1 Introduction

Rivers provide human beings with various ecological services such as water sources, biological protection, and landscape, and promote the development of cities with their natural, social, economic, and environmental values. However, with the development of human society, the disturbance to rivers is increasing day by day. Dam construction, water intake, diversion, turning straight, blocking the Han River and the solidification of the river bank, and the destruction of the riparian vegetation zone have disturbed the river flow pattern and hydrological cycle process, and have synergistic effects with water pollution and excessive utilization of aquatic organisms, resulting in the degradation of the river ecosystem (Poff et al. 1997; Aguiar et al. 2010 ; Chovanec et al. 2015). The river’s ecological degradation has threatened public interests, which makes it increasingly urgent in social needs to improve the quality of the river’s ecological environment. The health of the river ecosystem has attracted the attention of governments and academia. A healthy river ecosystem has become an important management goal.

The connotation of "health" constitutes the focal point of the conceptual framework for ecosystem health assessment. Due to its ambiguous and abstract nature, the precise definition of ecosystem health has been a subject of controversy (Costanza et al. 1992; Rapport 1989; Peng et al. 2007; Liu et al. 2010). Over the past few decades, scholars have engaged in extensive theoretical discussions on the definition of "health" (Schaeffer et al. 1988; Meyer 1997; Boulton 1999). Previous research has defined ecosystem health from diverse disciplinary perspectives and case studies, which can be broadly categorized into biological ecology definitions and ecological economics definitions. The former underscores the natural ecological aspects of the ecosystem while disregarding socioeconomic and human health factors. The latter regards humans as integral parts of an ecosystem and accounts for the health of the ecosystem, as well as the extent to which it meets human needs and desires, namely, ecosystem services. In the early stages of research, the definition of river health was mainly focused on the natural properties of rivers. Karr defined river ecological integrity as health (Karr 1999), while Simpson defined river health as the main process of support and maintenance by the river ecosystem to restore its previous undisturbed capabilities (Simpson et al. 1999). The undisturbed state of rivers was considered as healthy. However, due to the severe impact of human activities on urban rivers during the process of urbanization and development, it is difficult to return to an undisturbed state even after the ecological restoration of urban rivers. Furthermore, urban rivers not only need to sustain their own ecosystem structure and functions but also need to provide corresponding ecological services for urban residents.

Currently, most scholars maintain that ecosystem health pertains to the ability of regional ecosystems to sustainably maintain spatial structure and ecological processes, self-regulation, restoration, and meet the reasonable needs of human society. Researchers have applied this concept to river health management and have scientifically evaluated the status of river ecosystems from the perspectives of water quality, biology, and ecology, with the aim of enhancing river management. The United States, Australia, the United Kingdom, South Africa, and other countries have successively carried out research and practice on river health evaluation and formed a series of evaluation methods such as RIVPACS, AUSRIV AS, IBI, RCE, ISC, and RHP (Bain et al. 2000; Ladson et al. 1999; Karr 1999; Munné  et al. 2003; Brizga et al. 2000).

As the concept of ecosystem health was applied to urban rivers, the assessment and research on urban river health achieved some development in China. Some Chinese experts learn from the advanced experience of foreign countries to screen different indicators, trying to establish a river health index system and apply it to river management. Such as Zhao and Yang built an index system based on five factors including water quantity, water quality, aquatic organisms, physical structure, and riparian zone, and used it to evaluate river health in the city of Ningbo (Zhao et al. 2005). Deng et al. established the indicator-based assessment system of urban river health and applied it to river management in Lijiang. The system was composed of three first-tier indexes, including natural biology, social economy, and landscape environment, as well as twenty-four second-tier indexes (Deng et al.2014). Peng constructed the ecological assessment system of rivers and lakes based on the five aspects of hydrology and water resources, the physical form of rivers and lakes, water quality, aquatic organisms, and the social service function of rivers and lakes (Peng 2018). Some scholars have considered the impact of human activities when assessing the health of natural ecosystems and incorporated such factors into their models (Liao et al. 2018; Sun et al. 2019), while there has been a lack of reported evaluation models that include the influence of human activities in river health assessments.

The Yangtze Estuary has a unique geographical location, which is affected by both rivers and offshore. The environmental factors are complex and changeable, and the biological communities are rich. There are both broad coastal wetlands and towns with high-density populations, which also makes river ecosystem health management in the region challenging.

Chongming Island is located in the Yangtze Estuary and is surrounded by water, and the island is full of ditches and rivers, and the water system is rich, including urban rivers and suburban rivers, and the river’s ecological environment is complex and changeable. With the development of the social economy, the river ecosystem of Chongming Island encountered some difficulties, And there are differences between different regions, for example, rivers near urban and rural areas are mainly affected by domestic sewage, and human activities, while suburbs are mainly affected by agricultural production (Sun et al. 2009; Qian et al. 2011; Shen et al. 2017). However, in recent years, the situation has developed in a good direction, and the Chongming district government pays more and more attention to river ecological environment protection and adheres to the priority of ecological protection, the green transformation of the industry, and has carried out a series of river ecological governance and restoration projects, the river ecological environment has changed (Zhang et al. 2013a, b; Tian 2021; Xin 2022), in order to better manage the river, as well as build a world-class ecological island, it is necessary to assess the river ecological health status of Chongming Island that research results can provide a reference for river health management.

There is a wide divergence between different regions in terms of urban river characteristics in China which leads to the fact that river health assessment indicators and criteria for a city are not applicable elsewhere. So a regional river ecological health assessment system needs to consider the local actual situation to select the appropriate indicators and standards to establish. Taking into account the aforementioned considerations and in accordance with the concept of ecological health, we propose a definition of river health that draws upon commonly utilized river health indicators both domestically and internationally, while also considering the actual circumstances of the Yangtze Estuary. This definition comprises the following criteria. (1) Healthy river water quality that conforms to the corresponding water function standards, devoid of eutrophication, and without the risk of pollutant discharge into the water. (2) Healthy riverbank environments characterized by a certain level of vegetation coverage, capable of fulfilling ecological buffer functions, providing biological habitats, and contributing to aesthetic beauty and recreation. (3) Healthy aquatic organisms characterized by relatively rich species composition and population size, with a certain level of biodiversity. (4) A physically healthy river possessing a diverse and meandering flow state, maintaining a certain longitudinal continuity, and capable of preserving hydraulic connections without disrupting flood control. (5) The last healthy river received moderate human disturbance.

2 Study area and methodology

2.1 Survey sites

Chongming Island is located in the Yangtze Estuary and is surrounded by water, and the island is full of ditches and rivers, and the water system is rich, urban rivers and rural rivers are interconnected and run through the entire island, but they have their own characteristics (Fig. 1). Urban rivers serve the city and are often densely populated along the banks. Their main functions include flood control, drainage, transportation, recreation and entertainment, environmental improvement, coordination of urban development, and adjustment of local microclimates. In contrast, rural rivers often run through farmland, forests, and wasteland, mainly serving flood control, drainage, irrigation, domestic water supply, aquaculture, ecosystem restoration, and recreation and entertainment needs. In order to establish a more suitable and comprehensive river health assessment system for the Yangtze Estuary, we selected 6 river sections with different significance in Chongming Island. The S1 river section is located in the south-central part of Chongming Island, east–west trending, close to the densely populated Bao town (Fig. 1). It belongs to the urban type of river. The S2 river section is located in the northwest of Chongming island, with both sides of the distribution of ecological farmland and wetland park. It belongs to the countryside suburban rivers, less disturbed by a human. The S3 section of the river is located at the eastern end of Chongming Island, close to the Chongming Dongtan National Bird Sanctuary, with many ecological farms and wetland parks nearby, and it belongs to the countryside suburban rivers, less disturbed by a human. The S4 river section is located at the western end of Chongming Island, near Chongming Xisha National Wetland Park and Mingzhu Lake. Towns and Villages are scattered on both sides of the river, and the impact of human activities is more obvious. The S5 river section also passes through the densely populated area, but the river is north–south, connecting the north branch and the south branch. It belongs to the urban type of river. The water level of the Chongming inland river is flexibly regulated by way of water diversion from the south sluice and drainage from the north sluice. The S6 river section is located in the middle of Chongming Island, which is a small section, but it is connected with Huabo Cultural Park, which makes us very interested (Fig. 1). We carry out satellite remote sensing observations on these river sections to obtain data on the vegetation coverage of the river bank, the proportion of farmland area, the proportion of road area, and the proportion of building area to analyze the impact of human activities. And hydrological surveys such as river width, connectivity, and curvature were conducted on these river sections. In addition, we also select a site in each section of the river water quality, macrobenthos, zooplankton, and algae survey.

Fig. 1
figure 1

Locations of study areas showing river sections and sites, The blue line represents the river network formed by the rivers of Chongming Island

Full size image

2.2 Survey and analytical methods

Drawing on Wu and Wang's research, 12 alternate indicators were selected, and the final indicators were determined after screening through correlation analysis and principal component analysis (Wu et al. 2006; Xu et al. 2022). The data is obtained as follows.

Water quality indicators were monitored in accordance with the Technical Specifications for Monitoring Surface Water and Sewage (HJ/T91-2002) (Fig. 2a).

Fig. 2
figure 2

(a) Collecting water samples; (b) Collecting phytoplankton samples; (c) Collecting zooplankton samples; (d) Preservation and fixation of samples; (e) Collecting macrobenthos samples; (f) Separating and screening samples; (g) Laboratory identification

Full size image

Plankton survey methods are carried out in accordance with the Code for the Survey of Fishery Resources in Reservoirs (SL 167–2014). Three indicators were selected for a comprehensive evaluation, including the number of algae taxa, the Shannon–Wiener diversity index of algae, and the Berger–Parker dominance index of algae. The indicators were standardized first, and then, the arithmetic means the sum of the three indicators was calculated. Finally, get the Comprehensive index of phytoplankton diversity (Fig. 2b, 2c, 2d).

Macrobenthos are obtained by 1/16 Peterson mud miners (large). Samples are fixed on-site (if desired) and then refrigerated and brought back to the laboratory for identification, counting, and biomass determination. The index integration method was the same as that of phytoplankton (Fig. 2e, 2f, 2g).

The extent of vegetation cover and the impact of human activities on the banks is obtained by analyzing satellite remote sensing data.

The satisfaction degree of residents regarding the landscape close to the river and the degree and influence of river reconstruction were obtained by the method of questionnaire survey.

Data on vegetation coverage and human activities such as the proportion of coastal farmland area, the percentage of coastal hardened roads area, and the proportion of coastal buildings area are obtained from remote sensing imagery (Sentine-2 and SuperView-1).

The data of indicators in terms of the physical structure of rivers such as river width, connectivity, and river curvature obtained by field survey measurements.

2.3 Identification of index weight

There are many ways to determine the weight of river health assessment indicators (Geng et al. 2006; Gao et al. 2007; Wang et al 2007; Xia et al. 2007), While studying the quality evaluation of river habitats, Wang found that the evaluation system established by principal component analysis was more in line with the real situation (Wang et al. 2017). Therefore, this study uses the method of principal component analysis to determine the weight of the index. The specific methods are as follows. First, standardize the raw data. Standardization of raw data by minimum range standardization. Then, the correlation coefficient matrix and principal component model are established, the eigenvalues of the principal components of each index are obtained, and the weights of each index are derived. Of course, these calculations can be made with the statistical analysis software SPSS18.0.

2.4 River health assessment

The model of river health assessment is expressed as,

$${I}_{CH}={\sum }_{i=1}^{n}{B}_{i}\times {W}_{i}$$

which \({I}_{CH}\) is the river health evaluation index; Bi is standardized for river health indicator; Wi is the weight of river health assessment indicator. River health standards are divided into five grades, 0.8–1 is excellent, 0.6–0.8 is good, 0.4–0.6 is ordinary, 0.2–0.4 is somewhat inferior, and 0–0.2 is inferior.

3 Results and analysis

3.1 Assessment indicators

In order to avoid overlapping between indicators, the indexes under assessment were further analyzed and selected based on applied correlation analysis and principal component analysis. According to the Pearson correlation coefficient of all indicators, most indicators did not have a significant correlation. Only a few indicators, such as the proportion of hardened roads are significantly negatively correlated with the satisfaction rate of citizens, and there is a significant positive correlation between vegetation coverage and citizen satisfaction, so the indexes of citizen satisfaction were deleted. Besides the proportion of farmland land has a significant positive correlation with water quality and the proportion of coastal buildings. In order to avoid overlapping, the index of the proportion of farmland and cultivated land is deleted.

From the weights of the indicators, we can see that indicators such as water quality, aquatic organisms, river connectivity, and river width play a major role in the comprehensive assessment of rivers (weight > 0.2) (Table 1), while indicators such as vegetation coverage and human impact play a lesser role (wight < 0.05). Furthermore, the indicators such as the index of macrobenthos diversity and the proportion of building area along the coast are negative indicators, that is, the larger their values, the smaller the comprehensive evaluation result of the river.

Table 1 Health assessment indicators of rivers and weight
Full size table

3.2 Comprehensive health index

We can see that there are two sites in summer, namely S2 and S3, the comprehensive health index is greater than 80, reaching a very excellent level, only the S6 site is somewhat inferior level (Fig. 3). In the autumn, the comprehensive health index of most sites is in the good and ordinary level, and no site reaches the excellent level, so the overall comprehensive health evaluation results in summer are better than those in autumn. In summer, the comprehensive results of each site are quite different, while the difference in autumn is reduced, which is more similar to the results of the site.

Fig. 3
figure 3

Results of the comprehensive river health assessment at each site, S1 to S6 represent different river sections

Full size image

In addition, we can also see that, regardless of summer or autumn, the comprehensive health evaluation results of S2 and S3 are the best, while S6 is the worst.

Comparing the comprehensive evaluation results in summer and autumn, it can be found that there are large changes between S1 and S2. By analyzing the radar map of S1, we can find that the increase in phytoplankton diversity and zooplankton diversity leads to an increase in comprehensive results in autumn than in summer (Fig. 4, Fig. 5). By analyzing the radar map of S1, we can find that the increase in negative indicators of macrobenthos diversity, and the reduction of zooplankton diversity lead to an increase in comprehensive results in autumn than in summer (Fig. 4, Fig. 5).

Fig. 4
figure 4

Score distribution of different indexes of the river section in summer. The indicators represented by each letter have the following meanings (a) Vegetation coverage; (b) River connectivity;(c) River width; (d) Phytoplankton diversity; (e) Zooplankton diversity; (f) Macrobenthos diversity; (g) Proportion of building area; (h) River winding rate; (i) Proportion of hardened area;(j) Water quality

Full size image
Fig. 5
figure 5

Score distribution of different indexes of the river section in autumn, the index represented by each letter is the same as above

Full size image

4 Discussion

In river health evaluation, the determination of weight is one of the important links. Once the weight deviates, it will directly lead to incorrect evaluation results and cannot match the actual situation. In the past river health evaluation, single-factor evaluation was performed on different indicators or a simple summation of multiple factors was performed, resulting in inaccurate results (Kleynhans 1996; Hao 2014). As a reliable weight determination method, principal component analysis is widely used in various evaluations (Yang et al. 2015; Shen et al. 2020; Zou et al. 2021), and Wang’s study found that the results calculated by the principal component analysis method are more scientific and more reasonable than the entropy weight method (Wang et al. 2017). Therefore, in this study, principal component analysis was used to calculate the weights. Through this method, ten indicators related to the comprehensive health evaluation of Chongming River were screened out (Table 1), and their respective weights were calculated. These weight of indicators are consistent with the research results of Su (2019) and Wu (2006). The assessment system can be successfully applied to the Chongming Island river health assessment, indicating that the assessment system is applicable to the river system in the Yangtze Estuary. In addition, a total of ten secondary indicators were selected in the evaluation system in this study. Compared with 17 indicators in Wu's study and 17 indicators in Xu's study, it is simpler and more convenient in practical application (Wu et al. 2006; Xu et al. 2022).

The evaluation model of this study not only screened out the indicators such as water quality, aquatic organisms, river connectivity, and river width that have an important impact on the overall health of the river but also comprehensively considered the impact of vegetation coverage and human impact. Because Chongming Island not only has relatively concentrated urbanized townships but also many farmland and countryside which is also a characteristic of the river ecosystem in the Yangtze Estuary, the general urban river health assessment model or mountain river model is not applicable here. So adding indicators such as human activities can better reflect the real situation and are more conducive to the later ecological management of rivers.

In addition, the weights of benthic biodiversity indicators and the proportion of building area in this model are negative, which is rare in other models. First of all, the construction area along the coast is negatively correlated with the health of the river, which is easy to understand. The larger the proportion of building area, the denser the population in this area, and the greater the negative impact of human activities such as water abstraction and sewage discharge on the river. According to the research findings of Ge et al. (2022) the degree of land use is the main factor affecting ecosystem health and the impact of the proportion of construction land on ecosystem health increases over time. The layout used in urban land use planning has a significant impact on ecosystem health. While the negative correlation between benthic indicators and river health is mainly because the benthic organisms in this study area are mainly fouling-tolerant benthic organisms. The more fouling-resistant benthic organisms, the worse the health of the river (Deng et al. 2005; Wang et al. 2012a, 2012b, Hooper et al. 2013). Negative weights rarely appear in the research of other scholars, because they have especially standardized these negative indicators in advance (Wang et al. 2017). However, this will reduce the differences between sites and the results are not so intuitive. In order to ensure accuracy and scientific research, this study does not make special treatment for these negative indicators. To enhance the ecological health of rivers, it is essential to develop a rational riparian land use plan that regulates coastal construction and minimizes disturbances. Additionally, strict control measures should be enforced to prevent sewage discharge and domestic waste dumping into the river. Regular dredging of the river should also be conducted to mitigate sludge deposition.

In addition, the weights of the comprehensive water quality index, river connectivity, river width, comprehensive index of zooplankton diversity, the composite index of macrobenthos diversity, and comprehensive index of phytoplankton diversity are relatively high, which indicates that these indexes play an important role in river health assessment of the Yangtze Estuary. Therefore, more attention should be paid to these indexes.

Analysis of different research sites shows that the integrated assessment results of rural rivers S3 and S2 are the best (Fig. 3), which contradicts Jiang's research—the conclusion that the eutrophication of the northern water body of Chongming Island is more serious. The reason is that Jiang's research only takes the single dimension of nutrients as the starting point of the research (Jiang et al. 2019). In the comprehensive assessment of river health, a single indicator cannot be used. For example, if only considering the impact of human activities, the southern part of Chongming Island's result will be the worst with urbanization, the existence of more garbage floating objects, the high proportion of building areas, and the evaluation of a high degree of river bank solidification (Xu et al. 2005; Wang et al. 2021). However, this is not the case, river health assessment needs to consider multi-dimensional indicators (Peng et al. 2014), for example, S2 has higher vegetation coverage, river curvature, river width, and diversity of zooplankton and phytoplankton, so the comprehensive S2 evaluation result will be high, and the comparison is in line with the actual situation (Fig. 4, Fig. 5). When studying the seasonal differences at these two sites, it was found that the water quality assessment score of S3 in autumn was lower (Fig. 4, Fig. 5), which may be due to the influence of autumn agricultural harvesting activities. After the autumn harvest, there is a large amount of straw in the farmland, and returning the straw to the soil will increase the content of nitrate nitrogen in the soil. This will be washed into the river by rainwater, thereby affecting the water quality (Zhao et al. 2010). Although S2 also belongs to rural rivers, the S2 river is not directly connected to farmland but is separated by forest, vegetation, etc. The low score of the benthic biota index in S2 indicates that the sediment pollution in this river is more severe. In response to this situation, regular cleaning and dredging of the river should be carried out. Similarly, the comprehensive score of S6, which is also a rural river, is the worst. This is because the river is narrow, with poor connectivity and sediment accumulation. In addition, it is connected to the HuaBo Cultural Park, and the construction of HuaBo Cultural Park and a large number of tourists will have a negative impact on the health of the river. Therefore, it can be seen that when carrying out rural river management, it is necessary to retain buffer zones such as trees and vegetation on both sides of the river, improve connectivity, appropriately widen narrow rivers, regularly clean and dredge rivers, and reduce human activities that interfere with river health. Compared to rural rivers, urban rivers (S1, S4, and S5) show a generally average performance in a comprehensive evaluation. The high proportion of building area on both sides of the river and the low coverage of vegetation are the main factors limiting the evaluation results. Therefore, it is recommended to improve the health of urban rivers by rational planning of urban riverbank buildings and increasing vegetation coverage.

As for the seasonal differences of each site, the diversity of zooplankton, phytoplankton, and macrobenthos. For example, the autumn assessment results of S1 have improved compared with summer, mainly due to the increase in the diversity of zooplankton and phytoplankton, The assessment result of S2 in autumn is lower than that in summer, mainly because the diversity of macrobenthos is increased, which is a negative indicator, so the results of S2 in autumn are lower. The difference between the stations decreased in autumn, and the main reason may be related to the regulation system of "absorbing water in spring and draining in autumn" in the river channel of Chongming Island (Zhang et al. 2013a, b; Pang et al. 2016) because the rivers of Chongming Island are connected to each other, the large-scale drainage and water exchange in autumn reduces the difference between the stations. The evaluation results of S6 were the worst, mainly because the diversity of zooplankton, phytoplankton, and benthic organisms were all low, which shows that the diversity of zooplankton, phytoplankton, and benthic organisms is particularly important in river health assessment. When monitoring and ecological restoration should focus on these indicators.

5 Conclusions

This study constructed an ecological health assessment method for the Yangtze Estuary and applied it to the Chongming Island river network. The main conclusions are as follows.a) According to the characteristics of the rivers in the Yangtze River estuary, an ecological health assessment system suitable for the rivers in the Yangtze River estuary was constructed with five dimensions including water quality, river landscape, aquatic organisms, river hydrology, and human interference, and ten indexes including comprehensive water quality index, vegetation coverage, the proportion of building area along the coast, comprehensive index of zooplankton diversity, the composite index of macrobenthos diversity, comprehensive index of phytoplankton diversity, river connectivity, river winding rate, river width, the proportion of hardened area along the coast.b) The ecological health assessment system is applied to quantify the river health on Chongming Island. The evaluation results showed that the health of rural rivers in the northwest and east of Chongming Island (S2, S3) was the best, reaching an excellent level, while the central rural river (S6) was the worst, ranging somewhat inferior level. The assessment of urban rivers reached the good or ordinary level. Overall, the summer river health assessment results were better than those in autumn on Chongming Island, with slightly greater differences between locations in summer than in autumn. The seasonal differences at different locations were mainly due to changes in the diversity of plankton, phytoplankton, and benthic fauna indicators. These indicators should be given special attention in monitoring and ecological restoration.c) Based on the results of this study, it is recommended to adopt different measures for managing and restoring urban and rural rivers, taking into account their respective characteristics. In urban rivers, it is required to improve their health by rationalizing the planning of buildings along the river and increasing vegetation coverage. For rural rivers, it is better to preserve buffer zones such as forests and vegetation along the riverbanks, improve the connectivity of the river, widen narrow river sections as appropriate, and carry out regular dredging to reduce human interference on the health of the river.