Comprehensive evaluation of watershed ecohydrological risk and ecohydrological satisfaction characteristics from the perspective of hydrological variability

From a hydrological perspective, flow regimes directly influence the vertical and horizontal structure of aquatic ecosystems, and the flow processes throughout the year serve as important stimuli for aquatic biota's life activities (Thompson et al. 2021). Indicators characterizing flow regimes are not only components reflecting the characteristics of river processes but also important factors highly correlated with ecology. Many researchers use flow regime indicators to describe and analyze the changes in river hydrological conditions and their ecological impacts (Masikini et al. 2018; Su et al. 2021; Gocic & Amiri 2023). Early studies grouped indicators by the variability of flow, the pattern of flood conditions, and the degree of intermittent conditions (Hart & Finelli 1999; Shiau & Wu 2004). Richter et al. (1996) proposed five statistical feature types to describe hydrological changes caused by human impacts and established indicators of hydrologic alteration (IHA). These categories include the magnitude of flow, the magnitude and duration of the annual maximum, the timing of the maximum condition, the rate and frequency of flow changes, and the frequency and duration of high and low flow pulses (Richter et al. 2006). Later scholars used these five categories to study rivers worldwide (Wu et al. 2017; Gabiri et al. 2018; Cui et al. 2020). However, human activities on a global scale have been continuously intensifying, altering the watershed hydrological processes and the distribution of water resources through changes in land use, vegetation cover, and hydropower projects (Best 2019; Zhang & Yu 2021). Regarding flow regimes in rivers, reservoir regulation has been identified as a significant stressor (Su et al. 2020). With the development of clean energy sources like hydropower, the water conservancy project's regulatory impact on rivers has been strengthened, resulting in increased hydrological complexity, especially in the overall variation and distribution of flow processes throughout the year. The existing IHA indicator system lacks descriptions for these aspects, necessitating a more comprehensive supplement to the hydrological indicator system to adapt to the changing environment.

The variation of flow sequences contains periodic features and local fluctuations at various scales, exhibiting a multi-time scale structure (Zhang et al. 2014). Moreover, the hydrological processes in watersheds are complex, with flow regimes in rivers consisting of floods and low flows at different frequencies, along with external driving forces. At the intra-annual scale, the changes in flow processes display high complexity and uncertainty (Zhang et al. 2019). Entropy is a measure of disorder and can evaluate the overall system changes from a global perspective. Zhang et al. (2019) used multi-time scale entropy to analyze the complex fluctuation characteristics of water and sediment loads and qualitatively assessed the impact of reservoir operation. Huang et al. (2019) designed a flow diversity index based on Shannon entropy and analyzed the relationship between flow diversity and average fish aggregation, demonstrating that entropy theory is an effective indicator with general ecological significance. Therefore, entropy theory is an effective method that can be used to determine the complexity of changes in hydrological conditions.

The Yangtze River is the largest river in China, with abundant water resources and rich biodiversity (Guo et al. 2011). The Chinese government has carried out large-scale development in the Yangtze River basin over the past century, constructing a number of major hydropower projects. Especially since the 21st century, the number of large-scale reservoirs has increased significantly, and human activities have intensified (Wang et al. 2020a, 2020b; Yan et al. 2021). Among them, the Three Gorges Reservoir (TGR) built on the mainstream of the Yangtze River is the world's largest water conservancy hub (Tian et al. 2019). Since its operation began in 2003, TGR has had a profound impact on the hydrological condition downstream (Wang et al. 2016; Li et al. 2021). Moreover, the annual rainfall in the middle reaches of the Yangtze River has shown an increasing trend over the past 60 years, leading to an increase in the frequency of extreme rainfall events and causing frequent regional floods and droughts (Xue & Zhang 2023). In this case, the water resources required by aquatic ecosystems have changed, and there may be periods of water shortage, increasing the risk of ecological degradation. It is necessary to conduct a multi-time scale assessment of the ecological hydrological risk in the watershed to develop adaptive ecological conservation measures. Additionally, aquatic ecosystems are sensitive to external disturbances, and there is a response relationship between hydrological situation, habitat, and species for river ecosystems (Taylor et al. 2014; Ma et al. 2016). When changes in the hydrological situation exceed the self-regulating and recovery capacity of the biota, species will face the threat of decline, endangered status, and extinction, causing damage to the river ecosystem (Souchon & Tissot 2012; Shen et al. 2018; Hung et al. 2022). However, existing research still lacks a satisfactory characterization of the river's response to hydrological changes.

In conclusion, the purpose of this study is to comprehensively assess the changes in the watershed hydrological conditions and understand the ecohydrological risks and habitat conditions under changing environmental conditions. We have introduced innovation by extending the IHA index system with entropy theory and five additional new indicators to account for the intra-annual distribution characteristics of hydrological processes. Subsequently, we proposed a comprehensive quantification of the degree of hydrological change in the MLYR basin by combining the range of variation approach (RVA) and the river impact (RI) factor method. Furthermore, this study has introduced the ecohydrological satisfaction index (ESI) to characterize the satisfaction level of river ecosystems. By integrating the use of eco-surplus and eco-deficit indicators, we have conducted a comprehensive assessment of the ecohydrological risk and habitat conditions of river systems from a hydrological variability perspective. The results of this study provide useful references for revealing the evolution of river ecosystems under changing environments and potential risks.

The Yangtze River is the largest river in China, originating from the Qinghai-Tibet Plateau with a total length of approximately 6,380 km and a total area of 1.8 × 10⁶ km². The river has a total drop in elevation of around 5,400 m. The MLYR basin includes the Dongting Lake basin, the Han River basin, and the Poyang Lake basin (Figure 1). The Yangtze River basin experiences a prevalent subtropical monsoon climate, with a majority of the region classified as humid. The summer season is characterized by high temperatures and abundant precipitation. This climate is conducive to agricultural production and the region is known for its rich fishery resources, making it an important economic zone in China. The Yangtze River experiences an annual flood season, occurring from May to October, during which approximately 80% of the annual precipitation occurs. The Yangtze River possesses abundant water resources, with an annual average water volume of approximately 9.96 × 10¹¹ m³, accounting for around 35% of China's total water resources. The construction of the TGR has significantly impacted the hydrological conditions and ecological environment of the MLYR.

We selected daily flow data from representative hydrological stations of the MLYR, including Yichang, Hankou, Datong, Chenglingji, Huangzhuang, and Hukou stations, provided by the Yangtze River Water Conservancy Committee. Except for the Huangzhuang station, which has data from 1965 to 2021, the data lengths for the other stations are from 1960 to 2021. Yichang, Hankou, and Datong stations control the hydrological conditions of the mainstream of the Yangtze River. Huangzhuang station is located on the mainstream of the Han River and can represent the hydrological regimes of the Han River basin with a drainage area of 142,100 km². Chenglingji station is the control station for outflow in the Dongting Lake basin, located at the confluence of the Jing River and Dongting Lake, controlling a drainage area of 262,823 km². Hukou station is the outflow hydrological station for the Poyang Lake basin, located at the confluence of Poyang Lake and the mainstream of the Yangtze River, controlling a drainage area of 162,200 km².

Richter et al. (1996) proposed the IHA to describe the changes in the ecohydrological regime of rivers, but with the increasing external pressures, the changes in river systems have become more complex. Traditional IHA indicator systems lack attention to the distribution of flow processes and only focus on rising rates, falling rates, and the number of reversals in terms of variability and frequency, failing to comprehensively characterize the total features of within-year flow variations. In this study, we established 38 indicators from six aspects (Magnitude, Timing, Duration, Frequency, Distribution, and Variability) to describe the changes in the hydrological conditions of the MLYR (Table 1). The newly added indicators include skewness, kurtosis, concentration degree, and concentration period, which characterize the distribution of within-year flow, as well as sample entropy and coefficient of variation, which represent variability and frequency. The distribution of within-year flow reflects changes in river conditions, affecting the suitability of key life stages for aquatic organisms, such as fish spawning periods, and also altering the transport characteristics of nutrients (Dang et al. 2018). Sample entropy and coefficient of non-uniformity describe the complexity and variability of within-year hydrological processes, and complex flow variations can lead to frequent inundation and exposure of riparian floodplains, impacting the survival and reproduction of riparian biota (Zeng et al. 2018).

Similarly, 0 ≤ OA ≤ 1, and OA is more sensitive to changes in flow regime. The RVA method categorizes the degree of hydrological change into the following classes: [0, 0.33]: low degree of change; (0.33, 0.67]: moderate degree of change; (0.67, 1]: high degree of change (Zhang et al. 2015). On the other hand, the RI method classifies the degree of hydrological change influenced by external factors into the following categories: [0, 0.2): drastic change; [0.2, 0.4): severe change; [0.4, 0.6): moderate change; [0.6, 0.8): incipient change; [0.8, 1]: low change (Haghighi et al. 2014). Based on these definitions, we calculate the result of OA and classify hydrological regime changes as follows: [0, 0.4): slight change, [0.4, 0.6): moderate change, [0.6, 0.8): high change, [0.8, 1]: severe change.

Sort the pre-TGR period to obtain the annual FDC and seasonal FDC corresponding to the 75th and 25th percentiles, respectively, as the ecohydrological risk thresholds. The eco-surplus risk is the area above the 75th percentile FDC, normalized according to the method proposed by Gao et al. (2012), while the eco-deficit risk is the area below the 25th percentile FDC, also normalized. These two indicators, respectively, characterize the hydrological elements that are above and below the needs of the riverine ecosystem for that element.

The flow conditions of rivers are crucial for fish habitats. Both excessively high and low flow events are not conducive to the suitability of fish habitats, but higher flows have better ecological effects than lower ones (Gao et al. 2012; Bestgen et al. 2020). Based on the optimum flow conditions, this study proposes an ESI (0 ≤ ESI ≤ 1) to reflect the satisfactory level of water required by the current flow to meet the river habitat, while considering both flood and non-flood hydrological conditions. This method assumes that within a certain range of the optimum ecological flow, the satisfactory level of the river habitat with respect to the flow condition can be evaluated as 1. The farther the flow deviates from the most appropriate flow, the lower the satisfactory level of the river habitat. When the lowest flow condition occurs, the satisfactory level of the river habitat is considered to be 0, and the aquatic ecosystem may experience irreversible degradation. In addition, in order to consider the ecological impact of high flow conditions, we set a reduction coefficient α.

The construction of the TGR has had a profound impact on the hydrological condition in the MLYR. In this study, the research period was divided into a pre-TGR period (1961–2003) and a post-TGR period (2004–2021) based on the regulation time of the TGR in 2003. The changes in ecohydrological indicators in the MLYR were then analyzed.

The frequency of low-flow pulses at Yichang station experienced a high degree change (−91%) after 2003. Most high and low flow pulse events in the MLYR basin showed moderate changes, and changes in high and low flow pulses imply changes in aquatic biological behavior signals and can also affect the soil moisture content on both sides of the river, impacting vegetation succession.

The overall degree of change in the ecohydrological condition was analyzed for the six hydrological stations in the MLYR (Table 2). Yichang, Hankou, Datong, Chenglingji, Huangzhuang, and Hukou stations had 6, 4, 4, 3, 3, and 4 sets of hydrological indicators that showed changes of low degree or higher, respectively. The overall degree of change in the hydrological situation calculated by RVA corresponded to 0.61, 0.57, 0.41, 0.38, 0.43, and 0.46 for each station. Among them, Yichang and Hankou stations had two and one sets of indicators that showed a high degree of change, respectively. In addition, according to the results of the RI method, the RI values of Yichang, Hankou, and Datong stations were 0.53, 0.65, and 0.89, respectively, and those of Chenglingji, Huangzhuang, and Hukou stations were 0.55, 0.56, and 0.71, respectively. Overall, the hydrological regime at Datong station has experienced a moderate change, while the hydrological regimes at the other hydrological stations have undergone a high degree of change.

In terms of the longitudinal view, there is a strong correlation between the annual variations of eco-surplus and eco-deficit at Yichang and Hankou stations, with correlation coefficients of 0.81 and 0.70, respectively. The correlation between Yichang and Datong stations is relatively low, with coefficients of 0.61 and 0.60, respectively. Analyzing the relationship between the changes in the main stem and tributaries, the correlation coefficients of the eco-surplus and eco-deficit variations at Yichang and Chenglingji are 0.53 and 0.51, respectively, while those at Huangzhuang and Hankou are 0.42 and 0.35, respectively. Whether it is eco-surplus or deficit, the connection between the eco-flow index changes of the main and tributaries is strongest at Yichang and Chenglingji, especially in autumn, with correlation coefficients of eco-surplus and eco-deficit variations of 0.70 and 0.70, respectively. These results indicate that the influence of the TGR on the changes in ecohydrological risk indicators is spatially limited, and the regulation and storage effects of the tributaries and lakes mitigate the dam effect.

Natural hydrological processes are the main driving factors for the evolution and maintenance of species diversity in riverine wetland ecosystems. In this study, the IHA index system was supplemented, and the degree of hydrological variation was comprehensively evaluated using the RVA and RI methods, and then the ESI was proposed to characterize the degree of satisfaction of river systems with hydrological conditions. We found that after the operation of the TGR, the concentration degree of flow in the MLYR and the timing of the concentration period were delayed. The SampEn values representing the fluctuations in flow within a year increased. These changes indicate a high degree of alteration in the hydrological regimes in the MLYR basin, which is consistent with (Wang et al. 2015; van Oorschot et al. 2018; Yang et al. 2022). Additionally, according to the indicators proposed in this study, the ecohydrological risk in the mainstream of the MLYR increased, while the satisfaction level of ecohydrology decreased. On the other hand, the ecohydrological risk and ecohydrology satisfaction in the tributaries remained relatively stable (Figures 8 and 11).

For the calculation of the ESI proposed in this study, it is essential to first determine the probability distribution function that best fits the monthly average flow data. The goodness-of-fit test method and the associated index are crucial components of this process. In this study, the K-S method with a significance level of α = 0.05 was used to assess the goodness of fit. The optimal distribution function was determined based on the highest test probability (P) value. Li et al. (2011) conducted research on ecological water demand in the Yellow River and calculated the average P values for each station and each month using the K-S method. They found that the GEV distribution had a higher average P-value (0.885) compared to the P-III distribution (0.757). Jiang et al. (2021a, 2021b) determined eco-flow thresholds for the Laohe River basin using the K-S method and selected the probability distribution function with the smallest test statistic, which was the GEV distribution. According to the results of this study, the P values for the optimal distribution functions of monthly average flow from six hydrological stations in the MLYR ranged between 0.420 and 0.999. Furthermore, P values below 0.7 occurred only nine times. This suggests that the monthly flow data in the MLYR fit well with the GEV distribution. Additionally, the computed P values in this study align with previous research findings, where in most cases, the P values for the optimal probability distributions were greater than 0.7. Therefore, it can be considered that the goodness-of-fit assessment results based on the K-S method are reliable and suitable.

Climate factors, primarily precipitation and temperature, are the main drivers of hydrological changes. Precipitation is the primary source of runoff, directly impacting the total runoff volume within a watershed. Meanwhile, climate factors like temperature indirectly influence runoff by affecting evapotranspiration, which, in turn, leads to changes in the flow regime (Chu et al. 2019). Studies have identified changes in precipitation and temperature within the Yangtze River basin. Reduced water availability is the direct cause of the decline in ecohydrological satisfaction in the MLYR. Additionally, human activities significantly alter the hydrological cycle within the basin. Xia et al. (2017) quantified the impacts of different drivers on runoff changes in the Han River basin. They found that during the period from 1961 to 2013, 56.5–57.2% of the runoff reduction could be attributed to human activities. Yang et al. (2022) indicated that human activities accounted for 57–98% of the impact on runoff in the Dongting Lake basin. Forest and Cropland are the predominant land types in the MLYR basin. The Chinese government has implemented ecological restoration projects nationwide, such as the ‘Grain for Green’ program (Zhang et al. 2020a, 2020b). However, there has been a reduction in forest and arable land within the MLYR basin between 1980 and 2020. Economic development has accelerated urbanization, leading to an increase in urban land area (Guo et al. 2023). Vegetation can influence the infiltration process through canopy interception and transpiration. The reduction in vegetation cover partly contributes to increased runoff. Urbanization results in increased impervious surfaces, hindering surface runoff infiltration and increasing river runoff. Furthermore, reservoir regulation and the construction of water diversion projects also play roles in influencing flow regime changes.

The changes in the annual flow regime provide a stimulus signal for aquatic life activities (Li et al. 2015; Wang et al. 2022). The high degree of change in SampEn at the Yichang station implies the disruption of fish spawning signals within the river section. In addition, with the changes in high and low flow pulses, the dominant species and structure of river ecosystems may undergo succession. Moreover, under the influence of the TGR, the concentration of the flow regime decreased, and the concentration period was delayed, which means that the annual hydrological process was softened, leading to vertical stratification of the river and stable flow regimes, affecting nutrient exchange (Komatsu et al. 2007; Liu et al. 2019). In addition, the risk of algal blooms may increase due to the warming of the river caused by increasing air temperatures. Some studies have found that reservoir operation may be a potential driving factor for the deterioration of the river water environment (Soja & Wiejaczka 2014; Li et al. 2016; Zhao et al. 2021). On the other hand, eco-surplus reflects the change in the high-level flow regime. Its increase means an increase in flow rate and water level, which is conducive to the growth of submerged vegetation and the reproduction of adhesive egg-laying fish (Bi et al. 2020; Larson et al. 2020). However, a high water level may also encroach on the living space of terrestrial organisms. The increase in eco-deficit reflects the change in the low-level flow regime, which may manifest as consecutive years of extremely low water levels in the MLYR during the dry season. Although this is beneficial for some bird species that feed on submerged plant roots and rhizomes, it threatens the survival of benthic organisms (Fan et al. 2020).

In this study, we proposed an ESI based on the optimal flow to evaluate the extent to which the current flow conditions meet the habitat requirements of the river ecosystem. The method distinguishes between flood and non-flood seasons and takes fish, representing the top of the river food chain, as representatives of aquatic organisms. By using a dimensionless formula, it considers the range of the suitable flow and different hydrological conditions deviating from this range. Zhang et al. (2018) proposed a new index (ratio of environmental flow to streamflow) to evaluate habitat quality and environmental flow satisfaction between different rivers and established a river health assessment model. While this index considers the different needs of aquatic organisms in different periods, it overlooks the range of the suitable flow and the fact that high flow conditions have better ecological impacts than low flow conditions. For example, the most suitable flow rate for spawning of Prenant’ sschizothoracin ranges from 0.2 to 0.8 m/s, and spawning is not suitable when the flow rate is below 0.07 m/s or above 1.50 m/s (Shao et al. 2015). Qiu et al. (2023) established the relationship between the habitat area of Four Major Chinese Carps (FMCC) and the flow in the middle reach of the Yangtze River and concluded that the most suitable flow range for spawning of FMCC in the Yangtze River during the spawning season (May–June) was 10,000–22,500 m³/s. The optimal flow determined by the probability density method in this study also falls within this range, and the ESI in low flow conditions is assigned lower values than in high flow conditions (Figure 10). Therefore, our ecohydrological satisfaction assessment method is simple and scientific and provides insights into the current habitat conditions of the river ecosystem from a hydrological perspective. According to the results, the ecohydrological satisfaction in the mainstream of the Yangtze River has significantly declined, which may lead to a reduction in the lateral connectivity of the river, causing extensive fragmentation of habitats and affecting the succession of riparian plant communities.

Qian et al. (2022) investigated the impact of climate and human factors on runoff in the Yangtze River basin and found that human activities were the dominant factor in the runoff changes from 2001 to 2020. The upper reaches of the Yangtze River have significant hydropower development potential, and the Chinese government has planned numerous large-scale reservoirs in this area. By 2030, the total surface area of interconnected cascade reservoirs, including those in the Jinsha River, Yalong River, Dadu River, Jialing River, Wu River, and the mainstream of the upper Yangtze River, is expected to reach nearly 3,000 km² (Zhou et al. 2019). Dam regulation has already altered flow events in terms of magnitude, timing, frequency, and other aspects, including changes in hydrological processes (Li et al. 2015; Peng et al. 2022). In this study, it was found that the ecohydrological satisfaction at the Yichang station significantly decreased, especially in January–February and September–November. The hydrological regime at the Yichang station is mainly influenced by the TGR on the intra-annual scale. With the cumulative effects of the cascade reservoirs continuously strengthening in the future, the risk of ecological degradation at the Yichang station will increase. Additionally, the IPCC report indicates a high likelihood of continued warming in the future (IPCC 2022). Under the context of future climate warming, extreme events may become more frequent and severe, and the propagation of extreme events from the meteorological system to the hydrological system may further disrupt the hydrological situation.

It is worth noting that the MLYR basin is home to China's largest and second-largest lakes (Poyang Lake and Dongting Lake, respectively) and is also an internationally significant wetland conservation area. The ecosystems in this region are sensitive to changes in water resources. Moreover, as lakes are connected to the Yangtze River, the impacts of reservoirs constructed on the main stream can also be transmitted to the hydrological conditions of the lakes (Zhang et al. 2020a, 2020b). In this context, there is an urgent need for a more comprehensive assessment of ecohydrological conditions to understand the river system's degree of satisfaction with the hydrological conditions and the risk of degradation.

The study has some limitations. Three distributions were used to fit the multi-year monthly average flow during the pre-TGR period, and there are many probability density functions to choose from. Different function choices may result in different orders of magnitude for the optimal flow. However, the Gamma distribution is generally considered to be the optimal distribution for Chinese precipitation (Jiang et al. 2021a, 2021b). In this study, the optimal distribution function for streamflow was mainly the GEV distribution, but the fit of the Gamma distribution was only slightly lower than that of the GEV distribution. This is because streamflow is influenced not only by precipitation but also by many factors at the intra-annual scale, such as underlying surface conditions, vegetation, and reservoir construction (Qian et al. 2022). This indicates that the impact of different function choices on the results is minimal. On the other hand, this study did not quantify the specific impacts of human activities and climate change, making it challenging to provide specific conservation measures from the perspective of basin management. In the future, it will be necessary to improve the accuracy of research methods based on the specific protection goals of the basin, and to determine the specific direction of the role of human activities in conjunction with distributed hydrological models.

This study conducted a comprehensive evaluation of the ecohydrological condition of rivers within the basin from multiple perspectives. The traditional IHA index system was complemented with entropy theory and five additional indicators to describe the changes in intra-annual flow distribution. The RVA method and RI method were then employed to evaluate the extent of hydrological changes in the MLYR. Subsequently, ecohydrological risk indicators and the proposed ESI were used to characterize the potential risks associated with hydrological changes in the region.

• (1)

  The hydrological conditions in the MLYR basin showed a high degree of variation. The concentration degree of flow decreased, and the timing of the concentration period was delayed, indicating an increased complexity in the intra-annual hydrological process. In terms of flow regimes, except for the Datong station with an OA value below 0.5, the other five stations all had OA values above 0.6, signifying a high degree of hydrological change.

• (2)

  On an annual scale, the average eco-deficit at the Yichang, Hankou, and Datong stations was higher than the eco-surplus, indicating a high risk of ecosystem degradation, but with a decreasing trend along the river. For the Chenglingji station, the increase in eco-surplus was mainly concentrated in the winter, while the increase in eco-deficit was mainly concentrated in the summer and autumn seasons. For the Huangzhuang station, the eco-surplus increased more in the autumn season. For the Hukou station, except for the summer season, the eco-deficit decreased in other seasons.

• (3)

  From 2004 to 2021, the ESI at the Yichang station showed a significant decrease, with a large-scale low ESI condition occurring from January to February and September to November. The most severe decrease occurred in February, with a decrease of 0.35. Similarly, the ESI at the Hankou and Datong stations also decreased, with the most severe decrease occurring in January (−0.20 and −0.11, respectively). The ecohydrological satisfaction status of the tributary was relatively stable, with the ESI at the Huangzhuang station being above 0.8 for more than 80% of the period from 2004 to 2021.

Future studies should incorporate the suitability of aquatic biota into the overall assessment of river health, taking into account river health risks under medium to long-term future climate scenarios. Additionally, it should integrate meteorological–hydrological models to determine the specific impacts of different influencing factors. This will provide a scientific basis for the development of specific ecological conservation and restoration measures.

This study was supported by the Innovative Research Group Project of the National Natural Science Foundation of China (51779094), and the Henan Province Science and Technology Innovation Talent Program (16HASTIT024), and the Science and Technology Project of Guizhou Provincial (KT202008).

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.