontal migration of zooplankton: costs and benefits of inhabiting the littoral. Freshwater Biology. 2002;47(3):343–65.
- 75. Meerhoff M, Iglesias C, De Mello FT, Clemente JM, Jensen E, Lauridsen TL, et al. Effects of habitat complexity on community structure and predator avoidance behaviour of littoral zooplankton in temperate versus subtropical shallow lakes. Freshwater Biology. 2007;52(6):1009–21.
- 76. Mozetič P, Francé J, Kogovšek T, Talaber I, Malej A. Plankton trends and community changes in a coastal sea (northern Adriatic): Bottom-up vs. top-down control in relation to environmental drivers. Estuarine, Coastal and Shelf Science. 2012;115:138–48.
- 77. Kimmerer WJ, Thompson JK. Phytoplankton Growth Balanced by Clam and Zooplankton Grazing and Net Transport into the Low-Salinity Zone of the San Francisco Estuary. Estuaries and Coasts. 2014;37(5):1202–18.
- 78. Huber ED, Wilmoth B, Hintz LL, Horvath AD, McKenna JR, Hintz WD. Freshwater salinization reduces vertical movement rate and abundance of Daphnia: Interactions with predatory stress. Environ Pollut. 2023;330:121767. pmid:37146869
- 79. Attayde JL, Hansson L-A. Effects of nutrient recycling by zooplankton and fish on phytoplankton communities. Oecologia. 1999;121(1):47–54. pmid:28307888
- 80. Feniova IYu, Sakharova EG, Gladyshev MI, Sushchik NN, Gorelysheva ZI, Karpowicz M. Effect of Fish on the Transfer Efficiency of Carbon, PUFA, and Nutrients from Phytoplankton to Zooplankton under Eutrophic Conditions. Biol Bull Russ Acad Sci. 2021;48(8):1284–97.
- 81. Wang H, Bouwman AF, Wang J, Yu Z, Ran X. Competitive advantages of HAB species under changing environmental conditions in the coastal waters of the Bohai Sea, Yellow Sea and East China Sea. Continental Shelf Research. 2023;259:104991.
- 82. Ruoss JP, Mason BE, Wuellner MR, Koupal KD. Resetting the Clock: How a Reconnection Event Reshaped Cove Fish Assemblages in a Nebraska Reservoir. North American Journal of Fisheries Management. 2022;43(2):327–38.
- 83. Utne KR, Pauli BD, Haugland M, Jacobsen JA, Maoileidigh N, Melle W, et al. Poor feeding opportunities and reduced condition factor for salmon post-smolts in the Northeast Atlantic Ocean. ICES Journal of Marine Science. 2021;78(8):2844–57.
- 84. Waples RS, Zabel RW, Scheuerell MD, Sanderson BL. Evolutionary responses by native species to major anthropogenic changes to their ecosystems: Pacific salmon in the Columbia River hydropower system. Mol Ecol. 2008;17(1):84–96. pmid:18268786
- 85. Morita K, Suzuki T. Shifts of food habit and jaw position of white‐spotted charr after damming. Journal of Fish Biology. 1999;55(6):1156–62.
- 86. Dieterman DJ, Galat DL. Variation in Body Form, Taste Buds, and Brain Patterns of the Sicklefin Chub,Macrhybopsis meeki, in the Missouri River and Lower Yellowstone River, USA. Journal of Freshwater Ecology. 2005;20(3):561–73.
- 87. Nesteruk I, Passoni G, Redaelli A. Shape of Aquatic Animals and Their Swimming Efficiency. Journal of Marine Biology. 2014;2014:1–9.
- 88. Franssen NR, Harris J, Clark SR, Schaefer JF, Stewart LK. Shared and unique morphological responses of stream fishes to anthropogenic habitat alteration. Proc Biol Sci. 2013;280(1752):20122715. pmid:23235710
- 89. Eagderi S, Esmaeilzadegan E, Pirbeigi A. Morphological responses of Capoeta gracilis and Alburnoides eichwaldii populations (Cyprinidae) fragmented due to Tarik Dam (Sefidrud River, Caspian Sea basin, Iran). Iranian Journal of Ichthyology. 2014;1:114–20.
- 90. Gaston KA, Lauer TE. Morphometric variation in bluegill Lepomis macrochirus and green sunfish Lepomis cyanellus in lentic and lotic systems. J Fish Biol. 2015;86(1):317–32. pmid:25425144
- 91. Bianco PG. The Status of the Twaite Shad, Alosa agone, in Italy and the Western Balkans. Marine Ecology. 2002;23(s1):51–64.
- 92. Vajravelu M, Martin Y, Ayyappan S, Mayakrishnan M. Seasonal influence of physico-chemical parameters on phytoplankton diversity, community structure and abundance at Parangipettai coastal waters, Bay of Bengal, South East Coast of India. Oceanologia. 2018;60(2):114–27.
- 93. Chawla A, Jay DA, Baptista AM, Wilkin M, Seaton C. Seasonal Variability and Estuary–Shelf Interactions in Circulation Dynamics of a River-dominated Estuary. Estuaries and Coasts: J CERF. 2008;31(2):269–88.
- 94. Sarma VVSS, Krishna MS, Prasad VR, Kumar BSK, Naidu SA, Rao GD, et al. Distribution and sources of particulate organic matter in the Indian monsoonal estuaries during monsoon. JGR Biogeosciences. 2014;119(11):2095–111.
- 95. Yin K. Monsoonal influence on seasonal variations in nutrients and phytoplankton biomass in coastal waters of Hong Kong in the vicinity of the Pearl River estuary. Mar Ecol Prog Ser. 2002;245:111–22.
- 96. Lee Y-J, Lee S-J, Bae D-Y, An K-G. Fish Community Dynamics in the Artificial Fishways of Three Different Watersheds, and Fish Passage Evaluations Using a Multi-Metric Fishway Model. Pol J Environ Stud. 2019;28(5):3307–21.
- 97. Silva AT, Lucas MC, Castro‐Santos T, Katopodis C, Baumgartner LJ, Thiem JD, et al. The future of fish passage science, engineering, and practice. Fish and Fisheries. 2018;19(2):340–62.
- 98. Sun J, Galib SM, Lucas MC. Rapid response of fish and aquatic habitat to removal of a tidal barrier. Aquatic Conservation. 2021;31(7):1802–16.
- 99. Ysebaert T, van der Hoek D-J, Wortelboer R, Wijsman JWM, Tangelder M, Nolte A. Management options for restoring estuarine dynamics and implications for ecosystems: A quantitative approach for the Southwest Delta in the Netherlands. Ocean & Coastal Management. 2016;121:33–48.
- 100. Bouchard C, Boutron O, Lambremon J, Drouineau H, Lambert P, Nicolas D. Impacts of environmental conditions and management of sluice gates on glass eel migration. Estuarine, Coastal and Shelf Science. 2022;279:108139.
- 101. Bessert ML, Ortí G. Genetic effects of habitat fragmentation on blue sucker populations in the upper Missouri River (Cycleptus elongatus Lesueur, 1918). Conserv Genet. 2008;9(4):821–32.
- 102. Kidd IM, Fischer A, Chai S, Davis JA. A scenario-based approach to evaluating potential environmental impacts following a tidal barrage installation. Ocean & Coastal Management. 2015;116:9–19.
- 103. Elliott M, Day JW, Ramachandran R, Wolanski E. A Synthesis: What Is the Future for Coasts, Estuaries, Deltas and Other Transitional Habitats in 2050 and Beyond?. Coasts and Estuaries. Elsevier. 2019. 1–28.
- 104. Quinn TP, Bond MH, Brenkman SJ, Paradis R, Peters RJ. Re-awakening dormant life history variation: stable isotopes indicate anadromy in bull trout following dam removal on the Elwha River, Washington. Environ Biol Fish. 2017;100(12):1659–71.
- 105. Des Roches S, Post DM, Turley NE, Bailey JK, Hendry AP, Kinnison MT, et al. The ecological importance of intraspecific variation. Nat Ecol Evol. 2018;2(1):57–64. pmid:29203921
- 106. Fugère V, Hendry AP. Human influences on the strength of phenotypic selection. Proc Natl Acad Sci USA. 2018;115(40):10070–5.