Assessment of suitable minimum ecological flow downstream of a dam in the Upper Yellow River based on native fish conservation

In recent decades, with the continuous development of industrialization and urbanization, the requirements for water resources and energy have increased dramatically (Qiu & Hu 2018; Hough et al. 2022). Human beings have built dams, weirs, and different structures in the rivers for water supply and power generation to fully exploit and utilize water resources. However, the presence of these hydraulic projects modified the hydrological regimes of the natural river (Bednarek 2001; Zhao et al. 2018; Homsi et al. 2020; Gălie et al. 2021; Yan et al. 2021), and thus threaten the health and sustainability of the riverine ecosystems (Anderson et al. 2017; Ostad-Ali-Askari et al. 2019; Shamshirband et al. 2019). To balance the contradiction between the economic and social water use and the ecological environment, and further ensure the health of river ecosystem and habitat integrity, ecological flow came into being (Poff & Matthews 2013; Gates et al. 2015). The science and practice of the ecological flow as an approach for protecting important ecological services and ensuring the sustainable development of human socioeconomics by managing water flow regimes is of great practical significance (Arthington et al. 2018; Wu et al. 2022). Especially, for the increasingly constructed hydropower stations, the determination of ecological flow is not only the premise and basis for studying their ecological operation, but also one of the core requirements of sustainable hydropower development (Qiu & Hu 2018; Zhang et al. 2019; Yu et al. 2021).

It is predicted that, by 2050, the global water demand will increase by 55% (Lucinda et al. 2016), the growing large dams and hydropower stations worldwide will be built to meet the demand for water (ICOLD 2020), which further aggravates the contradiction between water use inside and outside the river (Li et al. 2013; Salik et al. 2016). Thus, ecological flow, which can reflect the flow-ecology relationships within river systems, has become a hot issue in the study of how to better manage water resources and protect riverine ecosystems (Mezger et al. 2021). Relevant experts and scholars have performed numerous studies on ecological flow assessment over time, and many methods have emerged, which have mainly been differentiated into hydrological methods, hydraulic methods, habitat simulation methods, etc. (Jowett 1997; Ahmadi-Nedushan et al. 2010; Pastor et al. 2014; Bussettini & Vezza 2019).

The hydrological method is widely used because of the availability of hydrological data (Hughes & Hannart 2003), included in, Tennant, Texas, etc., and it is considered as the simplest approach (Li & Xu 2012; Bussettini & Vezza 2019). But the obtained ecological flow value is small and can only support the short-term basic survival of fish and other aquatic organisms (Zou & Wang 2007). The hydraulic method measures the changes in simple hydraulic variables of single sections of a particular river, to substitute habitat factors known or assumed to be target biological limiting factors, aimed at determining the minimum or preservation flows required (Caissie & El-Jabi 2003; McDonough et al. 2017), the wetted perimeter and R2-Cross method are widely employed. In the case of complete filed data, more detailed hydraulic data can be provided for the habitat simulation method (Dunbar et al. 1997). The habitat simulation method, the second most widely adopted method worldwide, determines ecological flow based on the hydraulic conditions required by indicator species, which includes IFIM and CASiMiR (Noack et al. 2013; Ma et al. 2020). As a natural extension of the hydraulic method, the habitat simulation method is considered a valuable tool for assessing instream flows, but it tends to be applied for targeted species and for specific life stages (Jowett & Davey 2007).

To obtain a more reasonable ecological flow, it is important not to rely on only one method, but rather on the best available knowledge pertaining to all instream flow approaches (Caissie & El-Jabi 2003). Hence, the combined approach is adopted to evaluate ecological flow, such as the combination of two or more methods of hydrological, hydraulic, and habitat simulation (Zhao et al. 2020), aimed at providing a more suitable physical habitat for aquatic organisms.

With the increasing awareness of ecological protection, we should not only pursue efficient utilization of water resources and maximization of economic benefits for the numerous water conservancy projects in the Upper Yellow River, but also pay attention to the protection of the health and sustainability of river ecosystems. In the proposed C hydropower stations, native fish were considered as indicator species of the reach downstream of the dam for ecological conservation through field surveys. The 2D shallow water model with high-efficiency and high-resolution, the ecological flow calculation methods combined with Tennant, the hydraulics method, and habitat suitability model were applied. Habitat conditions, such as hydraulics, flow patterns, and river morphology of the reach during non-spawning periods and effective habitat areas of native fish during the spawning period under different discharges were analyzed. Through the detailed analysis, a more reasonable and suitable minimum ecological flow downstream of the dam was obtained to construct and develop eco-friendly hydropower projects.

According to the statistical analysis of the runoff series from 1919 to 2016, the mean annual runoff at the dam site of C hydropower station is 629 m³/s and the average annual discharge of the low flow year is 424 m³/s. The Tennant method is applied to propose the recommended minimum ecological flow with fixed percentages based on the average annual flow of 629 m³/s. Combined with the measured flood surface line survey, the downstream water levels of the study reach under the corresponding discharges are obtained by interpolation, as shown in Table 2.

Through the field surveys of the habitat status of native fish under natural conditions, the cross-sectional hydraulic parameters are studied, to pay attention to the non-spawning habitat conditions of native fish. According to the living habits of native fish on the plateau, species, body length, feeding reproductive habits, etc., as well as the preferred flow pattern of fish, it is found that cross-sectional basic hydraulic parameters, such as maximum water depth ⁠, average water depth ⁠, average velocity ⁠, water width B, discharge section area A, and wetted perimeter W, have a certain impact on fish habitats. Meanwhile, the flow pattern can be divided in detail into rapids (≥1 m/s), relatively rapids (0.5–1 m/s), relatively tranquil (0.3–0.5 m/s), and tranquil flow (<0.3 m/s) based on the cross-sectional average velocity. Furthermore, on the basis of field observation and judging by the cross-sectional maximum water depth, pool and riffle are identified. When the cross-sectional maximum water depth is greater than 10 m, it is considered as a pool. While the maximum water depth is less than 0.5 m in the range of 5 m, and the bank slope near the river is less than 10°, it is treated as a riffle.

Generally, the cross-sectional maximum water depth should be considered as 2–3 times the total length of fish (the longest fish is 46.2 cm) to meet the swimming and survival requirements of the longest fish. So the maximum water depth is greater than 1.4 m. Through the field survey and analysis, compared with the average annual discharge of 424 m³/s in low flow years, minimum standards for other hydraulic parameters, flow form, and river morphology for the survival of native fish are acquired, as shown in Table 3. Here, the water surface area radio is the percentage of the cross-sectional water surface area at a certain discharge to the discharge of 424 m³/s. Meanwhile the wetted perimeter percentage is the percentage of the cross-sectional wetted perimeter at a certain discharge to the discharge of 629 m³/s. The number of cross sections M with the rapid and relatively rapid flows and the number of cross sections with riffle are also considered.

Table 3

Minimum requirements for basic hydraulic parameters, flow form, and river morphology required by native fish of the cross sections in the reach

Habitat parameters .	Standard .	Percentage of minimum river reach .
(m)	≥1.4	95%
(m)	≥0.5	95%
(m/s)	≥0.3	95%
(m)	≥30	95%
(%)	≥50	95%
(m2)	≥30	95%
(%)	≥70	100%
	No obvious change	Reduction rates of river lengths with rapids and relative rapids <20%
	No obvious change	100%

Habitat parameters .	Standard .	Percentage of minimum river reach .
(m)	≥1.4	95%
(m)	≥0.5	95%
(m/s)	≥0.3	95%
(m)	≥30	95%
(%)	≥50	95%
(m2)	≥30	95%
(%)	≥70	100%
	No obvious change	Reduction rates of river lengths with rapids and relative rapids <20%
	No obvious change	100%

View Large

Moreover, basic hydraulic parameters of the selected cross sections of the study reach were calculated through the hydraulic method. According to the calculation results, the percentages of cumulative river length conforming to the standards are analyzed statistically (see Table 5). Combined with Table 3, it can be seen that at the discharge of 62.9 m³/s, the cumulative river length ratio that meets the minimum standard of cross-sectional maximum water depth, average water depth, average velocity, and wetted perimeter percentage was 92.12, 93.76, 93.83, and 93.43%, respectively. Because these percentages should not be less than 95% as required, it showed that the above cross-sectional hydraulic parameters did not meet the requirements at this discharge. While the discharge is 75.5 m³/s, all basic hydraulic parameters met the requirements of minimum standard and the proportions of cumulative river length. As the discharge increased, all basic hydraulic parameters were also suitable for native fish.

Table 5

Statistical results for cross-sectional basic hydraulic parameters in the reach

Discharge (m3/s) .	Ratio of river reach of habitat parameters (%) .						(%) .
	.	.	.	.	.	.
62.9	92.12	93.76	93.83	96	93.43	100	75.1
75.5	96.27	98.24	96.23	97	100	100	78.8
87.5	98.64	99.35	98.63	98.90	100	100	78.8
94.4	100	100	100	100	100	100	78.8
106.9	100	100	100	100	100	100	85.8
125.8	100	100	100	100	100	100	87.2

Discharge (m3/s) .	Ratio of river reach of habitat parameters (%) .						(%) .
	.	.	.	.	.	.
62.9	92.12	93.76	93.83	96	93.43	100	75.1
75.5	96.27	98.24	96.23	97	100	100	78.8
87.5	98.64	99.35	98.63	98.90	100	100	78.8
94.4	100	100	100	100	100	100	78.8
106.9	100	100	100	100	100	100	85.8
125.8	100	100	100	100	100	100	87.2

View Large

Based on the cross-sectional average velocity, the number of cross sections, and the cumulative river lengths with different flow patterns, the number of cross sections of riffles under different discharges were counted, as shown in Tables 6 and 7. Compared with the discharge of 424 m³/s in the low flow year, it could be seen that when the discharge was greater than 62.9 m³/s, the reduction rates of cumulative river lengths with rapids and relatively rapids flow were less than 20%, and the number of cross sections of riffle did not much vary. Thus, it was concluded that flow pattern and riffle habitat suitable for survival of native fish in the non-spawning periods could be satisfied when the discharge was greater than 62.9 m³/s.

Table 6

Flow patterns of the study reach under different discharges

Flow pattern .		Discharge (m3/s) .
		62.9 .	75.5 .	87.5 .	94.4 .	106.9 .	125.8 .
Rapids and relatively rapids flow		10	11	11	11	12	12
	River length (km)	7.16	7.91	7.91	7.91	8.55	8.55
	Reduction rates of river lengths	12.86	5.92	5.92	5.92	0	0

Flow pattern .		Discharge (m3/s) .
		62.9 .	75.5 .	87.5 .	94.4 .	106.9 .	125.8 .
Rapids and relatively rapids flow		10	11	11	11	12	12
	River length (km)	7.16	7.91	7.91	7.91	8.55	8.55
	Reduction rates of river lengths	12.86	5.92	5.92	5.92	0	0

View Large

Table 7

River morphology of the study reach under different discharges

River morphology .		Discharge (m3/s) .
		62.9 .	75.5 .	87.5 .	94.4 .	106.9 .	125.8 .
Riffles		3	2	2	2	2	1

River morphology .		Discharge (m3/s) .
		62.9 .	75.5 .	87.5 .	94.4 .	106.9 .	125.8 .
Riffles		3	2	2	2	2	1

View Large

As can be seen from Table 8 and Figure 7, at low discharge, effective habitat areas both in the concentrated spawning reach and the entire study reach began to increase with the increase of discharge. When the discharge increased to a certain value, effective habitat areas decreased to some extent with the continuous increase of the discharge. Then, it began to increase again and showed a decreasing trend after reaching the maximum values as the discharge further increased. Combined with hydraulic parameters of several sections obtained by the hydraulic method, it was found that the reason for this phenomenon was the existence of some pools and riffles in the study reach, affecting the flow conditions. Thus, it was indicated that pool and riffle had a certain impact on fish habitat. Table 8 shows that when the discharge was 503.2 m³/s, the effective habitat area (WUA) suitable for spawning and breeding of native fish has reached its maximum, which was 322,659.074 m² in the study reach and 264,601.299 m² in the concentrated spawning reach, respectively. It showed that the physical habitat of river is very suitable for fish spawning under this discharge.

Meanwhile, according to the results in Table 8, the percentages of effective habitat areas under different discharges were calculated based on the maximum effective habitat area. It can be seen from Figure 7 that when the discharge was 87.5–629 m³/s, the effective habitat area of the whole study reach accounted for more than 80% of the maximum. In which, the effective habitat area of the concentrated spawning reach also reached more than 90% when the discharge was 87.5–139 and 338–562.5 m³/s, which put the spawning and breeding habitat of native fish in a better state.

To realize the sustainable development of river ecosystems, the exploitation of hydropower resources should take into account water ecological protection and include reasonable planning. In this paper, for the proposed C hydropower stations of the Upper Yellow River, the 2D shallow water model and ecological flow calculation methods combined with Tennant, hydraulics method, and habitat suitability model were applied to focus on the impact of hydropower development on the survival and reproduction of native fish in the study reach. The habitat conditions of the reach considering fish survival during non-spawning periods and effective habitat areas of native fish during spawning periods under different discharges were analyzed. Through detailed analysis, the following conclusions can be drawn:

• I.

  Based on the standards of cross-sectional habitat conditions obtained from field surveys that meet the survival requirements for native fish, the ecological flow downstream of C hydropower station should be greater than 75.5 m³/s to be suitable for native fish survival during non-spawning periods.

• II.

  Under the proposed discharges, the effective spawning habitats of native fish are mainly concentrated in the reach close to the dam site and the downstream reach of Yehuxia Gorge. The trends of the weighted used areas of the study reach and the concentrated spawning reach were consistent and showed double peaks with the increase of the discharge. Combined with the hydraulic parameters of several sections obtained by the hydraulic method, it was found that this was due to the wide and shallow river, and the existence of some pools and riffles, which affect the flow conditions.

• III.

  The effective habitat area of the entire study reach accounting for more than 80% of the maximum at the discharge was 87.5–629 m³/s, while the effective habitat area of the concentrated spawning reach even exceeded 90% when the discharge was 87.5–139 and 338–562.5 m³/s, fish habitat can maintain a better state under these conditions.

To sum up, the ecological flow downstream of C hydropower station should be greater than 87.5 m³/s to be suitable for the breeding and survival of native fish in whole life stages. Thus, it is suggested that the suitable minimum ecological flow was 87.5 m³/s to construct and develop the eco-friendly hydropower projects, and to take into account the health and sustainable development of water ecology while meeting human needs. This result not only provides basic data for the optimal operation of C hydropower station based on ecological response but also has important ecological significance for the protection of native fish habitat in the Yellow River Basin.

This work is partly supported by the Numerical Simulation of Flood Process in Urban Areas with Fine Terrain and Lack of Pipe Network Data (Grant No. 52079106); Chinesisch – Deutsches Mobilit taprogramm: High-Resolution Numerical Simulating and Predicting Methods for Urban Floods (Grant No. M-0427); High-Resolution Numerical Simulation of Sediment Carrying Capacity Mechanism and Erosion Process of Over surface Flow on Whole Sand Slope (Grant No. 52009104); Key R&D program of Shaan Province, China (Grant No. 2021SF-484).

Conceptualization and Methodology: L.Y., J.H., Y.T.; Software: L.Y., J.H.; Validation: L.Y., Y.T., P.W., C.S.; Formal analysis: L.Y., J.H., P.W., L.C.; Investigation: Y.L., J.L., S.X.; Data Curation: P.W., L.C., C.S.; Writing – original draft, Writing – review and editing: L.Y., J.H., Y.T.; Funding acquisition: J.H.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.