The Han River Basin (106°12′–114°14′E, 30°8′–34°11′N) is located in the transition zone between the northern subtropical zone and the warm temperate zone of China. This unique geographical position contributes to its diverse landscape, making it one of the most important biodiversity hotspots in China (Wang and Tan, 2017).

This basin features a complex topography characterized by distinct geomorphological divisions. The upper reaches are predominantly mountainous, whereas the middle and lower reaches consist primarily of plains. The significant elevation gradient along the main river channel results in abundant hydropower resources, offering substantial potential for energy development.

The basin is situated within the subtropical monsoon zone, it is characterized by a distinctly seasonal climate. The region experiences an annual mean temperature ranging from 15°C to 17°C, with annual precipitation fluctuating between 700 mm and 1,200 mm.

The main types of vegetation in the Han River Basin include crops, grasslands, evergreen coniferous forests, evergreen broadleaf forests, deciduous coniferous forests, and deciduous broadleaf forests. The primary growing season for the vegetation is from April to October.

The Han River Basin features a well-developed fluvial system characterized by numerous tributaries and sinuous river channels. The flood season occurs from May to October, while the dry season lasts from December to February of the following year.

As the largest tributary of the Yangtze River, the Han River holds significant economic value. It provides essential water resources for domestic and industrial use to dozens of towns and cities. In addition, it serves as the water source for several major water diversion projects, such as the Middle Route of the South-to-North Water Diversion Project and the Hanjiang-to-Weihe River Water Diversion Project. The Middle Route of the South-to-North Water Diversion Project draws water from the Danjiangkou Reservoir to supply Beijing, Tianjin, Hebei and other areas along the route, with an annual water transfer capacity of approximately 9.5 billion cubic meters. In contrast, the Hanjiang-to-Weihe Water Diversion Project channels water from the Huangjinxia Reservoir to the Wei River, with an annual transfer capacity of 1.5 billion cubic meters. The Han River functions as a crucial hub, connecting the southern and northern regions and facilitating east–west integration.

However, in recent years, the construction of water conservancy projects and the increase in the proportion of urban land use have increased rates of pollution and caused frequent algal blooms, and these have led to further habitat fragmentation (Li et al., 2008). Therefore, considering the habitat condition, intensity of human activities, and vertical connectivity, a total of 20 zoobenthos sampling sites were set up along the Han River mainstem from upstream to downstream. Sampling was conducted in May 2023. Because various types of aquatic organisms can reflect the health of the ecosystem more comprehensively, we also collected plankton at 12 of the selected zoobenthos sampling sites (Figure 1).

The data used in this study are shown in Table 1. Water temperature (WT), electrical conductivity (EC), dissolved oxygen (DO), and pH were determined on site using YSI. The LS300-A portable flow velocity meter was used to measure the water velocity (WV) and water depth (DTW). Each parameter was measured three times to minimize errors, and the mean value was used for analysis. Additionally, we collected 1.5 L water samples at each sampling site to measure total nitrogen (TN), total phosphorus (TP), ammonia nitrogen (NH₃-N), and chemical oxygen demand (COD). All parameters were analyzed in the laboratory following standard laboratory procedures.

For zoobenthos, quantitative sampling was performed using a Surber net and the samples were preserved with 10% formalin in 100 mL bottles. For phytoplankton, a 5 L water sampler was used to collect 1.5 L of water, which was then fixed with 20 mL Lugol’s reagent. For zooplankton, a 5 L water sampler was used to collect a total of 50 L of water, which was then filtered through a No. 13 mesh silk net and concentrated to 100 mL in a sampling bottle, fixed with 10% formalin. All samples were sent to the laboratory under controlled conditions for the acquisition of aquatic organism density and biomass data.

To obtain the urbanization level and climatic conditions of the sampling sites, land use data and meteorological data were processed using ArcGIS (ESRI, Redlands, CA). The accuracy of the land use data was 30 m and the extracted land cover types included cropland, woodland, grassland, shrubland, water bodies, artificial surfaces, bare ground, and wetland. Multi-year average precipitation was calculated based on the daily precipitation of 23 meteorological stations in the Han River Basin from 1979 to 2019 and interpolated using the inverse distance weighting method to obtain precipitation at different sampling sites.

Figure 2 presents the spatial characteristics of internal variables along the river. The red labeled points in the horizontal coordinates are sampling sites where all three types of aquatic organisms (zoobenthos, zooplankton, and phytoplankton) were collected. There were significant differences in pH and EC between upstream and downstream, with the mean value of pH upstream being greater than downstream, while the mean value of EC upstream was less than downstream. In contrast, other internal factors did not differ significantly between upstream and downstream. The external variables, including urbanization level and vertical connectivity, were significantly smaller upstream than downstream, while the difference in the multi-year average precipitation was not significant.

The result of Pearson correlation analysis showed that TN was positively correlated with TP (r = 0.53, P < 0.05) and NH₃-N (r = 0.60, P < 0.01) and negatively correlated with EC (r = −0.5, P < 0.05). In addition, TP was positively correlated with WT (r = 0.37, P < 0.05).

Figure 3 shows the changes in relative abundance at the phylum level of different types of aquatic organisms. A total of 75 zoobenthos species belonging to 4 phyla, 41 families, and 70 genera occurred in 18 sampling sites. The density of zoobenthos ranged from 24 ind/m² to 912 ind/m², with an average of 264 ind/m². The sampling site with the lowest density was S10, which was located in the reservoir area of Danjiangkou Reservoir and the sampling site with the highest density was S1, which was the closest to the headwater. The biomass of zoobenthos varied from 2.02 mg/L to 241.13 mg/L, with an average biomass of 43.81 mg/L. The site with the smallest biomass was S11 and the one with the highest biomass was S14. We found a significant difference in zoobenthos community composition between upstream and downstream. The proportion of Mollusca gradually increased and the proportion of Arthropoda gradually decreased from the headwater to the river mouth. Moreover, Arthropoda accounted for the highest percentage at 70.72% while Platyhelminthes accounted for the lowest percentage and were only found at S1.

A total of 63 zooplankton species belonging to 3 phyla, 29 families, and 50 genera were found in 12 sampling sites. The zooplankton community was dominated by Protozoa, with a relative abundance higher than 70%. The density of zooplankton ranged from 216.93 ind/L to 12,366.73 ind/L, with an average of 4,358.34 ind/L. S13 was the sampling site with the lowest zooplankton density while S5 had the highest density. The biomass of zooplankton ranged from 0.005 mg/L to 4.03 mg/L, with an average of 1.27 mg/L. The sampling site with the lowest biomass was S13 and the highest was S5. Similar to the zoobenthos community, the zooplankton community composition differed between upstream and downstream, with the abundance of Rotifer gradually decreasing.

A total of 64 phytoplankton species belonging to 6 phyla, 23 families, and 40 genera were found in the 12 sampling sites. The phytoplankton community was dominated by Bacillariophyta which accounted for 68.26% of the community. The density of phytoplankton varied from 7.95 × 10⁵ cells/L to 853.46 × 10⁵ cells/L, with an average density of 181.35 × 10⁵ cells/L. The biomass varied from 0.496 mg/L to 32.651 mg/L, with an average density of 4.945 mg/L. Interestingly, S12, located in the Danjiangkou Reservoir, had both the lowest density and lowest biomass of phytoplankton, while S2, close to headwater, had the highest density and highest biomass. In general, the density and biomass of aquatic organisms upstream were greater than those downstream, and the richness was also relatively greater.

The local diversity showed a clear trend from the headwater to the river mouth (Figure 4). Interestingly, the four diversity indexes of zoobenthos, zooplankton, and phytoplankton exhibited similar trends from upstream to downstream, initially decreasing, then increasing, and finally decreasing again. However, the Margalef index of zooplankton showed a distinctly different trend compared with the other three diversity indexes. The discrepancy is likely due to the significantly higher richness of zooplankton in the midstream and downstream, resulting in a sharp increase in the Margalef index.

The maximum Shannon–Wiener indexes for zoobenthos, phytoplankton, and zooplankton occurred at S2 (2.30), S12 (2.60), and S4 (2.61), respectively, and the minimum occurred at S17 (0.88), S17 (0.94), and S18 (0.51), respectively. Normally, a higher Shannon–Wiener index indicates higher biodiversity at the sampling site. Similar to the Shannon–Wiener index, the maximum Simpson indexes for zoobenthos, phytoplankton, and zooplankton occurred at S2 (0.88), S12 (0.90), and S4 (0.91), respectively, and the minimum occurred at S19 (0.22), S17 (0.40), and S18 (0.22), respectively. The Simpson index indicates higher diversity with larger values and emphasizes the relative abundance of dominant species. The maximum Pielou indexes for the three types of aquatic organisms were found at S11 (0.97), S12 (0.84), and S4 (0.86), respectively, while the minimum were at S19 (0.32), S11 (0.44), and S18 (0.20), respectively. Unlike the Shannon–Wiener index and Simpson index, the Pielou index measures the evenness of species distribution, with higher values indicating a more even distribution. We also found that the maximum Margalef indexes for the three types of aquatic organisms occurred at S1 (2.20), S12 (1.55), and S10 (2.59), respectively. The minimum indexes occurred at S9 (0.54), S17 (0.43), and S13 (1.30), respectively. The Margalef index considers the number of species and the total number of individuals at the sampling site, with larger values indicating a site with fewer individuals but more species. Overall, the mean diversity indexes of the three types of aquatic organisms showed a geographical pattern of being highest upstream, followed by the midstream, and lowest downstream.

Figure 5 shows the results of zoobenthos, zooplankton, and phytoplankton resilience. The resilience of zoobenthos and zooplankton showed an obvious trend of higher values upstream compared with downstream, while the resilience of phytoplankton did not show significant variations. Excluding the reference site, the resilience of zoobenthos ranged from 0.46 to 0.98 and had an average value of 0.77, while the resilience of zooplankton ranged from 0.20 to 0.90, with an average value of 0.62. We also calculated the resilience of phytoplankton which varied within the range of 0.38–0.96 and had an average value of 0.70. The results indicated that the biodiversity loss due to disturbance is greatest for zooplankton, followed by phytoplankton, and the least for zoobenthos. We found that the maximum resilience value always occurred upstream and the minimum resilience value always occurred downstream. For instance, the maximum resilience values of zoobenthos and zooplankton were found at S3 and S10, respectively, while the minimum values were found at S12 and S18, respectively.

However, significant differences were observed in the resilience values of the three types of aquatic organisms. At S12, the resilience value of zoobenthos was 0.46, indicating low resilience, while the resilience value of phytoplankton was 0.96, indicating high resilience. Thus, the resilience of a single type of aquatic organism could not reflect the real status of ecosystem resilience. On the basis of the variability of the three aspects of resilience, comprehensive resilience was adopted to represent the state of ecosystem resilience. Comprehensive resilience ranged from 0.48 to 0.82, showing a pattern of higher values upstream compared with downstream. This indicated that upstream experienced lower disturbance intensity compared with downstream.

Figure 6 shows the resistance results of the three types of aquatic organisms and there is an obvious decreasing trend from upstream to downstream. The average resistance of zoobenthos, zooplankton, and phytoplankton were 0.57, 0.53, and 0.50, respectively. Although the average resistance of the three types of aquatic organisms was relatively similar, there were substantial discrepancies among different types of aquatic organisms at the same sampling site, especially downstream. Different types of aquatic organisms showed varying sensitivities to disturbance, resulting in changes in biodiversity and resistance. Therefore, we adopted the measure of comprehensive resistance which ranged from 0.21 to 0.98 and had an average of 0.53. Comprehensive resistance showed a decreasing trend from upstream to downstream, similar to the pattern of resilience. The results suggested that the intensity of disturbance was lower upstream compared with downstream.

Comprehensive resilience and resistance showed a significant correlation with each other (r = 0.99) (Figure 7). The main environmental factors that affected resilience and resistance were TN, TP, EC, and DO. Among them, comprehensive resilience was positively correlated with TN (r = 0.71), TP (r = 0.66), and DO (r = 0.72), while it was negatively correlated with EC (r = −0.77). Similarly, comprehensive resistance was positively correlated with TN (r = 0.69), TP (r = 0.67), and DO (r = 0.72), while it was negatively correlated with EC (r = −0.74).

The RDA model showed a significant relationship overall between internal and the spatial distribution of zoobenthos, zooplankton, and phytoplankton (P < 0.05) (Figure 8). The spatial distribution of zoobenthos was affected by six environmental factors, including DO, TN, WV, TP, EC, and WT. The distribution of Mollusca was positively correlated with EC, and the distribution of Arthropoda was positively correlated with WV. The first and second axes explained 56.16% and 3.94% of the variation in the zoobenthos distribution, respectively. TN, TP, EC, DTW, WV, and DO were the main environmental factors affecting the distribution of zooplankton with the first and second RDA axes explaining 64.39% and 14.51% of the community variation, respectively. EC, DTW, and WV explained significant proportions of the total variation in the zooplankton community, with Protozoa mainly influenced positively by DTW and WV and negatively influenced by EC. The environmental factors affecting phytoplankton biomass were relatively numerous, with the first and second axes explaining 64.72% and 15.28% of the community variation, respectively. The distribution of Bacillariophyta was positively correlated with TN, TP, and pH. Among the various environmental factors, EC, DO, and TP were the main factors affecting the communities of the three types of aquatic organisms.

On the basis of the RDA results and the Pearson correlation analysis results between resilience, resistance, and environmental factors, we selected TP, EC, and DO as the three main environmental factors. Through the urbanization level, precipitation, dam construction, and the selected environmental factors, the SEM explained 96.7% and 93.5% of the spatial variation in resilience and resistance, respectively (Figure 9). The model’s Goodness of Fit Index (GFI) and Comparative fit index (CFI) were greater than 0.95, and the Standardized Root Mean Square Residual (SRMR) was equal to 0.05, indicating a good fit. Urbanization level, hydraulic engineering construction, and precipitation had significant negative impacts on ecosystem stability, and urbanization was the most influential factor. Moreover, hydraulic engineering could indirectly affect resilience and resistance by altering environmental variables. Of all the environmental factors, TP was the most significant factor with an impact coefficient of 0.729 on resilience and 0.811 on resistance. The elevations of TP, DO, and EC contributed to enhancing the ecosystem stability, which was characterized by high resilience and high resistance.

Quantifying ecosystem stability is vital for sustainable ecosystem management. In this study, we applied the methods for calculating the resilience and resistance of terrestrial ecosystems to river ecosystems and used changes in biodiversity to reflect the stability of river ecosystem. Considering the continuity of river ecosystems, we redefined resilience and resistance. Resilience indicates the ability of an ecosystem to recover to its original state after disturbance. In this study, it referred to the difference in biodiversity between other sampling sites and the reference site. Resistance usually refers to the ability of an ecosystem to withstand external disturbances while maintaining its structure and function (Domínguez-García et al., 2019). We used the range of biodiversity changes as a measure of sensitivity to quantify resistance. Correlation analysis results in 3.4 showed a positive relationship between resilience and resistance (r = 0.99), indicating that for an ecosystem, the smaller the degree of recovery needed, the stronger its resistance when disturbed (Domínguez-García et al., 2019). Interestingly, we found that sampling sites with high resilience and high resistance always occurred upstream (S3 and S4) rather than downstream (S18 and S20). This aligned with the general trend that upstream areas typically have better natural environments and experience less human interference compared with downstream (Wang J. et al., 2021).

We proposed a comprehensive method for evaluating resilience and resistance to disturbance based on biodiversity. In previous studies, the stability of river ecosystems was calculated based on the biomass variation of a single species. However, in our study, we found that the average resilience value calculated using zoobenthos was 0.77, which was higher than the values obtained using phytoplankton and zooplankton, at 0.70 and 0.62, respectively. Interestingly, a similar pattern was observed in the results of resistance. The average resistance value calculated using zoobenthos was 0.57, which also exceed the values for phytoplankton and zooplankton, at 0.50 and 0.53, respectively. This indicates that the diversity of zoobenthos exhibits smaller variations compared to phytoplankton and zooplankton. We speculate that this may be related to the differences in the mobility of species. Most of the zoobenthos we collected during this period were aquatic insects belonging to Diptera, Ephemeroptera, and Trichoptera. In spring with subtropical regions, these insects are predominantly in their aquatic larval stage, with limited movement confined to river channels and relatively weak migration and reproductive capacities (Krajenbrink et al., 2019). However, during the summer, they develop into terrestrial adults with enhanced aerial migration capabilities, allowing them to travel long distances by flight (Li et al., 2023).

Moreover, we also found that there are differences in the resilience and resistance of different types of aquatic organisms at the same point. At S13, the resistance calculated using zooplankton was 0.14 while the value for zoobenthos and phytoplankton was 0.92 and 0.78, respectively. Similarly, at S18, the resilience calculated using zooplankton was 0.20 while the value for zoobenthos and phytoplankton was 0.76 and 0.50, respectively. These results supported the notion that the ecosystem stability cannot be assessed solely based on a single type of organism. Therefore, when using biodiversity to quantify resilience and resistance, it is essential to select types of aquatic organisms from different trophic levels to comprehensively reflect the status of the ecosystem.

The results of SEM revealed that urbanization and hydraulic engineering in human activities, as well as precipitation among climatic factors, not only directly affect ecosystem resilience and resistance but also exert indirect impacts by altering internal variables of river ecosystem. Moreover, urbanization was the most significant factor affecting ecosystem stability with a direct impact coefficient of 0.435 on resilience and 0.507 on resistance. Urbanization-driven changes in land use will alter the nature and structure of the underlying surface. Following up, increased impermeable surfaces reduce vegetation cover along the river, leading to increased surface runoff that carries non-point source pollutants into the river, thereby resulting in the deterioration of water quality (Liao et al., 2018). Our results showed that increasing urbanization levels can increase nutrient levels (TP) in the river which is one of the main environmental factors influencing phytoplankton biomass, particularly affecting Bacillariophyta. Excessive nutrient inputs can lead to eutrophication, causing rapid proliferation of phytoplankton and the formation of algal blooms, which result in oxygen depletion in the water environment and mass mortality of other aquatic organisms (Andersen et al., 2020; dos Santos et al., 2023). In the downstream of the Han River, Bacillariophyta is the dominant species and the frequent occurrence of Bacillariophyta blooms is one of the most prominent ecological problems (Xin et al., 2020). Therefore, nutrient emissions caused by urbanization should be controlled to prevent algal blooms. Also, correlation analysis results showed that urbanization was negatively correlated with the biomass, abundance, and richness of the three types of aquatic organisms. In this study, we did not find any zoobenthos at site 16 which surrounded by extremely high urbanization levels. Previous studies indicated that increasing urbanization levels could reduce species richness. Yang et al. (2022) found that in the Houxi River increasing urbanization levels could significantly reduce phytoplankton species richness and resource use efficiency. Yang et al. (2024) and Qiao et al. (2022) also found that increasing the level of urbanization decreased the taxonomic diversity and functional diversity of fishes. These results are consistent with our findings because the richness and diversity of the three types of aquatic organisms decreased with increasing urbanization levels in the Han River.

The construction of dams also affects the level of nutrients in rivers because hydraulic engineering alters the hydrodynamic conditions of water, limiting the transport of nutrients from upstream to downstream (Maavara et al., 2020). We found that the construction of reservoirs in the Han River increased the retention of TP, leading to a decrease in TP fluxes downstream. For example, the average TP content of the sampling sites below the Danjiangkou Reservoir was 0.026 mg/L, which was only half of that upstream. Moreover, the construction of dams also fragments river integrity, alters hydrological conditions, and affects aquatic community structure (Cheng et al., 2015). Around reservoirs and dams with lower flow velocity, a large number of Chironomidae and Oligochaeta species, which prefer still water, appeared. Results in 3.1 and 3.2 showed that the upstream of the Han River, with fast flow and high DO, created a suitable living environment for Ephemeroptera, Plecoptera, Trichoptera (EPT) taxa. Therefore, EPT taxa were mostly found in sampling sites near the headwater without reservoir construction. In addition to the cascade reservoirs, the operation of water diversion projects along the Han River has changed the hydrological factors, exacerbating environmental pollution. Hydrological factors such as flow velocity and discharge can directly influence aquatic organisms and can indirectly disrupt community structure by affecting habitats, thus influencing the ecosystem stability (Descy et al., 2017). Therefore, the impacts of these hydrological factors on resilience and resistance should be considered.

Besides human activities, changes in climatic conditions can also affect ecosystem stability. Increased precipitation leads to higher river velocity and flow, enhancing water exchange and diffusion, which in turn introduces organic pollutants such as nitrogen and phosphorus from the land into the river (Liu et al., 2024b). As a result, the levels of DO, nutrients, and conductivity decrease (Liu Y. et al., 2024). However, this depletion tends to be in excess of nutrient inputs (Smith, 2003). Therefore, increased precipitation negatively impacts the diversity of aquatic organisms, and reduces ecosystem stability. Our findings are consistent with the findings of Kim et al. (2016). Our findings were also related to the severe non-point source pollution in the Han River. Pollutants from non-point sources, such as fertilizers used in agricultural production and heavy metals emitted from vehicle exhaust, are washed into the river by rain and snowmelt runoff, leading to a deterioration in water quality and aquatic biodiversity loss (Sadiqi et al., 2024).

Unraveling how ecosystem stability responds to external variables, such as human activities and climatic factors, is beneficial for the conservation and restoration of river ecosystems. Our findings support the adoption of targeted management strategies based on the resilience and resistance mechanisms to human activities, precipitation, and water quality parameters. For instance, we found that urbanization is the most significant factors affecting ecosystem stability. To reduce urbanization-induced changes in the natural shape of riparian zones, restoring river morphology should follow the “bend where appropriate” principle. Additionally, we can reduce artificial water barriers and expand habitat areas for aquatic organisms to maintain ecosystem stability. Reintroduce large aquatic plants can address biodiversity loss from urbanization and dam construction and enhance the self-purification capacity of rivers as they can not only absorb heavy metals and pollutants but also provide essential habitats (Favas et al., 2018; Xiao et al., 2021).

Moreover, our results indicate that TP is also an important factor influencing ecosystem stability, and the input of non-point source pollution in rivers can lead to elevated TP concentration. Thus, to manage non-point source pollution caused by runoff from rainfall, which is closely related to land use and urban domestic and industrial water use, it is necessary to control urban sewage discharge and the rate of urban expansion to maintain good water quality.

We also found that hydraulic engineering increases nutrient accumulation in reservoir areas and reduces nutrient levels in the lower reaches of rivers, which may lead to algal blooms. Therefore, we suggest improving river flow velocity in reservoir areas and reducing nitrogen and phosphorus loads. In the Han River, the abundance and biomass of toxic cyanobacteria, such as Microcystis, are extremely high. Their toxins, however, can affect water safety, necessitating the monitoring of toxic cyanobacteria trends to ensure the safety of drinking water.