Numerical Simulation of Habitat Restoration for Floating Fish Eggs in the Upper Yangtze River Tributaries

Abstract

The upper reaches of the Yangtze River not only serve as a crucial waterway in China’s southwestern region but also function as a conservation area for rare fish species. Recognizing the tendency of current navigation channel improvement projects to overlook the impact on aquatic habitats and the limitations of existing habitat assessment systems, this study specifically focuses on the bifurcation section of the upper Yangtze River. To address these issues, a two-dimensional mathematical model is established to simulate and validate various improvement schemes. An analysis of the flow conditions necessary for fish migration and spawning in this section is conducted, determining optimal flow rates and water levels during spawning and migration periods. A suitability assessment system for spawning and migration is then established, utilizing the four major Chinese carps as representative fish species for evaluation. Through a comprehensive analysis of the results, rational improvement schemes are identified. The findings underscore the importance of considering elevation of dam crest during the construction of sub-dams to regulate the navigation channel, particularly regarding its impact on fish habitats during the dry season. In this plan, the weighted available area of the spawning ground increased by 189,251 m², 165,860 m², 89,527 m², 66,542 m², and 47,182 m² under five conditions. Additionally, the evaluation indexes PPA, HFI, and CQI showed increases of 5.011%, 4.176%, and 2.901%, respectively. Moreover, this study refines fish habitats based on their reproductive and living habits, thereby enhancing existing habitat assessment models.

1. Introduction

The Minjiang River and the Jinsha River, as the main tributaries of the Yangtze River, converge in Yibin City, Sichuan Province, forming the Yangtze River. The upper reaches of the Yangtze River refer to the main stream of the Yangtze River, spanning 384 km from the confluence of the Minjiang River and the Jinsha River to the Three Gorges Dam. This segment of the Yangtze River serves not only as a major maritime transportation route in southwestern China [1] but also as a national-level ecological conservation area for fish [2].

From the perspective of navigation and socio-economic factors, the navigation capacity of the upper Yangtze River section from Yibin to Chongqing is 220 million tons, with a navigation development rate of 41.04%, and according to the development threshold of 80%, there is tremendous development potential [3]. The quantified socio-economic benefits of the navigation channel amount to CNY 33.724 billion, which is relatively small compared to provinces along the Yangtze River in a similar development stage [4]. From an ecological conservation and sustainable development perspective, the continuous construction and operation of cascade hydropower projects in the upper reaches of this river segment have fragmented habitats, leading to a decline in early fish resources [5,6,7,8]. Although various ecological compensation measures have been taken, such as research and construction of ecological fish lanes and laying up ecological reefs, fish protection still faces many problems [9]. Yang Zhi’s research [10] shows that the cascade reservoir has the most obvious influence on the endemic fish of the upper Yangtze River, benthic fish, algal fish, and drifting fish.

Xiao [11] utilized a physical habitat suitability model in conjunction with a two-dimensional hydrodynamic numerical model to investigate the impact of the Three Gorges Dam on the spawning habitat suitability and potential spawning locations of the four major Chinese carps (FMCCs) in the fluctuating backwater area (FBA) of the upper reaches of the Yangtze River. The research findings indicate that the reservoir impoundment operations of the Three Gorges Dam have resulted in a decrease in the suitability of hydrodynamic factors. The migration pattern of the fish has been blocked by the construction of Gezhouba Dam, reducing the natural spawning site length to less than 7 km along the Yangtze River [12]. The findings provide further basis for the protection and restoration of Chinese sturgeon spawning habitats, especially in the lower reach of the Yangtze River.

Therefore, researching how to protect and construct an environment that is more conducive to the survival of fish while ensuring navigation conditions is of great value.

Fish habitat refers to the sum of environmental factors in the aquatic environment that fish rely on for survival. Functionally, it can be divided into the “three fields and one channel”, namely spawning grounds, foraging grounds, overwintering grounds, and migration channels. From a biological point of view, reproduction is a key process to ensure the continued existence and evolution of populations [13]. For fish with sexual reproduction and spawning, the process of spawning and spawning migration is an important link to ensure the continuation of the population. Therefore, in the “three fields and one channel”, spawning grounds and migration channels are particularly important. Current research on fish habitat mainly focuses on the hydrodynamic indicators of fish habitats, including water depth, flow velocity, flow velocity gradient, Froude number, etc. [14,15,16,17]. Studies by researchers such as Wang Shipeng [18], Lou Qun [19], Liu Ya [20], Shih [21], Im [22], and Deng [23] have shown that the improvement of structures is conducive to enriching fish habitats and promoting the recovery and reconstruction of aquatic ecosystems. Yi et al. [24] studied the impact of the Gezhouba and Three Gorges Dams on habitat suitability for carps in the Yangtze River. By coupling the habitat suitability curves and the 1-D mathematical model, an HSI model for the YFMCS was established.

Noor Ben Jebria et al. [25] explored the relationship between juvenile fish behavior and hydraulic cues by integrating three-dimensional computational fluid dynamics (CFD) simulations with two-dimensional telemetry tracking of fish positions. Their findings provide new insights that contribute to the design of downstream migration passage facilities.

However, most existing studies utilize a single suitability curve for flow velocity and water depth to calculate the weighted usable area and evaluate the habitat suitability [26]. Since fish engage in different activities in different seasons and in habitats serving different ecological functions, the criteria for evaluating the water flow conditions of habitats supporting various ecological functions also differ [27]. For example, when analyzing spawning grounds, the focus is on whether the water flow conditions are conducive to the fertilization of buoyant fish eggs and stimulate fish spawning. During fish migration upstream, considering the limited swimming abilities of fish, it is crucial to assess whether the flow velocity exceeds the maximum swimming capacity of fish, thus evaluating whether the river section is suitable for fish migration. The most ideal habitat restoration project should aim to meet the hydraulic needs of fish at different stages of their life cycle, making the river section a multi-functional habitat.

In the main stream of the Yangtze River, the fish producing floating eggs are less threatened in the development and maturation process and belong to the dominant species in the main stream of the Yangtze River [28]. Among them, the fish spawning field represented by the four major fish has a wide range, and the spawning process is closely related to the water flow conditions in flood season. In addition, the four kinds of fish are China’s unique economic fish, and in history, the Yangtze River is the main supply base and natural seedling production accounted for about 70% of the total output. The Yangtze river drainage fish genetic characteristics are significantly better than for other drainages, and they constitute unique genetic resources, which cannot be replaced from other drainages or by artificial means [29,30,31].

Therefore, this study intends to select the four major Chinese carps as research subjects. First, the study selects the river sections and proposes habitat restoration measures based on the characteristics of the river sections and determines the conditions that need to be calculated. Second, a mathematical model is constructed and validated. Third, this study analyzes the habitat restoration effect based on the suitable flow conditions for spawning and migration, which are critical during the life cycles of these fish. Finally, we analyze and discuss the computational results.

2. Habitat Restoration Plan

2.1. River Section Overview

This river section is located approximately 16 km downstream from the confluence of the Yangtze River and Tuo River, within the jurisdiction of Luzhou, along the upstream navigable stretch of the Yangtze River at 892.7–897.0 km. In this section, the river is divided into left and right channels by the Yijiang Heart Island (Furong Dam), creating a straight to slightly curved bifurcated segment. The left channel has shallow water and is not suitable for navigation during low water levels, while the right channel serves as a navigable passage during the dry season. The satellite real scene map is shown as Figure 1.

2.2. Design of the Habitat Restoration Plan

(1)

Improvement Plan Design Concept

Due to the non-navigable nature of the left channel and its shallow depth during the dry season, a strategy is proposed to enhance navigational conditions and prevent the left channel from evolving into the main channel. This involves constructing a submerged lock dam at the entrance of the left channel. By doing so, the flow velocity in the left channel is reduced, simultaneously raising the water level in the right channel. This improvement aims to enhance navigational conditions and, at the same time, prevents the left channel from becoming the primary navigational route.

Additionally, the selection of a submerged lock dam can contribute to the sedimentation of the tributary, leading to the refinement of bed sand particle sizes. To avoid disruptions caused by waterfalls, cross currents, and eddies, the dam axis of this model is strategically placed at the upstream point of the left channel at the 895 km mark along the navigational stretch. The specific position of the dam axis in this river section is shown in Figure 2.

2.3. Arrangement of Habitat Restoration Plan

(1)

Submerged Lock Dam Size Design

Referencing the relevant design specifications for lock dams in the upper reaches of the Yangtze River, the dam body design parameters are as shown in

Table 1. Design of dimensions for submerged dams.

Plan	Elevation of Dam Crest (m)	Top Width (m)	Upstream Slope Ratio	Backwater Slope Ratio
Plan A	219	3	1:2	1:3
Plan B	220	4	1:2	1:3
Plan C	221	5	1:2	1:3
Plan D	222	6	1:2	1:3

.

Using Plan B as an example, the specific parameters of the dam body and the model’s performance are illustrated in Figure 3.

(2)

Calculation of Operating Conditions Design

2.3.1. Selection of Representative Conditions for Reproductive Migration Period

The entire reproductive season for the four major fish species spans from late April to late July [32]. Prior to this period, fish are stimulated by the rising water temperatures and initiate reproductive migration, typically occurring from January to April. Through statistical analysis, it has been determined that the monthly average flow in this river segment remains consistently at 5000 m³/s from January to May over multiple years. Therefore, 5000 m³/s is chosen as the representative flow rate for the reproductive migration period of the four major fish species.

2.3.2. Selection of Representative Conditions for Spawning Period

The reproductive spawning of the four major Chinese carps not only requires suitable temperature conditions but also necessitates conditions conducive to rising water levels. Based on the theory of rising water conditions proposed by Shen Chen [33] as an identification criterion for the peak reproductive period, a statistical analysis of effective rising water processes from April to July for the years 2013 to 2020 at the Zhu River hydrological station is presented in Figure 4. Figure 4 shows water rising processes that meets discriminant criteria and may stimulate fish reproduction during the breeding season from 2013 to 2020.

The flow on the last day of each effective rising water process in the graph above represents the flow during the peak spawning period. After statistical analysis of the peak spawning period flows, it was observed that the peak reproductive flows are distributed across six flow intervals. After calculating the averages and rounding to the nearest integer, six representative flow rates for the peak spawning period were obtained, which will be used as simulated operating conditions. According to the above analysis, the simulated operating conditions are shown in

Table 2. Design of simulation conditions.

	Conditions	Flow (m³/s)	Water Level (m)
Migration Period Conditions	M0	5000	220.43
Spawning Period Conditions	M1	5300	220.58
	M2	6500	221.07
	M3	7500	221.55
	M4	8500	222.04
	M5	9400	222.42
	M6	11,000	223.67

.

3. Mathematical Model

3.1. Model Principles and Parameters

When conducting three-dimensional numerical simulations of water and sediment, the vertical direction is considered along the water depth. By integrating the fundamental equations along the vertical direction and then taking the average, the three-dimensional problem is transformed into a two-dimensional problem. The continuity equation and the motion equation for water flow are as follows:

Continuity equation:

$$\frac{\partial \mathrm{h}}{\partial \mathrm{t}}+\frac{\partial \mathrm{h}\overline{\mathrm{u}}}{\partial \mathrm{x}}+\frac{\partial \mathrm{h}\overline{\mathrm{v}}}{\partial \mathrm{y}}=\mathrm{h}\mathrm{S}$$

(1)

X and Y direction motion equations:

$$\frac{\partial \mathrm{h}\overline{\mathrm{u}}}{\partial \mathrm{t}}+\frac{\partial {\overline{\mathrm{u}}}^2}{\partial \mathrm{x}}+\frac{\partial \mathrm{h}\overline{\mathrm{u}\mathrm{v}}}{\partial \mathrm{y}}=\mathrm{f}\overline{\mathrm{v}}\mathrm{h}-\mathrm{g}\mathrm{h}\frac{\partial \mathrm{\eta }}{\partial \mathrm{x}}-\frac{\mathrm{h}}{{\mathsf{\rho }}_0}\frac{\partial {\mathrm{P}}_{\mathrm{\alpha }}}{\partial \mathrm{x}}-\frac{\mathrm{g}{\mathrm{h}}^2}{2{\mathsf{\rho }}_0}\frac{\partial \mathsf{\rho }}{\partial \mathrm{x}}+\frac{{\mathrm{\tau }}_{\mathrm{s}\mathrm{x}}}{{\mathsf{\rho }}_0}-\frac{{\mathrm{\tau }}_{\mathrm{b}\mathrm{x}}}{{\mathsf{\rho }}_0}-\frac{1}{{\mathsf{\rho }}_0}(\frac{\partial {\mathrm{s}}_{\mathrm{x}\mathrm{x}}}{\partial \mathrm{x}}+\frac{\partial {\mathrm{s}}_{\mathrm{y}\mathrm{y}}}{\partial \mathrm{y}})+\frac{\partial }{\partial \mathrm{x}}(\mathrm{h}{\mathrm{T}}_{\mathrm{x}\mathrm{x}})+\frac{\partial }{\partial \mathrm{y}}(\mathrm{h}{\mathrm{T}}_{\mathrm{y}\mathrm{y}})+\mathrm{h}{\mathrm{u}}_{\mathrm{s}}\mathrm{S}$$

(2)

$$\frac{\partial \mathrm{h}\overline{\mathrm{v}}}{\partial \mathrm{t}}+\frac{\partial {\overline{\mathrm{v}}}^2}{\partial \mathrm{y}}+\frac{\partial \mathrm{h}\overline{\mathrm{v}\mathrm{u}}}{\partial \mathrm{x}}=-\mathrm{f}\overline{\mathrm{v}}\mathrm{h}-\mathrm{g}\mathrm{h}\frac{\partial \mathrm{\eta }}{\partial \mathrm{y}}-\frac{\mathrm{h}}{{\mathsf{\rho }}_0}\frac{\partial {\mathrm{P}}_{\mathrm{\alpha }}}{\partial \mathrm{y}}-\frac{\mathrm{g}{\mathrm{h}}^2}{2{\mathsf{\rho }}_0}\frac{\partial \mathsf{\rho }}{\partial \mathrm{y}}+\frac{{\mathrm{\tau }}_{\mathrm{s}\mathrm{y}}}{{\mathsf{\rho }}_0}-\frac{{\mathrm{\tau }}_{\mathrm{b}\mathrm{y}}}{{\mathsf{\rho }}_0}-\frac{1}{{\mathsf{\rho }}_0}(\frac{\partial {\mathrm{s}}_{\mathrm{y}\mathrm{x}}}{\partial \mathrm{x}}+\frac{\partial {\mathrm{s}}_{\mathrm{y}\mathrm{y}}}{\partial \mathrm{y}})+\frac{\partial }{\partial \mathrm{x}}(\mathrm{h}{\mathrm{T}}_{\mathrm{x}\mathrm{y}})+\frac{\partial }{\partial \mathrm{y}}(\mathrm{h}{\mathrm{T}}_{\mathrm{y}\mathrm{y}})+\mathrm{h}{\mathrm{v}}_{\mathrm{s}}\mathrm{S}$$

(3)

In the formulas, t represents time; $\mathrm{\eta }$ is the water level; h is the total water depth; u, v is velocity component; $\mathrm{f}$ is the Coriolis force coefficient, $\mathrm{f}=2\mathsf{\omega }\sin \mathsf{\phi }$; $\mathsf{\omega }$ is the angular velocity of the Earth’s rotation; $\mathsf{\phi }$ is the local latitude; g is the local latitude; $\mathsf{\rho }$ is the density of water; S is the source term; ${(\mathrm{u}}_{\mathrm{s}},{\mathrm{v}}_{\mathrm{s}})$ is source flow velocity.

Based on the aforementioned equations, for the convenience of numerical solution, we discretized the entire river section using the finite volume method for spatial discretization and solved the time integration using the semi-implicit Euler method. We employed the same approach to construct different grids for different engineering schemes to make the model as realistic as possible and reduce errors. Taking Plan A as an example, the total area of the model is 7,644,061 m², divided into 67,393 cells, with grid node spacings ranging from 5 m to 35 m. There may be slight differences in various parameters between different grids. We set the time step to 30 s, the number of calculation steps to 360, and the total calculation duration to 10,800 s. After ensuring the stability of the model’s water flow, we extracted the model calculation data and performed post-processing analysis.

3.2. Model Validation

The measured water levels at the entrance section and exit section of the model, as well as the calculated water levels for each element along the cross-section, are illustrated in Figure 5.

Table 3. Statistical Analysis of Model Deviations.

Cross-Section	Measured Water Level	Average Calculated Water Level	Deviation (m)
Inlet Section	224.27	224.2886032	0.0186023
Outlet Section	222.85	222.8231845	0.0268155

presents the average water levels of the elements along the section, the measured water levels, and their respective deviations.

The figure and table provide the calculated values of the model water surface profile under measured flow rates, along with the measured water levels. A comparison reveals a close agreement between the two, with differences less than 0.03 m, meeting the required accuracy standards.

4. Habitat Evaluation Indicators and Calculation Methods

As mentioned above, four species will be used to study the habitat restoration of their respective habitats. Beforehand, we need to establish a suitability assessment system to evaluate the effectiveness of the restoration efforts. The focus of the habitat improvement plan is to enhance the flow conditions during fish spawning and migration periods. Therefore, it is necessary to calculate the flow and water levels representing fish spawning migration and spawning periods separately. We referred to the hydraulic conditions suitable for spawning of these four fish species and their swimming abilities [34,35,36,37] and constructed different assessment models for spawning and migration behaviors.

4.1. Spawning Suitability Assessment

Apart from water temperature conditions, the suitability of a river section for fish spawning is mainly related to flow velocity and water depth. However, the suitability of a natural river section for fish habitation and reproduction cannot be determined by a single factor; it depends on the combined suitability of multiple factors for fish. This paper establishes an assessment model for the suitability of spawning grounds in the river section based on the weighted usable area (WUA) method [38]. The specific calculation formula is as follows:

$$\mathrm{W}\mathrm{U}\mathrm{A}={\sum }_{\mathrm{i}=1}^{\mathrm{n}}\mathrm{C}\mathrm{S}\mathrm{F}\left({\mathrm{V}}_{1\mathrm{i}},{\mathrm{V}}_{2\mathrm{i}},{\mathrm{V}}_{3\mathrm{i}}\cdots {\mathrm{V}}_{\mathrm{m}\mathrm{i}}\right)\times {\mathrm{A}}_{\mathrm{i}}$$

(4)

In the formula: $\mathrm{W}\mathrm{U}\mathrm{A}$ represents the weighted usable area; ${\mathrm{A}}_{\mathrm{i}}$ is the projected area of the $\mathrm{i}$-th element; ${\mathrm{C}\mathrm{F}\mathrm{S}}_{\mathrm{i}}\left({\mathrm{V}}_{1\mathrm{i}},{\mathrm{V}}_{2\mathrm{i}},{\mathrm{V}}_{3\mathrm{i}}\cdots {\mathrm{V}}_{\mathrm{m}\mathrm{i}}\right)$ is the value of the comprehensive suitability of the $\mathrm{i}$-th cell in this river section; $\left({\mathrm{V}}_{1\mathrm{i}}\cdots {\mathrm{V}}_{\mathrm{m}\mathrm{i}}\right)$ represents $\mathrm{m}$ different evaluation indicators. This paper only considers the comprehensive suitability based on flow velocity and water depth. The calculation formula is as follows:

$${CSF}_i={(u_i\times h_i)}^{1/2}$$

(5)

In which $u_i$ and $h_i$ are the assigned values for flow velocity and water depth suitability of the $\mathcal{i}$-th cell, with reference to the literature [36]. The specific suitability curves are illustrated in Figure 6.

4.2. Migration Suitability Assessment

In this model, the river section is a natural one, and the successful migration of fish depends mainly on whether the flow velocity in this section is within the swimming capabilities of the fish [39,40]. The swimming ability of adult fish may vary slightly due to species, body length, age, and water temperature, but it generally falls within a certain range [41]. Referencing fishway design standards and relevant literature [42], this paper adopts 0.2 m/s as the flow velocity sensed by fish.

The continuous swimming speed and maximum swimming speed of mature fish are calculated using empirical formulas.

$${\mathrm{V}}_{\mathrm{n}}=(2~3)\mathrm{L}$$

(6)

In which ${\mathrm{V}}_{\mathrm{n}}$ represent the sustained swimming speed of fish, cm/s; $\mathrm{L}$ is the length of fish, cm.

$${\mathrm{V}}_{\mathrm{m}\mathrm{a}\mathrm{x}}=\mathrm{n}\mathrm{L}+\mathrm{m}$$

(7)

${\mathrm{V}}_{\mathrm{m}\mathrm{a}\mathrm{x}}$ represent maximum swimming speed, cm/s; $\mathrm{n}$ and $\mathrm{m}$ are swimming speed coefficient of the four major Chinese carp species, $\mathrm{n}=3$, $\mathrm{m}=65$.

Fish reproductive migration occurs after the sexual maturity of fish. Different fish species take varying amounts of time to reach sexual maturity, and their lengths upon reaching sexual maturity also differ. In this paper, the median length of mature fish at the age of two and three is taken as 50 cm, and calculations for the sustained swimming speed are performed based on this length. That is, ${\mathrm{V}}_{\mathrm{n}}$ = 100 cm/s~150 cm/s, ${\mathrm{V}}_{\mathrm{m}\mathrm{a}\mathrm{x}}=215 \mathrm{c}\mathrm{m}/\mathrm{s}$. The spawning migration suitability curve is shown in Figure 7.

5. Spawning Ground Suitability Evaluation

5.1. Analysis of Natural River Spawning Ground Suitability

From Figure 8a,c,e,g,h,k, it can be observed that both in the natural river channel and the channel after habitat improvement, the suitable spawning area tends to increase with the increase in flow rate. Taking the natural river channel before the project as an example, an analysis is conducted on the relationship between flow rate and the suitable spawning area.

One noticeable area with significant changes is the left bank of the outlet section of the left channel. Due to the presence of the bend-sealing dam, the water depth and flow velocity in this bend-sealing area are not suitable for fish spawning during low water levels. As the water level increases, the water level and flow velocity in this area gradually increase. When the flow rate increases to 6500 m³/s, the spawning suitability in this area improves significantly. It continues to increase, reaching a suitability value above 0.8 when the flow rate reaches 9400 m³/s.

Another area with noticeable changes is the main channel area of the right channel. The overall trend of spawning suitability in this area decreases as the flow rate increases. The main factor contributing to the decrease in suitability is the flow velocity. As the flow rate increases, the suitable water depth for spawning in this area remains relatively high, but the suitability of flow velocity decreases significantly. This results in an overall decrease in spawning suitability when the flow rate increases from the M3 condition to the M4 condition.

5.2. Evaluation of the Effectiveness of Plan A

The distribution of spawning suitability before and after Plan A for various conditions is shown in Figure 8, and the values corresponding to different colors in the subsequent cloud map are as shown in Figure 8m.

Due to the lower elevation of the dam crest in Plan A and limited interception capacity, the spawning suitability under various conditions after the project is not significantly different from that of the original river channel. Moreover, under the M3 and M4 conditions, there is a noticeable reduction in the weighted usable spawning area. The change in spawning suitability before and after the project is more pronounced in the M4 condition, with a significant decrease in spawning suitability to below 0.4 in most areas of both left and right channels.

Under the M5 and M6 conditions, the overall spawning suitability of the river section remains basically unchanged before and after the project, and the weighted usable spawning area is comparable to that of the original river channel. The change values before and after Plan A are shown in

Table 4. Summary of the Before and After Effects of PLAN A.

Conditions	WUA ($\mathbf{C}\mathbf{F}\mathbf{S} > 0$)			WUA ($\mathbf{C}\mathbf{F}\mathbf{S} > 0.6$)
	Before (m2)	After (m2)	Growth Rate (%)	Before (m2)	After (m2)	Growth Rate (%)
M1	2,412,004	2,406,522	−0.22	2,079,686	2,067,085	−0.60
M2	2,719,174	2,719,371	0.00	2,429,256	2,428,986	−0.01
M3	3,168,574	2,937,111	−7.30	2,886,769	2,632,777	−8.79
M4	3,113,533	2,727,965	−12.38	2,814,238	2,434,727	−13.48
M5	3,208,826	3,208,397	−0.01	2,924,498	2,924,493	0.00
M6	3,460,038	3,455,152	−0.14	3,199,333	3,193,875	0.17

.

As shown in the above table, the habitat improvement plan has no significant impact on the spawning grounds for eggs.

5.3. Evaluation of the Effectiveness of Plan B

The distribution of spawning suitability before and after Plan B for various conditions is shown in Figure 9.

By comparing the parts circled in the box in Figure 9a,b, it can be found that even when the discharge is low, after the implementation of Plan B, the suitability of the right branch channel is obviously improved.

In the M1 condition, compared to the original river channel, there is a significant improvement in the spawning suitability of the right channel after the project, with an increase of 189,251 square meters in the weighted usable spawning area for the entire river section. Under the M2 condition, some areas in the mid-lower reaches of the right channel show an increase in spawning suitability to above 0.6. In the M3 condition, there is a decrease in the spawning suitability of the right channel after the project, attributed to increased flow velocity resulting from water diversion to the left channel, leading to a decrease in spawning suitability. Additionally, there are small areas both upstream and downstream of the submersible dam where the spawning suitability falls below 0.8. This results in a decrease in spawning suitability after the project under the M3 condition. Under the M4 condition, some areas at the location of the left channel dam exhibit an increase in spawning suitability to above 0.8, and there is an improvement in spawning suitability in the entrance section of the right channel and the mid-upper reaches of the right channel. Although the improvement in spawning suitability is not visually evident in the cloud map under the M4 and M5 conditions, the numerical values show an increase in the weighted usable spawning area by 66,542 square meters and 47,182 square meters, respectively. This indicates that, while there may not be significant improvements in specific regions, the overall spawning suitability of the entire river section has increased. Overall, although there is a decrease in spawning suitability under the M3 condition, there is an improvement in spawning suitability under the remaining five conditions. The change values before and after Plan B are shown in

Table 5. Summary of the Before and After Effects of PLAN B.

Conditions	WUA ($\mathbf{C}\mathbf{F}\mathbf{S} > 0$)			WUA ($\mathbf{C}\mathbf{F}\mathbf{S} > 0.6$)
	Before (m2)	After (m2)	Growth Rate (%)	Before (m2)	After (m2)	Growth Rate (%)
M1	2,412,004	2,601,255	7.84	2,079,686	2,312,242	11.18
M2	2,719,174	2,885,034	6.09	2,429,256	2,614,609	7.63
M3	3,168,574	3,072,401	−3.03	2,886,769	2,784,441	−3.54
M4	3,113,533	3,203,060	2.87	2,814,238	2,906,073	3.26
M5	3,208,826	3,275,368	2.07	2,924,498	2,959,421	1.19
M6	3,460,038	3,507,220	1.36	3,199,333	3,203,596	0.13

.

5.4. Evaluation of the Effectiveness of Plan C

With the increase in the height of the left channel dam, the flow interception capacity of the dam has improved. At low flow rates, changes in the river channel flow field are more significant. Under the M1 condition, there is a significant decrease in the spawning suitability in the entrance sections of both the left and right channels after the project. This is because, at low flow rates, the presence of the left channel dam leads to a reduction in the diversion flow of the left channel and an increase in the diversion flow of the right channel, resulting in a decrease in water depth in the left channel and an increase in flow velocity in the entrance section of the right channel, thereby affecting spawning suitability. However, from a numerical calculation perspective, the post-project weighted usable spawning area increased by 77,252 square meters compared to before the project. In the M2 condition, with the increase in flow rate and rise in water level relative to M1, the diversion ratio of the left channel has increased, leading to a significant improvement in the spawning suitability of the left channel under the post-project M2 condition compared to the M1 condition. Moreover, the post-project weighted usable spawning area increased by 170,710 square meters compared to the original river channel under the same condition. Under the M3 condition, there is a slight decrease in the weighted usable spawning area, but it is the smallest decrease compared to other scenarios. Under the M4, M5, and M6 conditions, the downstream area of the dam shows significant improvement, with the weighted usable spawning area increasing by 187,543 square meters, 171,053 square meters, and 144,530 square meters, respectively. The changes before and after the implementation of Plan C are shown in Figure 10.

The change values before and after Plan C are shown in

Table 6. Summary of the Before and After Effects of PLAN C.

Conditions	WUA ($\mathbf{C}\mathbf{F}\mathbf{S} > 0$)			WUA ($\mathbf{C}\mathbf{F}\mathbf{S} > 0.6$)
	Before (m2)	After (m2)	Growth Rate (%)	Before (m2)	After (m2)	Growth Rate (%)
M1	2,412,004	2,489,256	3.20	2,079,686	2,085,593	0.28
M2	2,719,174	2,889,884	6.27	2,429,256	2,573,547	5.93
M3	3,168,574	3,137,432	−0.98	2,886,769	2,820,210	−2.30
M4	3,113,533	3,301,076	6.02	2,814,238	2,995,284	6.43
M5	3,208,826	3,379,879	5.33	2,924,498	3,073,458	5.09
M6	3,460,038	3,604,568	4.17	3,199,333	3,309,706	3.44

.

5.5. Evaluation of the Effectiveness of Plan D

Plan D has a higher dam elevation, so under the M1 condition, the left channel is intercepted by the dam, and no water flows through the left channel, resulting in the loss of a large area of spawning grounds. Moreover, with a significant increase in the diversion of the right channel, both flow velocity and water depth have significantly increased, leading to a decrease in spawning suitability in most areas of the right channel. As the flow rate increases, under the M2 and M3 conditions, although there is water flow through the left channel and the usable spawning area has increased, the spawning suitability of the right channel has improved, but compared to the original river channel, the weighted usable spawning area has significantly decreased. With further increases in flow rate and water level, under the M4, M5, and M6 conditions, there is some improvement in spawning suitability in the downstream area of the left channel dam, with suitability increasing to 0.8 or above. However, the suitability of the large area in the right channel has decreased compared to the original river channel. Nevertheless, under these three conditions, the post-project weighted usable spawning area has increased, with increases of 4997 square meters, 56,213 square meters, and 128,387 square meters, respectively. The changes before and after the implementation of Plan D are shown in Figure 11.

Overall, although this plan improves the spawning grounds under high flow conditions, it results in a significant reduction in the area of fish spawning grounds under low flow conditions. This indicates that in future renovations of the right channel, improving low-flow navigation conditions by intercepting flow from the left channel and raising the water level in the right channel may not be a viable approach.The change values before and after Plan D are shown in

Table 7. Summary of the Before and After Effects of PLAN D.

Conditions	WUA ($\mathbf{C}\mathbf{F}\mathbf{S} > 0$)			WUA ($\mathbf{C}\mathbf{F}\mathbf{S} > 0.6$)
	Before (m2)	After (m2)	Growth Rate (%)	Before (m2)	After (m2)	Growth Rate (%)
M1	2,412,004	1,995,148	−17.28	2,079,686	1,707,235	−17.91
M2	2,719,174	2,455,119	−9.71	2,429,256	1,941,709	−20.07
M3	3,168,574	2,834,993	−10.53	2,886,769	2,425,206	−15.99
M4	3,113,533	3,118,530	0.16	2,814,238	2,754,160	−2.13
M5	3,208,826	3,265,039	1.75	2,924,498	2,941,213	0.57
M6	3,460,038	3,588,425	3.71	3,199,333	3,288,635	2.79

.

6. Fish Migration Improvement Analysis

6.1. Evaluation Indicators for Fish Migration

In contrast to the assessment of spawning grounds, where fish autonomously seek suitable breeding locations in a large area, the reproductive migration of fish involves swimming from the downstream to the upstream of a river. During this process, fish cannot instantaneously traverse essential river sections and must navigate through them. If the flow velocity is too low, fish may fail to sense the water flow stimulus, leading to a loss of migration direction. Conversely, if the flow velocity is too high or if there is a significant gradient, fish may surpass their swimming abilities, resulting in migration obstacles or failure. Therefore, when evaluating the suitability of fish migration channels, it is crucial to consider the connectivity of these channels. Drawing on research methodologies by Zhang Xinhua [9] and others, we introduce metrics such as patch area ratio, migration channel fragmentation index, migration channel connectivity index, and migration channel comprehensive quality index to assess the improvement effects on fish migration channels following different restoration schemes.

(1) Percentage of Patch Area

$$\mathrm{P}\mathrm{P}\mathrm{A}=\sum _{\mathrm{m}=1}^{\mathrm{k}}\frac{{\mathrm{P}}_{\mathrm{m}}}{\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{i}}}$$

Here, ${\mathrm{P}}_{\mathrm{m}}$ represents the area of the m-th migratory micro-patch, and q denotes the number of large patches formed by the connectivity of micro-patches.

(2) Habitat Fragmentation Index

The habitat fragmentation index is defined as the ratio of the average area of each patch to the total area of the entire river section. Its value ranges from 0 to 1, with smaller values indicating more severe fragmentation of the migratory channel, which is less favorable for fish migration.

$$\mathrm{H}\mathrm{F}\mathrm{I}=\frac{\sum _{\mathrm{m}=1}^{\mathrm{k}}{\mathrm{P}}_{\mathrm{m}}/\mathrm{q}}{\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{i}}}$$

The parameters’ interpretations and values are the same as mentioned above in the equation.

(3) Habitat Connectivity Index

Habitat connectivity index (HCI) is an important parameter that characterizes the connectivity of habitats. Only when the connectivity is good can fish have a suitable living space. According to the needs of this study, the calculation formula for HCI given in reference [9] is modified as follows:

$$\mathrm{H}\mathrm{C}\mathrm{I}=\frac{\sum _{\mathrm{m}=1}^{\mathrm{k}}{\mathrm{P}}_{\mathrm{m}}}{\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{i}}}\left(\sum _{\mathrm{m}=1}^{\mathrm{k}}\frac{{({\mathrm{P}}_{\mathrm{m}}/{\mathrm{D}}_{\mathrm{m}})}^2}{\sum _{\mathrm{m}=1}^{\mathrm{k}}{\mathrm{P}}_{\mathrm{m}}/{\mathrm{D}}_{\mathrm{m}}}\right)=\frac{\sum _{\mathrm{m}=1}^{\mathrm{k}}{{(\mathrm{P}}_{\mathrm{m}})}^2/{\mathrm{D}}_{\mathrm{m}}}{\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{i}}}$$

In which ${\mathrm{D}}_{\mathrm{m}}$ is the shortest distance between the m-th micro-patch in the migratory channel and other micro-patches.

(4) Comprehensive Quality Index

When evaluating the overall quality of the migratory channel in this study, a weighted analysis is performed on the three evaluation indicators, namely PPA, HFI, and HCI. The specific calculation formula is as follows:

$$\mathrm{C}\mathrm{Q}\mathrm{I}=\sum _{\mathrm{\lambda }=1}^3{\mathsf{\beta }}_{\mathrm{\lambda }}{\mathrm{q}}_{\mathrm{\lambda }}$$

In which ${\mathsf{\beta }}_{\mathrm{\lambda }}$ is the weight of the evaluation index; ${\mathrm{q}}_{\mathrm{\lambda }}$ is the calculated value of the evaluation index; CQI is the comprehensive quality index of the migratory channel. A higher calculated value represents higher quality. When evaluating the suitability of the migratory channel, the fragmentation index and connectivity index should be given more consideration. Therefore, the weights for PPA, HFI, and HCI are taken as 0.3, 0.35, and 0.35, respectively.

6.2. Combined Analysis of Indicators with Cloud Maps

The calculation results for different parameters under each scheme are shown in the

Table 8. Evaluation of Migratory Channel Suitability.

	PPA %	HFI %	HCI %	CQI %
No Plan	22.067	18.389	5.344	14.927
Plan A	21.947	18.290	5.353	14.859
Plan B	27.078	22.565	4.969	17.760
Plan C	21.809	18.174	5.347	14.775
Plan D	19.423	16.186	5.644	13.467

.

The cloud maps of the original river channel and the post-construction fish migration suitability are shown above. In terms of the PPA index, those of Plan A, Plan C, and Plan D are lower than in the original river channel, indicating a reduction in the area suitable for upstream fish migration under these three schemes. Among them, Plan A and Three show a slight decrease, while Plan D exhibits a significant decrease, as seen in Figure 12a,b,d,e. The notable reduction is mainly attributed to two factors: firstly, during the spawning migration season of fish, the river segment experiences low water levels in the dry season, preventing water overflow over the left branch’s dam, resulting in significant shrinkage of areas available for fish migration, as shown in Figure 12e. Secondly, the diversion of water by the left branch’s dam causes an elevation in water levels and an acceleration of flow in the right branch, thereby reducing the suitable upstream migration area.

In the HFI, the original river channel is slightly higher than in Plan A and Three but generally maintained at the same level, indicating a similar degree of fragmentation in fish migration channels. Plan D’s HFI is significantly higher than that of the original river channel, suggesting that the fish migration channel is more fragmented and dispersed compared to the original river channel. This is due to the increased flow velocity in the right branch after the dam, causing the originally continuous fish migration channel to become scattered and reducing the suitability for fish migration, as shown in Figure 12e. In contrast to other schemes and the original river channel, Plan B exhibits a noticeable increase in the HFI. This is related to the elevated PPA index, indicating an increase in the suitable area for fish migration. The areas suitable for upstream migration created by Plan B result in an enlargement of the original fish migration channel, making some scattered micro-patches more continuous.

The relationship between the HCI values of each plan and the original river channel differs slightly from the other three indices. The sizes of the original river channel, Plan A, and Plan C are generally consistent, but Plan D is slightly larger than the original river channel and Plans A and C by around 0.3%, while Plan B is slightly smaller than the original river channel and Plans A and C by around 0.3%. This difference in the HCI values may be attributed to the nature of the fragmentation index. In light of this situation, a simplified error comparison analysis was conducted for Plan B and Plan D, considering the numerical calculation process and in conjunction with the cloud map. Since the same grid boundaries were used and the calculation area, i.e., the total area of the grid cells, is the same, the denominator $\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{i}}$ in the formula can be disregarded for its impact on the results. The numerator $\sum _{\mathrm{m}=1}^{\mathrm{k}}{{(\mathrm{P}}_{\mathrm{m}})}^2/{\mathrm{D}}_{\mathrm{m}}$ is divided into two parameters, ${\mathrm{P}}_{\mathrm{m}}$ and ${\mathrm{D}}_{\mathrm{m}}$, for consideration. As the suitable habitat area in Plan B increases, the number of participating grid cells also increases, i.e., the value of parameter k in the numerator becomes larger. Since the calculated value utilizes the average distance between grid centers, this would cause the value of $\sum _{\mathrm{m}=1}^{\mathrm{k}}1/{\mathrm{D}}_{\mathrm{m}}$ in Plan B to be larger than the actual value, resulting in a slightly underestimated overall result. Similarly, as the suitable habitat area in Plan D decreases, the number of eligible grid cells decreases, leading to a reduction in the value of parameter k, and the calculated value of $\sum _{\mathrm{m}=1}^{\mathrm{k}}1/{\mathrm{D}}_{\mathrm{m}}$ in Plan D is slightly smaller than the actual value, causing the HCI calculation result to be underestimated.

7. Discussion

The trend of spawning suitability in the original river channel and four engineering schemes varies with the flow conditions, as shown in the Figure 13.

In the selection of the optimal solution, the improvement of both spawning grounds and migration channels should be considered simultaneously. Regarding the evaluation of spawning grounds, as shown in Figure 13a, it can be observed that the effectiveness of Plan A is not significant. In all four scenarios, the weighted usable area (WUA) of spawning grounds for Plan A is nearly the same as the original river section, and in two scenarios, it is significantly smaller than the original river section. For Plan D, the WUA of spawning grounds is much smaller than that of the unaltered river section in the M1, M2, and M3 scenarios, approximately equal to that in the M4 scenario, and slightly larger than that in the M5 and M6 scenarios. Plan B exhibits a WUA of spawning grounds slightly smaller than that of the original river section only in the M3 scenario, while in the other five scenarios, it is significantly larger than in the original river section. Plan C’s WUA of spawning grounds is approximately the same as that of the original river section only in the M3 scenario; in the remaining scenarios, its WUA is much larger than in the original river section. In summary, Plan B and Plan C show significant improvements in spawning grounds, while the improvements in Plan C and Plan D are less notable.

Regarding the improvement of migration channels, as shown in Figure 13b, Plan A and Plan C have values of various migration evaluation indices that are nearly identical to those of the original river section, indicating that these two plans have little effect on improving fish migration. Plan D’s three evaluation indices are all significantly smaller than those of the original river section, with only a slight discrepancy in the HCI calculation due to inevitable errors in the process. Plan B exhibits values significantly larger than those of the original river section in three out of the four evaluation indices, with only HCI being slightly smaller (around 0.3%).

Considering both aspects, the optimal solution should be chosen from Plan B and Plan C. In this case, the consideration should also involve the construction quantity. Since the elevation of dam crest and top width of Plan B are smaller than those of Plan C, the construction quantity of Plan B is less than that of Plan C. Additionally, the lower elevation of dam crest in Plan B implies less difficulty for fish to migrate over the dam compared to Plan C. Taking these two factors into account, Plan B emerges as the optimal solution for habitat restoration.

The results of this study differ from those of related research. Through analyzing existing literature [11,17,18,19,20,38], it has been observed that the major difference among scholars in evaluating the effectiveness of habitat restoration lies in the establishment of evaluation models. Specifically, the variation stems from differences in the factors considered and the suitability values assigned when calculating weighted usable area (WUA). However, a commonality exists in the use of a fixed calculation method to compute WUA before and after habitat restoration projects. Subsequently, comparisons are made between the WUA values at different flow rates to assess the effectiveness of the restoration projects. Nonetheless, our analysis suggests that the optimal engineering solution may not necessarily result in the highest increase in WUA, a departure from some existing studies. This paper does not simply evaluate based on changes in WUA, rather, it considers habitat suitability in two aspects: whether it is suitable for spawning during the spawning season and whether it is conducive to migration during the migration season. Based on these two aspects, different influencing factors and suitability criteria are selected to construct two distinct evaluation systems. Finally, by comprehensively considering these two aspects, the effectiveness of habitat restoration is evaluated.

8. Conclusions

When studying habitat restoration in the bifurcation section of the river, the research methodology in this paper differs from existing methods. Firstly, instead of using a single evaluation model for the same river section, this paper establishes two evaluation models based on the different habitat functions that the river section needs to fulfill. This is an innovative approach not previously seen in past studies and represents a significant methodological contribution in this paper. Secondly, different influencing factors are considered and different suitability parameters are used when calculating WUA. This allows for the consideration of the varying hydraulic requirements of fish during different life activities, enabling the evaluation of restoration methods accordingly. Thirdly, when assessing the improvement of fish migration pathways, the paper simplifies the calculation methods of the four indicators (PPA, HFI, HCI, and CQO) mentioned in reference [9] based on two-dimensional spatial grid characteristics, which is also original.

In the upper reaches of the Yangtze River, many river sections have bends and branches. When implementing habitat restoration projects in these sections, engineering plans must be tailored to adapt to the unique navigation conditions and topographic features of each area. Despite this, the paper introduces a habitat enhancement assessment model based on numerical calculation principles, providing valuable insights for evaluating the results of other engineering works.