Research targets and methods

Among the 22 intensive surveys of the WPAs over the last 5 years (2015-2019), we used the data for 8 wetlands in 2016, 2017 and 2019 in which fish surveys had been conducted, to understand the diversity and distribution characteristics of the fish species inhabiting the WPAs (Kim et al., 2016; 2017; Lim et al., 2019) (Fig. 1). These wetlands consisted of stream type wetlands (n=5) and marsh type wetlands (n=3) (Table 1).

Intensive surveys of the WPAs followed the monitoring guidelines of the Ministry of Environment for Korean inland wetlands (ME, 2011). In large areas where general fish investigation methods (e.g., cast net and skimming net) were not possible (e.g., the Han River estuary and Nakdong River estuary), we enlisted the help of fishermen to capture individuals using fixed shore nets, gill nets, dredge nets, and other methods. Species of captured individuals were either identified in the field and released after recording the required data, or were fixed with 10% formalin and identified in lab if identification in the field was not possible. Fish species were identified with reference to Chae et al. (2019), Choi et al. (2002), Kim et al. (2005), and Kim and Park (2002), and the taxonomic sequencing was based on the National Species List of Korea (NIBR, 2019).

Data analysis

We compiled a list of the fish species based on the reports of the intensive surveys conducted for the WPAs to analyze the diversity and distribution of the emergent fish in the WPAs. We analyzed the fish community characteristics and similarities using the diversity and population data of emergent fish, as well as the richness index (Margalef, 1958), dominance index (McNaughton, 1967), evenness index (Pielou, 1975), and species diversity index (Shannon & Weaver, 1963).

All statistical analyses were performed using R (R version 3.5.2, R Core Team 2020). First, we performed a hierarchical cluster analysis using the ‘hclust’ (method='ward.D2') function in the Vegan package to understand the relationship between wetland types (Bang et al., 2006) and the fish community structure (Oksanen et al., 2020). The total species population of the collected individuals (S) was converted to log(S+1) to be used as the input data, and the species that emerged in only one wetland were excluded from the analysis to eliminate regional specificity.

Second, we applied non-metric multidimensional scaling (NMDS) using the ‘metaMDS’ function in the Vegan package (Oksanen et al., 2020), and analyzed the fish community structure and relationship in the investigated sites using ordination methods. Again, we converted the species population data to log(S+1) to produce the input data for the analysis. We calculated the distance between the investigation sites using the Bray-Curtis method (Bray & Curtis, 1957). In addition, we determined the correlation coefficients and levels of significance using the ‘envfit’ function in the Vegan package (Oksanen et al., 2020) to analyze the eating behaviors of fish and other correlations in the 2-dimensional visualization of the NMDS results. We referred to the “Biomonitoring Survey and Assessment Manual” (ME, 2016a) and “Biomonitoring Survey and Assessment Manual: Estuary Ecosystem (fish)” (ME, 2017a) to determine the ecological characteristics of the fish species.

Emergent fish fauna in the WPAs

A total of 15 orders, 45 families, and 134 fish species were identified from the 12,972 individuals collected from the 8 WPAs. Specifically, 28 species were identified in the Han River estuary, 21 species in the Upo wetland, 60 species in the Nakdong River estuary, 17 species in the Gonggumji reservoir, 18 species in the Damyang stream, 9 species in the Duung wetland, 46 species in the Cimsil wetland, and 36 species in the Hanbando wetland. The most species appeared in the mouth of the Nakdong River estuary, and the proportion of marine fish species was high. Next, the second most abundant species appeared in the Cimsil wetland (Table 2).

The Nakdong River estuary, a partially-mixed estuary, has an estuary bank and coastal or bay characteristics that have enabled the high occurrence of marine fish species as well as the largest overall number of species. Blockage of the freshwater inflow by the estuary bank had gradually transformed the brackish water zone into a marine environment; which is consistent with the observations of previous studies (Kang et al., 2012).

On the other hand, the Han River estuary does not have an estuary bank, and the brackish water zone is extended due to the direct influence of the tides. The Han River estuary is a well-mixed estuary that maintains the typical ecosystem structure of estuaries along the Korean western coast (Hwang & Rhow, 2010), therefore, estuarine fish species were observed to be the dominant species. The species diversity of the Han River and Nakdong River estuaries differed greatly, with the latter estuary having a relatively higher occurrence of marine fish species.

The Cimsil wetland, which was found to support the second highest number of species, has a widely spread out brackish water zone due to the lack of an estuary bank, and the water quality is relatively high due to the lack of a source of atmospheric pollution. The diverse endemic species in the Cimsil wetland may largely influence the high proportion of endemic species in this area (Baek et al., 2013; Jang et al., 2009). Among the collected fish, 45 species (33.58%) of Cyprinidae, 12 species (8.96%) of Gobiidae, and 8 species (5.97%) of Cobitidae (Fig. 2A) were identified. In the observed population, the emergent fish species and their associated rates (from highest to lowest) were Zacco platypus (19.5%), Zacco koreanus (8.2%), Lepomis macrochirus (6.3%), and Carassius auratus (5.4%) (Fig. 2B). Among the emergent fish, there were 7 legally protected species: Acheilognathus signifer, Acheilognathus majusculus, Culter brevicauda, Gobiobotia brevibarba, Hemibarbus mylodon, Pseudopungtungia tenuicorpa, and Lethenteron reissneri, that made up 5.2% of the total emergent species; 4 exotic species: Carassius cuvieri, Cyprinus carpio, L. macrochirus, and Micropterus salmoides, that made up 3.0% of the total emergent species; and 31 Korean endemic species: including Z. koreanus, Coreoperca herzi, Squalidus chankaensis tsuchigae, Odontobutis interrupta, A. signifer, and Acheilognathus yamatsutae, that made up 23.1% of the total emergent species (Fig. 3A). There were 36 brackish water species: including Hyporhamphus intermedius, Hyporhamphus sajori, Konosirus punctatus, and Thryssa kammalensis, that made up 26.9% of the total emergent species; 4 migratory species: Anguilla japonica, Coilia mystus, Coilia nasus, and Plecoglossus altivelis, that made up 3.0% of the total emergent species, 36 species of marine species: including Conger myriaster, Muraenesox cinereus, and Strongylura anastomella, that made up 26.9% of the total emergent species; and 12 species of riffle-benthic species, including Coreoleuciscus splendidus, G. brevibarba, and Koreocobitis rotundicaudata, that made up 9.0% of the total emergent species (Fig. 3B).

After analyzing the eating behavior and tolerance characteristics using the National Institute of Environmental Research (2017). Biomonitering Survey and Assessment Manual, Incheon: National Institute of Environmental Research. We identified 16 carnivorous: including A. japonica, C. brevicauda, Erythroculter erythropterus, and Channa argus, that made up 11.9% of the total emergent species; 7 herbivorous: including H. intermedius, H. sajori, and K. punctatus, that made up 5.2% of the total emergent species; 41 insectivorous: including Cobitis hankugensis, C. splendidus, and Repomucenus olidus, that made up 30.6% of the total emergent species; and 29 Omnivorous: including Oryzias sinensis, Abbottina rivularis, and Acheilognathus koreensis, that made up 21.6% of the total emergent species (Fig. 3C). In addition, there were 33 tolerant species: including O. sinensis, A. rivularis, and C. auratus, that made up 24.6% of the total emergent species; 44 species with intermediate tolerance: including A. japonica, H. intermedius, and C. nasus, that made up 32.8% of the total emergent species; and 16 sensitive species: including A. signifer, C. splendidus, and P. tenuicorpa, that made up 11.9% of the total emergent species (Fig. 3D).

Fish community indexes and similarity analysis

Among the investigated WPAs, the Nakdong River estuary and Duung wetland had the highest and lowest Diversity and Richness indexes of 3.0 and 7.77 and 1.45 and 1.68, respectively. The Han River estuary, Upo wetland, and Cimsil wetland had similar Evenness indexes of 0.75, 0.75, and 0.74, respectively, and the Duung wetland had the lowest Richness index of 1.68. On the other hand, the Duung wetland had the highest Dominance index of 0.73 because its small area and lack of water inflow resulted in the domination of C. auratus (Fig. 4).

The NMDS and hierarchical cluster analysis of fish communities in the eight WPAs showed that the wetlands can be classified into four categories: estuary-type, stream-type, lake-type, and coastal dune-type (Figs. 5, 6). The Han River and Nakdong River estuary-type wetlands were placed on the left side of the first axis, and were found to have a high incidence of marine fish species and brackish water species, such as H. intermedius, Chelon haematocheilus, and Synechogobius hasta (Fig. 6). It is noteworthy that herbivorous fish species dominated the estuary-type wetlands. The Cimsil and Hanbando wetlands (i.e., typical stream-type wetlands) were placed on the upper-right side of the first axis, and were found to have a high incidence of endemic species and riffle-benthic species, such as Z. koreanus, Sarcocheilichthys variegatus wakiyae, and C. herzi. The Upo and Gonggumji lake-type wetlands that are created by floodplains, were placed on the lower-right side of the first axis, and found to be dominated by exotic species, such as M. salmoides and L. macrochirus. On the other hand, the Damyang stream wetland was considered to be better characterized as a stream-lake-type wetland (e.g., the Upo and Gonggumji wetlands) than a stream-type wetland because a significant amount of this stream wetland has been lost by development activities, such as the renovation of a dammed pool for irrigation, installation of a detention pond, construction of artificial wetlands, and large-scale engineering works. The loss of this stream wetland has resulted in water pollution and eutrophication, and parts of the retained wetland are undergoing various geomorphic changes and rapid successional changes that are driving a change in the wetland type.

In conclusion, this study was based on the results of fish-related investigations for 11 sites included in the intensive surveys on WPAs in South Korea. Currently, 44 sites (1,553.13 km²) are designated as WPAs; however, studies that focus on emergent fish and their distributions despite the variations in shape, area, and type of WPAs are seriously limited

Analysis of the fish investigation data from the intensive surveys on WPAs revealed that there are four wetland types and one combined wetland type, and that the wetland types differed in species composition. Additionally, the predominant species Z. koreanus and Z. platypus contributed the most to the similarity of the fish communities, and are reportedly dominant species in domestic streams (Yoon et al., 2011). Tolerant or exotic species had high incidences in lakes, which is similar to the findings of a previous study where the proportion of omnivorous and tolerant species increased with the increase in stagnant water (Han & An, 2013).

Such environmental changes in wetlands have direct influences on fish communities; thus, it is high possible that they will be used as the main indexes in the management of wetland types. However, fish communities are influenced by a combination of waterfront vegetation, differences in water quality, and physical characteristics of river beds. It is, therefore, difficult to use limited data for fish communities to develop a comprehensive management plan for wetlands. This is especially relevant regarding estuarine WPAs, in which fish communities vary depending on the geomorphic characteristics of habitats and the salinity of the water source (e.g., freshwater, brackish water, saltwater) (Baek et al., 2013; Kim et al., 2018; Park et al., 2013). Data supplementation and an improvement of our understanding through the classification of geomorphological environment types and analysis of biota including fish communities within inland wetlands is necessary for the overall management of wetlands.