Response of Fish Community to Building Block Methodology Mimicking Natural Flow Regime Patterns in Nakdong River in South Korea

Abstract

Water regulation and flood control of rivers are changing due to streamflow depletion following industrialization and urbanization, significantly impacting aquatic ecosystems. Therefore, restoration of the ecological environment is necessary to maintain a healthy river ecosystem. For ecosystem restoration, the amount of discharge from dams must be controlled and the appropriate environmental flow must be calculated according to the fish species. The change in the flow through the dam due to hydropeaking directly impacts the fish. This study aimed to construct a building block methodology (BBM) using dam inflows in the Gudam Bridge basin upstream of the Nakdong River, build a River2D model of this area, and calculate the natural flow regime and the weighted usable area (WUA). The analysis of the scenarios for the whole period (2006–2020) and by flow regime showed that WUA decreased in some periods, but improved overall in the scenario reflecting the BBM. For Zacco platypus, a dominant fish species of the Gudam Bridge, WUA decreased by ~11% in some periods (in September) but the habitat improvement effect measured up to 79%. Changing the dam discharge pattern by considering the flow regime seemed more effective in improving the habitat of fish living downstream.

1. Introduction

Following industrial development and urbanization, the amount of river water has gradually decreased, and the quality of the water has deteriorated. The appearance of natural rivers has been artificially changed by the intensifying water regulation and flood control functions of rivers. As this has a great impact on the habitat of biota inhabiting rivers, it is necessary to create an appropriate river environment for ecological restoration. Understanding the appropriate habitat and physical characteristics of fish is important for conserving biodiversity and maintaining species abundance [1,2].

In order to restore aquatic ecosystems, it is important to control the amount of discharge from dams and estimate the appropriate environmental flow for most of the fish species. Dams not only provide a stable water supply by regulating the flow of rivers, but also help supply energy with hydroelectric power generation. Nevertheless, as the changes in flow caused by dams (such as power generation discharges) directly affect the habitats of organisms living downstream, there is an urgent need to study habitat changes caused by dam discharges.

The United States has made a major contribution to the development and application of in-stream flow estimation methods. As of 2002, 207 methods for estimating in-stream flow had been developed in 44 countries around the world, of which 77 methods had been developed and applied in the United States [3]. Researchers from the United States have predicted environmental ecological flow through habitat simulation techniques since the 1980s, and Mathews and Richter evaluated the ecological functions of each natural flow regime [4]. In Japan, the flow required to maintain the normal function of flow in rivers is defined as “maintenance flow,” and changes in the habitat and growth environment of aquatic plants are taken into account to calculate the in-stream flow [5].

Research on in-stream flow for habitat conservation of aquatic organisms started in the 1970s, and it has been commonly used for river management, including dam operation, water regulation, and catchment management [6]. The importance of fluctuations in flow regimes emerged in the field of in-stream flow research in the 2000s [7]. In South Korea, the assessment of the environmental flow using the Physical Habitat Simulation Model (PHABSIM) started in the 1990s, and eco-hydraulic modeling using River2D started in the mid-to-late 2000s [8].

River2D, a two-dimensional model, has been widely used to simulate the relationship between flow and fish habitat [9]. Several researchers have calculated the optimal environmental flow using the results of the physical habitat simulation for certain points. These results have also been used to plan river restoration [10,11], evaluate river health [12,13], and assess the impact of river works [14,15,16]. Recently, there have been studies on the ecological health of streams and management of water resources; furthermore, studies have analyzed the correlation between ecologic indexes and flow metrics to evaluate the health of aquatic ecosystems [17,18].

To estimate the environmental flow suitable for river aquatic ecosystems, it is necessary to study the discharge patterns (reflecting the flow regime characteristics) of dams located in the upstream section, their ecological impact, and the habitable areas. In this study, the Nakdong River basin, where two dams, the Imha and Andong Dams, are located upstream, was selected as the target area to analyze the changes caused by the building block methodology (BBM) mimicking the natural flow regime in the habitat of the fish community. We considered that it would be easy to analyze the effects of BBM as the two dams are located upstream. The desalination of the Imha and Andong Dams started in December 1991 and December 1975, respectively. The results of past aquatic ecology monitoring analysis based on the Nakdong River basin, which includes the target area of this study, indicate that many of the fish species that inhabited it in the past have disappeared [19,20,21,22,23]. The completion and hydropeaking of the Andong Dam and Imha Dam located upstream of the Nakdong River appear to have caused changes in fish species inhabiting the downstream section by shortening the favorable period for fish habitat suitability [24]. Therefore, it is necessary to minimize the impact of the construction and discharge patterns of the dam on the aquatic ecosystems and to establish a countermeasure for the operation of the dam by numerically analyzing the physical habitat of fishes. Since South Korea has four different seasons with a considerable effect of precipitation on the watershed, the efficient operation of hydraulic infrastructure, such as dams, is critical. As the flow regime acts as an important determinant of the physical habitat conditions of organisms in rivers [25,26], and changes in the natural flow regime in downstream rivers due to multi-purpose dams have a great ecological impact, it is necessary to study the dam operation plan considering the flow regime characteristics [27]. In South Korea, research related to flow regime analysis started with the flow regime analysis of major rivers in 1906, and recently, studies on the changes in flow regime caused by dam construction have been conducted [28,29]. Previously, the calculation of the flow duration curve and flow variation coefficient for each major point of the river was mainly conducted, with a scarcity of analysis on the ecological characteristics according to the flow regime in relation to the operation of the dam [27]. For studying river ecosystems, an integrated approach is required, which involves flow control by hydraulic structures based on the integrated approach of flow regime.

In this study, BBM flow pattern was constructed using a natural flow regime and used as input data for the model to analyze changes in the habitat of fish species. BBM is a type of holistic approach, developed in South Africa in the 1990s, that enables the consideration of monthly flow regime characteristics [7,30]. Alfredsen et al. (2012) [31] configured three stages (low-, medium-, and high-flow scenarios) using BBM in the Daleelva river (Norway), and analyzed the environmental flows at all three stages. Bakken et al. (2013) [32] conducted a study in the Godavari basin in India evaluating the method using a modified BBM as a tool for water allocation for irrigation, drinking water supply, and hydroelectric power generation in multi-purpose reservoirs. Bahukandi and Ahuja (2013) [33] argued the need for biodiversity maintenance and environmental ecological flow management in the Dehradun River basin in India based on the opinions of various experts in environmental flow (E-Flow; described as the quantity, timing, and quality of freshwater flows and levels necessary to sustain aquatic ecosystems) and devised a comprehensive strategy [34,35]. Kang et al. (2016) [36] evaluated the in-stream flow by estimating the monthly flow that can reflect the diversity of the flow regime by modifying the BBM to account for the domestic in-stream flow calculation guidelines and institutional aspects.

Similarly, in the BBM framework, weighted usable area (WUA) estimation using the River2D model enables numerical analysis of fish habitats and aquatic ecosystem health according to flow regime. In this study, BBM was developed by combining various flow patterns resembling an existing natural flow regime using past hydrologic data. This is simpler than the natural flow regime, but it is different from existing hydropeaking in that it mimics a pattern. The fact that the duration can be applied differently for each flow range is more significant than maintaining the same flow every month [5]. Therefore, this study aimed to (1) construct a BBM for dry years, normal water years, and wet years by analyzing the past hydrological data in the areas upstream of the Nakdong River; (2) build a River2D model of the study area; and (3) calculate the natural flow regime for two fish species in the study area, as well as the weighted usable area (WUA) according to BBM. Figure 1 shows the flowchart of this study.

2. Materials and Methods

2.1. Study Area

The Gudam Bridge basin downstream of Andong Dam was selected as the study area, where Andong Dam and Imha Dam (the multi-purpose dams) are located upstream of the Nakdong River basin. The model was built for approximately 8.5 km, including the Gudam Bridge located downstream of the Andong Dam. The Gudam Bridge basin was selected as the study area due to the advantage of being able to establish simulated input conditions with the presence of both flow and water level observation stations and a biological monitoring network, and to consider the effect of dam discharge control on securing in-stream flow with Andong Dam and Imha Dam. Figure 2 shows the map of the study area, including Andong Dam and Imha Dam. Figure 3 is a graph of the annual mean discharge from the two dams.

2.2. River2D Model

In this study, the River2D model, enabling hydraulic analysis and habitat simulation, was used to analyze the physical habitat of fish species [28,37,38]. The River2D model is a two-dimensional flow analysis model that uses a finite element scheme to analyze the two-dimensional shallow water equation, which consists of the continuity equation and the momentum equation as follows. Equation (1) represents the conservation of mass and Equation (2) represents the conservation of momentum.

$$\frac{\partial H}{\partial t}+\frac{\partial q_x}{\partial x}+\frac{\partial q_y}{\partial y}=0$$

(1)

$$\begin{gathered} \frac{\partial q_x}{\partial t}+\frac{\partial }{\partial x}\left(U{q}_x\right)+\frac{\partial }{\partial y}\left(V{q}_x\right)+\frac{g}{2}\frac{\partial H^2}{\partial x} \\ =gH\left(S_{ox}-S_{fx}\right)+\frac{1}{\rho }\left\{\frac{\partial }{\partial x}\left(H{\tau }_{xx}\right)\right\}+\frac{1}{\rho }\left\{\frac{\partial }{\partial y}\left(H{\tau }_{xy}\right)\right\} \end{gathered}$$

(2)

$$\begin{gathered} \frac{\partial q_x}{\partial t}+\frac{\partial }{\partial x}\left(U{q}_y\right)+\frac{\partial }{\partial y}\left(V{q}_y\right)+\frac{g}{2}\frac{\partial H^2}{\partial y} \\ =gH\left(S_{oy}-S_{fy}\right)+\frac{1}{\rho }\left\{\frac{\partial }{\partial x}\left(H{\tau }_{yx}\right)\right\}+\frac{1}{\rho }\left\{\frac{\partial }{\partial y}\left(H{\tau }_{yy}\right)\right\} \end{gathered}$$

(3)

where t denotes the time; x and y denote the flow direction and transverse coordinates, respectively; H denotes the depth; u and v denote the depth-averaged flow velocity in the x and y directions, respectively; q_x and q_y denote the flows per unit width in the x and y directions, respectively; S_ox and S_oy denote the bottom slopes in the x and y directions, respectively; $S_{fx}$ and $S_{fy}$ denote the friction gradients in the x and y directions, respectively; T_ij denotes the turbulent stress tensor; p denotes the water density; and g denotes the gravitational acceleration. The friction gradients in the x and y directions in Equations (2) and (3) were calculated as in Equation (4) below.

$$S_{fx}=\frac{n^{2}u\sqrt{u^2+v^2}}{H^{4/3}}, S_{fy}=\frac{n^{2}v\sqrt{u^2+v^2}}{H^{4/3}}$$

(4)

where n represents Manning’s roughness coefficient.

In this study, the suitability for the fish species was analyzed by physical habitat simulation based on the input information, such as water depth and flow velocity in the cross section for each flow, and the habitat suitability index (HSI). HSI can describe the suitability of the environment for a species and the distribution of that species accordingly. WUA was calculated using Equation 5 as follows:

$$\mathrm{WUA}={\displaystyle \sum }_{i=1}^{n}CS{I}_i{A}_i$$

(5)

where A_i denotes the area of the finite element corresponding to cell i and CSI_i denotes the composite suitability index (CSI) in cell i.

2.3. Fish Species and HSI

HSI was used as the input data for the habitat model of organisms as it quantitatively expresses the relationship between physical habitat characteristics and the preference of the species in terms of flow velocity and water depth. In general, the factors affecting the habitat of fish are not only physical factors such as depth and velocity, but also chemical factors and biological factors, such as nutrients and toxicity [39]. Particularly, as the periphytic algae grow under the influence of nutrients such as nitrogen and phosphorus, their growth is strongly correlated with the concentrations of total phosphorus and total nitrogen [40]. Each element varies with the characteristics of each water supply and river. In order to calculate the HSI, physical factors, water chemistry, and biotic variables should be considered together. Although suitable fish species are classified by water quality in the Framework Act on Environmental Policy in Korea, there are no quantitative data on fish habitat suitability related to biological and chemical factors [41]. In addition, the CSI of the River2D model is calculated as a combination of individual suitability indices for depth, velocity, and channel index [28]. Although in this study we only considered the water depth, velocity, and substrate, the HSI was constructed using the data obtained through on-site monitoring to increase the accuracy of the simulation. The closer the value of the suitability index (SI) to 1, the more suitable are the flow velocity and water depth for the species. The SI value of 0 indicates the least suitable habitat conditions. In addition, since the same fish species could have different HSIs depending on the river, accurate HSI calculations require expert judgment, on-site investigation, and analysis, as well as the HSI surveyed in the stream adjacent to the target stream [29]. The closer the value of the HSI is to 1 for each physical habitat characteristic, the more suitable the habitat environment is for the fish species. The HSI of fish is based on the number of fish species that appeared in the survey area. With the maximum number of individuals in the survey area set to 1, the SI was calculated as a ratio.

In order to calculate the environmental flow of a fish species using the River2D model, the HSI of the selected fish species in the survey area must be input. Therefore, in this study, HSI was established for the dominant and sub-dominant fish species selected through on-site monitoring. The HSI is obtained based on the number of fish species that appear in a specific monitoring point [42]. In this study, the maximum value was set to 1 by combining the maximum number of species that appeared during the survey in the Gudam Bridge, and the remainder values were set as relative ratios to the maximum value. The HSI was calculated by applying the binary method to substrate and a univariate curve to depth and velocity.

The HSIs of the dominant and sub-dominant fish species were established through field surveys. Since the same fish species could have different HSIs in each water system and stream, it was preferable to input the HSI investigated in the water system or stream adjacent to the survey area.

As for the Nakdong River basin including Gudam Bridge, the study area, a number of monitoring analyses and related ecological studies have been conducted. The results of previous studies on aquatic ecology monitoring analysis in the Nakdong River basin, including the Gudam Bridge, show that several fish species that appeared in the 1990s disappeared after the 2000s. Representative fish species that gradually disappeared include Rhodeus uyekii, Acheilognathus, Gobiobotia naktongensis, and Microphysogobio rapidus (see Figure A1 and

Table A1. Fish species appearing in Nakdong River Basin including the target area of this study.

Species Name	Time of Appearance
	1990s	2000s	2010s
Lethenteron reissneri			O
Carassius	O	O	O
Macropodus opercularis	O	O
Rhodeus uyekii	O
Cyprinus carpio	O		O
Pungtungia herzi	O	O	O
Squalidus gracilis majimae	O	O	O
Squalidus chankaensis tsuchigae	O	O	O
Coreoleuciscus splendidus	O	O	O
Pseudorasbora parva	O	O	O
Hemibarbus labeo		O	O
Hemibarbus longirostris	O	O	O
Belligobio			O
Pseudogobio esocinus	O	O	O
Microphysogobio yaluensis	O	O	O
Nipponocypris	O	O	O
Zacco platypus	O	O	O
Opsariichthys		O	O
Cultriculus eigenmanni			O
Rhynchocypris oxycephalus	O	O	O
Barbatula nuda			O
Cobitis koreensis			O
Misgurnus anguillicaudatus	O	O	O
Cobitis taenia	O	O	O
Koreocobitis	O		O
Niwaella			O
Niwaella multifasciata	O		O
Silurus asotus	O	O	O
Silurus microdorsalia			O
Liobagrus andersonii			O
Liobagrus mediadiposalis	O		O
Plecoglossus altivelis			O
Cottus pollux			O
Rhinogobius brunneus	O		O
Acheilognathus yamatsutae	O	O
Acheilognathus koreensis	O	O
Tanakia lanceolata	O
Acheilognathus macropterus	O	O
Rhodeus ocellatus	O
Sarcocheilichthys variegatus wakiyae	O	O
Acheilognathus	O
Abbottina springeri	O
Acheilognathus	O	O
Micropterus salmoides	O	O
Lepomis macrochirus	O	O
Odontobutis	O	O	O
Microphysogobio rapidus	O
Aphyocypris	O
Iksookimia longicorpa	O
Leiocassis ussuriensis	O
Rhinogobius giurinus	O	O
Gymnogobius urotaenia	O	O
Tridentiger brevispinis	O	O
Channa argus	O	O	O
Hemiculter eigenmanni	O
Culter alburnus	O
Misgurnus mizolepis	O
Pelteobagrus fulvidraco	O	O
Leiocassis ussuriensis	O
Coreoperca herzi	O	O	O
Siniperca scherzeri	O	O	O
Rhinogobius giurinus	O		O
Rhinogobius brunneus	O		O
Brachymystax lenok		O	O
Gobiobotia nakdongensis	O	O

Species that appeared at the time: O.

in Appendix A) [17,18,19,20]. Therefore, in this study, the field monitoring was performed to estimate the environmental flow reflecting the current state of the Gudam Bridge basin. The fish species, physical characteristics (water depth and flow velocity), and bed materials were investigated in the field survey conducted during the dry season (March to June 2021), when a decrease in river flow could affect the ecosystem. For the bed materials, 1 represents silt (0.062 mm or less), 2 represents sand (0.062 to 2.0 mm), 3 represents granules (2.0 to 16.0 mm), 4 represents pebbles (16.0 to 64.0 mm), 5 represents cobbles (64.0 to 256.0 mm), and 6 represents boulders (over 256.0 mm).

In this study, a total of four field surveys were conducted at a point located approximately 10 km downstream from the Gudam Bridge. Based on the results of the field survey, Zacco platypus was selected as the dominant species with the highest number of individuals, and Pseudogobio esocinus was selected as the sub-dominant species. A large Zacco platypus population was observed at a flow velocity of 0.2 to 0.45 m³/s, and a large Pseudogobio esocinus population appeared at a flow velocity of 0.3 to 0.4 m³/s. The bed materials constituting their habitat included sand (0.062 to 2.0 mm) and granules (2.0 to 16.0 mm). Figure 4 and Figure 5 below show the HSIs of Zacco platypus and Pseudogobio esocinus calculated according to the number of individuals. Although Zacco platypus and Pseudogobio esocinus did not show a significant difference in terms of their preferred water depth or flow velocity, great significance lay in applying the HSI constructed through field surveys for simulation in this study.

2.4. Building Block Methodology (BBM)

The BBM was introduced to modify dam operations using hydrologic and ecological data [43]. It is one of the holistic methods used for estimating environmental flows [44,45]. Other methods include static minimum flow allocation, percentage of flow allocation, seasonally adjusted minimum flow allocation, and seasonally adjusted minimum flow allocation with seasonal flushing flow.

In this study, inflow data from Andong Dam and Imha Dam, located upstream of the Gudam Bridge, were used to construct a dam discharge scenario that mimics the natural flow regime patterns. The methods of calculating the average flow by period and calculating the representative flow by duration were used to construct the scenarios. For the unification of the period, analysis was performed using hydrological data for 15 years from 2006 to 2020, when the data for both the inflow and outflow of Andong Dam and Imha Dam, as well as the flow of Gudam Bridge, were established. The average flow for each period was calculated as the flow range using the monthly average of past hydrological data, and the flow range for each period was calculated through the analysis of the duration. In South Korea, with clear seasonal changes, using the monthly average flow is an effective way to represent the river flow. The average flow for the whole period was calculated, and the year with the lowest flow was designated as the dry year, the year with the highest flow as the wet year, and the year with the average flow as the normal water year. The flow range was calculated using monthly average flows from each year.

This study configured two main scenarios: Scenario 1 consisted of habitat simulations based on the flow range using the monthly average flow of the Gudam Bridge basin; and Scenario 2 consisted of discharge simulations. Scenario 2 builds a BBM to determine the natural flow pattern by assuming natural flow for the inflow into Andong and Imha Dam. Five scenarios that reflected the characteristics of the entire period and each flow regime were constructed using BBM flow (

Table 1. Scenario configuration.

Division		Contents
Scenario 1		Whole period, Monthly (Observed Flow)
Scenario 2	2-1	Whole period, Range (BBM)
	2-2	Whole period, Average Monthly (BBM)
	2-3	Wet year, Monthly (BBM)
	2-4	Dry year, Monthly (BBM)
	2-5	Normal year, Monthly (BBM)

). Table 1 shows the scenarios and their descriptions. Scenario 2-1 was calculated by analyzing the flow range and duration of past hydrological data, and Scenario 2-2 consisted of the flow range calculated using monthly average flows. Scenarios 2-3 and 2-5 consisted of the flow range calculated using the monthly average flow data for the years selected as the wet year, dry year, and normal water year.

3. Results

3.1. Flow Regime Analysis and Flow Calculation by Scenario (BBM Calculation)

In this study, the BBM was constructed using the flow of the Gudam Bridge and the inflow data of Andong Dam and Imha Dam located upstream. The wet year, dry year, and normal water year were designated based on the hydrological data of the actual flow of the Gudam Bridge and the inflow data of Andong Dam and Imha Dam from 2006 to 2020, and the BBM discharge scenarios were constructed for the whole period, wet year, dry year, and normal water year. As the flow regime in rivers acts as an important determinant of the physical habitat conditions of organisms, BBM was constructed as a dam discharge scenario that mimics the natural flow regime. The shape of flow duration curves (FDCs) is affected mainly by the reservoir, land cover, and daily upstream water use [46]. Figure 6 shows the FDCs created using the BBM flows for the whole period, wet year, dry year, and normal water year. By using the FDC formula, the overall flow condition can be determined probabilistically. To create the FDC curve, the observed and BBM flow data were arranged in the order of the maximum flow to the minimum flow, and the number of days exceeding a specific flow rate was calculated as a percentage. In Figure 6, the basic flow regime curves show the FDC graph calculated for the observed value and BBM flow rate with the X-axis depicting the time percentage as a cumulative frequency distribution and the Y-axis depicting flow data corresponding to the time percentage. Equation 6 is an expression of the flow arrangement order as a percentage. An FDC known as a probability curve of percentage exceedance is a graphical tool for describing streamflow [46].

Percentage of days flow exceeded (%) = rank/number of data × 100

(6)

After arranging the BBM flow from the maximum flow to the minimum flow and calculating the flow regime, the results showed that BBM flow was lower in all scenarios (wet, normal, dry) except for the whole range. In addition, since the WUA was larger in the BBM flow rate, it is more effective as a habitat area when the discharge pattern is changed, and efficient river management is possible even with a lower flow.

The figures below show the flow for each scenario. Figure 7 is a graph showing the daily flow at Gudam Bridge and the BBM calculated when Scenario 1 considered the flow of Gudam Bridge as the observed flow and Scenario 2 considered the location of the dams to compose a dam discharge scenario mimicking the natural flow-regime. Figure 8 and Figure 9 show the inflow, discharge, and BBM flow of Andong Dam and Imha Dam, respectively (using hydrological data for 15 years from 2006 to 2020). Data obtained using the Water Resources Management Information System [47] were used for daily dam discharge.

3.2. River2D Model Construction and Calibration

In this study, a River2D model was constructed for an approximately 8.5 km section of the Gudam Bridge downstream of the Andong Dam and Imha Dam to simulate habitat changes due to the natural flow regime (Figure 10). To verify the accuracy of the numerical simulation results using the River2D model, the results were compared with the Hydrologic Engineering Center River Analysis System (HEC-RAS) simulation results, which were verified through the water level observation results. The comparison of the results showed that there was a river bed with a very steep slope, resulting in errors at some upper and lower points. The error range was within 2%, indicating that the simulation had progressed well. Figure 11 shows the verification result of the model, and Figure 12 shows the model construction result. In Figure 11, the channel thalweg is the curve connecting the deepest points of the river [48].

3.3. Comparison of WUA Results According to Natural Flow Regime and BBM

The BBM calculated using the actual values of the Gudam Bridge was used for Scenario 1, and the sum of the BBMs for each flow regime of Andong Dam and Imha Dam was used for Scenario 2, considering the topographical characteristics of the dams located upstream of the Gudam Bridge, flowing from the dams into the main stream of the Nakdong River. The difference between the flow from the dams and from the Gudam Bridge was calculated and reflected in the model, considering the distance to the Gudam Bridge discharge point scenario constructed in consideration of the natural flow regime flowing into the upstream of the dam. Physical habitat analysis was performed using the River2D model for the configured Scenario 1 and Scenario 2. Figure 13 and Figure 14 show the results of the physical habitat analysis performed for Zacco platypus, the dominant species, and Pseudogobio esocinus, the sub-dominant species, respectively. In order to represent the WUA, the flow and WUA for each scenario for each day were indicated on the graph. In the wet year, the WUA of Zacco platypus indicated a habitat improvement effect of 29.51–79.20% in Scenario 2; and the WUA of Pseudogobio esocinus indicated a habitat improvement effect of 43.07–78.11% in Scenario 2, reflecting the BBM. In the dry year, for Zacco platypus, there was a period with a habitat reduction effect of about 15% or more in Scenario 2, reflecting the BBM, but the WUA of Zacco platypus showed a habitat improvement effect of up to 29.51%. Moreover, for Pseudogobio esocinus, there was a period with a habitat reduction effect of about 10% in Scenario 2, reflecting the BBM, but it showed a habitat improvement effect of up to 43.64%. In the normal water year, there was a period with a habitat reduction effect of about 11% or more in Scenario 2, reflecting the BBM, but the WUA of Zacco platypus showed a habitat improvement effect of up to 77.89%. Moreover, for Pseudogobio esocinus, there was a period with a habitat reduction effect of about 18% in Scenario 2, reflecting the BBM, but it showed a habitat improvement effect of up to 44.45%. Based on the monthly average BBM of the whole period, the WUA of Zacco platypus indicated a habitat improvement effect of 30.90–67.26% and the WUA of Pseudogobio esocinus indicated a habitat improvement effect of 43.07–58.65% in Scenario 2, reflecting the BBM. Based on the periodic BBM of the whole period, there was a period with a habitat reduction effect of about 10.46%, but the WUA of Zacco platypus showed a habitat improvement effect of up to 30.90%. There was no period with a habitat reduction effect for Pseudogobio esocinus, indicating a habitat improvement effect of up to 58.65%. Based on the monthly average BBM, there was no period with a habitat reduction effect for Zacco platypus, indicating a habitat improvement effect of up to 67.26%. For Pseudogobio esocinus, there was a period with a habitat reduction effect of about 1%, but it showed a habitat improvement effect of up to 43%. Figure 15 is a graph showing the monthly efficiency of each scenario of Zacco platypus and Pseudogobio esocinus. The efficiency decreased in some periods but increased in most periods. When comparing the results of the two scenarios, there were some periods showing the effect of habitat reduction, but overall, an improvement in the habitat was observed with the application of the scenario, thereby reflecting the BBM. The present simulation results indicate that the impact of the flood on the habitat suitability changes depending on the stream morphology. Moreover, the impact of the flow regime on the habitat suitability changes depending on the stream morphology. This result is consistent with the findings of previous studies [49,50,51,52,53,54].

4. Discussion

In this study, the effect of habitat improvement was analyzed by calculating the flow range through the analysis of the past hydrological data and calculating the WUA accordingly, unlike the conventional method of discharging a uniform flow throughout the year. The results of this study are expected to be used as effective basic data for effective dam operation in consideration of the ecosystem in the future.

In this study, the 2D model River2D was used; the 2D numerical model was found to be effective compared with the 1D model, especially for spatially distributed phenomena, such as patterns of fish habitat quality under low discharge [55,56]. The River2D model uses triangular irregular network grids; it was determined that more accurate simulations were possible than estimating the habitat area using a one-dimensional model.

Several studies have compared 1D and 2D models or analyzed the effects of hydraulic structures, but studies on dam discharge scenarios and habitat area changes that have a direct effect on fish communities have not been conducted [57,58]. Furthermore, in this study, the simulation accuracy was improved by utilizing the results of on-site monitoring performed directly. River2D was applied as a mixture of sand and granules only in the case of substrate of the target area, reflecting the monitoring results. However, there is a possibility that the sediment supply from the upstream is blocked due to the discharge of the dam, and the composition of the substrate may change. In future studies, it is necessary to analyze the changes in habitat area due to the changes in substrate [59]. Nevertheless, the effect of habitat improvement suggested in this study has been estimated only considering the physical habitat and does not necessarily guarantee the improvement of the actual habitat. Fish habitats in rivers are formed in association with various conditions, in addition to physical habitats such as spawning grounds, and there is a limit to predicting them based on physical habitat only. In this study, habitat suitability for fish species was analyzed considering only discharge from dams (physical conditions). However, there are numerous factors that can affect the habitat suitability for fish species. Studies must be conducted in the future to show that the area of fish habitats changes according to the influence of various factors such as water chemistry and biotic variables (e.g., temperature, predation, prey, and competition) [60,61,62]. Therefore, it will be possible to find effective ways to improve habitats through future studies to analyze the effect through long-term monitoring in actual rivers and reconstruct related scenarios or through research reflecting various habitat conditions for fish species.

5. Conclusions

In this study, BBM was applied to construct discharge scenarios mimicking natural flow regimes using hydrological data from 2006 to 2020. Since the flow regime reflecting the characteristics of each season acted as an important determinant of the physical habitat conditions of living organisms, the BBM was constructed using the monthly average flow as a representative value by dividing the past hydrological data for the wet year, the dry year, and the normal water year. The HSI for the flow velocity, water depth, and bed materials was constructed for the fish species Zacco platypus and Pseudogobio esocinus, which were selected as the dominant and sub-dominant species, respectively, through direct monitoring of the Gudam Bridge basin by four field surveys. The downstream river hydraulic analysis and habitat simulation using the River2D model, which is a two-dimensional habitat model, showed that the WUA was larger in Scenario 2, mimicking the natural flow regime, than the WUA calculated based on the actual flow at the Gudam Bridge. There were some periods (in August and September of S2-4) in which the WUA of Scenario 1 was larger, but in general, a habitat improvement effect of up to about 80% was achieved in the scenario reflecting the BBM. Changing the dam discharge pattern in consideration of the flow regime seems more effective in improving the habitat of fishes living downstream. Since the WUA was larger in the BBM flow, it is more effective in improving the habitat area when the discharge pattern is changed, and efficient river management is possible even with a lower flow. In this study, it was confirmed that the WUA at BBM flow was greater under various flow regimes. It is possible to improve the fish habitat environment by controlling the amount of discharge from the dam and constructing a control dam downstream of the dam to continuously discharge and control the optimal flow. Furthermore, in order to create a healthy ecosystem around rivers, integrated management by controlling flow according to the flow regime, using hydraulic appendix structures, is essential.