The TGR is a crucial project for the developing and utilizing of water resource in the Yangtze River. From a geographical perspective, TGR is suited upstream of Yangtze River from Sandouping in Hubei province to Jiangjin County in Chongqing municipality (Xiang et al., 2021). The climate of the TGR area is the subtropical monsoon climate. The average annual air temperature of this area is 17.4°C, and the mean annual precipitation is 1204 mm (Cui et al., 2022). And the mean annual runoff of the TGR is 4.0 × 10¹¹ m³ (Yan et al., 2021).

Since the completion of the final impounding stage in October 2010, the TGR reached the design water level of 175 m, creating a massive reservoir spanning approximately 660 km in the mainstream of the Yangtze River, with a flood control capacity of 2.215 × 10¹⁰ m³ (Cui et al., 2022). Typically, the lowest water levels (around 145 m) occur in late May for flood control, while the highest water levels (around 175 m above sea level) are usually observed in October (Ye et al., 2022).

Because of the increased water level, downstream areas of the tributaries to the main channel of TGR in this region were flooded as the bay areas. Consequently, the watersheds for the flood tributaries formed by the impoundment of the TGR were divided into the Three Gorges Reservoir Region (Figure 1), which is the priority area for watershed management of the TGR (Zhang and Lou, 2011).

In order to conduct a thorough evaluation for the ecological water quality of TGR, a total of 173 sampling sites were selected to cover 22 tributary bays and 17 transects in the mainstream of TGR, providing a full coverage of the major tributaries and mainstream of former Yangtze River of the TGR (Figure 1). Specifically, in the reservoir area in Hubei province, there are 7 tributaries and 6 transects in the mainstream of TGR (Figure 1). While the reservoir area in Chongqing province contains 15 tributaries and 11 transects in the mainstream of TGR. The field sampling was carried out on 14–25 April 2015. Due to the loss of a few phytoplankton and water quality samples, a total of 164 sampling sites were ultimately used in our study.

All field sampling and lab analyses strictly followed the standard protocols of the Chinese Ecosystem Research Network (Cai, 2007). Prior to laboratory analysis, the samples for water quality and phytoplankton were collected using a 5-L Van Dorn sampler at a depth of 0.5 m underwater. Water chemistry samples were obtained using polyethylene bottles that had been pre-cleaned and stored in dark and cool environment before lab analyses. Total nitrogen (TN) and total phosphorus (TP) used in this study were analyzed using a segmented flow analyzer (Skalar San⁺⁺, Netherlands). A 600 ml water sample was filtered through Whatman filter (−1.2 μm, GF/C) to measure the concentration of Chlorophyll-a (Chl-a), using by the trichromatic method from APHA (1998).

For phytoplankton enumeration and identification, 1.2-L water sample was collected and preserved with neutral Lugol’s solution immediately in each site after sampling. The phytoplankton sample for each site was concentrated by the sedimentation method and then preserved with 4% formalin. Taxonomic identification of phytoplankton samples was carried out using the Fuchs-Rosenthal slide with an Olympus CX21 microscope at 10 × 40 magnification, with references to the works of Hu and Wei (2006) and John et al. (2002).

The Q index, developed by Padisak et al. (2006), was used to evaluate the ecological water quality in the present study. The Q index is rooted on the functional groups of the phytoplankton community and is calculated as follows: Q = ∑ j = 1 n p j F

Here, p _j is the proportion of the biomass of the j-th functional group to the entire phytoplankton biomass for the sample; and F is the factor number for the functional group in a given water body. The resulting Q index values can be ranged from 0 to 5 and can be divided into 5 grades: bad (0<Q≤1), poor (1<Q≤2), moderate (2<Q≤3), good (3<Q≤4), and excellent (4<Q≤5). The phytoplankton functional group-based Q index is a widely accepted index for the evaluation of reservoir ecological water quality worldwide (Becker et al., 2010; Wang et al., 2011; Shen et al., 2014; Korneva and Solovyeva, 2021; Çelekli and Lekesiz, 2021; Wu et al., 2023).

Here, the determinations of functional groups and the F factor mainly referred to the previous research carried out in the TGR by Wang et al. (2011). The details of phytoplankton species and the corresponding functional groups and the values of F factor could be found in Supplementary Table S1 in the supporting material. By utilizing the Q index, we are able to perform a comprehensive assessment of the ecological water quality of TGR and examine potential effects of changes in land covers on the ecological water quality of the tributary bays of TGR.

The land cover data were obtained from the Resource and Environment Science and Data Center (RESDC) (https://www.resdc.cn). Specifically, the data of land covers were interpreted from the Landsat images by specialists in RESDC. These images were captured around the same time as our field sampling. The land cover data were aggregated into 5 categories including forest, farmland, grassland, urban area, and others, to align with our research aims.

To determine the specific land cover information for each tributary bay of TGR, we extracted the watershed outline for each tributary of TGR using QSWAT (Dile et al., 2022). Then, the area of each type of land covers in the watershed was counted to investigate potential effects of changes in watershed land covers on the ecological water quality.

Our study revealed that forest and farmland were the predominant land cover types in all watersheds of the tributaries located within the Three Gorges Reservoir Region (Figure 2). Specifically, the forest area ratios in all tributary watershed of TGR ranged from 23.84% to 86.72%, while the farmland ratios ranged from 7.36% to 64.25%. Notably, we observed a higher proportion of forest area in the watersheds of tributaries near the Three Gorges Dam, which decreased as the distance from the Three Gorges Dam increased (Figure 2). Conversely, the ratios of farmland area exhibited an opposite pattern, with lower percentages in the tributary watersheds near the dam and higher ratios in the tributary watersheds away from the dam. The coverage of grass was lower than that of forest and farmland, with values ranging from 3.16% to 33.31%. Furthermore, the proportion of urban area in the tributary watersheds of TGR was extremely low, ranging from 0.02% to 1.81%.

The statistical summaries for the main chemical parameters were presented in Table 1. The average concentrations of TN and TP within the backwater regions of TGR’s tributaries were 1.689 and 0.094 mg/L, both of which are marginally lower than the corresponding averaged values observed in the mainstream of TGR. The averaged concentration of Chl-a in the tributary area was 12.47 μg/L, which is significantly higher than the mean value (0.78 μg/L) in the mainstream of TGR. Moreover, for the tributaries, we found the variations of water quality parameters in tributary bays are higher than the values in the main channel (Table 1).

The Q index-based assessment showed that the ecological water quality of most sites in the main channel of TGR are good, while most sites in the backwater regions of TGR’s tributaries exhibit a status with moderate or bad ecological water quality (Figure 3). Specifically, the ecological water quality assessment found that 84% of the sampling sites in the main channel of TGR exhibited good or high ecological water quality, whereas 12% and 4% of the sites falling into moderate and poor status, respectively (Figure 3A). In contrast, only 21% of the sites in the tributary bays were in good status, with 47%, 26%, and 6% of sites exhibiting moderate, poor, and bad ecological water quality, respectively (Figure 3B). Notably, we found that the tributary bays close to the Three Gorges Dam exhibit a better ecological water quality than the bays in the upstream (Figure 3C). Overall, the ecological water quality of tributary bays degraded with the distance from the Three Gorges Dam.

We found that the percentages of farmland and urban areas had significant negative correlations with the Q index-based ecological water quality of the tributary bays (Figure 4A, B). This finding supports our hypothesis that an increase in the urban and farmland area will degrade the ecological water quality in the backwater regions of TGR’s tributaries (H1). On the other hand, our study did not fully support the hypothesis based on the natural land cover (H2), as the percentage of grass area had a nonsignificant positive correlation with the ecological water quality (Figure 4C). However, the percentage of forest area had a marginally significant (p = 0.058) positive correlation with the ecological water quality (Figure 4D), suggesting increasing forest area may improve the ecological water quality in the backwater regions of TGR’s tributaries.

Here, we evaluated the ecological water quality of TGR and examined the effects of changes in watershed land covers on ecological water quality in the backwater regions of TGR’s tributaries. With the aid of the Q index, we have gained a comprehensive understanding of the ecological water quality of the entire TGR. Our investigation indicated that the ecological water quality of most regions within the mainstream (previously known as the Yangtze River) of TGR is good, while the water quality problem primarily manifests in the backwater regions of tributary bays. This finding is in accordance with the previous study based on Carlson’s trophic state index (Xu et al., 2010; Tan et al., 2014) as well as the reported algal blooms in the Three Gorges Reservoir Region (Ouyang et al., 2021; Xu et al., 2021; Ye et al., 2022). From the perspective of ecological water quality, our study highlights the significance of managing the reservoir water quality in the backwater regions of TGR’s tributaries.

Water quality is a crucial aspect of ensuring the health and safety of our environment and public health (Su et al., 2022). The accurate assessment of water quality of a given water body is an essential concern for the successful management and protection of water resources (European Environment Agency, 2000; Cosgrove and Loucks, 2015; Bhateria and Jain, 2016; Zhang et al., 2022). Compared to the traditional chemical water quality indices such as TN and TP, our study found that Q index-based ecological water quality can better indicate the land cover changes in the watersheds of TGR’s tributaries.

In our study, we observed a robust association among ecological water quality in the backwater regions of TGR’s tributaries and the land covers within the corresponding watershed. However, we did not observe a reasonable correlation between the concentrations of TN or TP in the tributary bays of TGR and the watershed land cover (Figure 5). Specifically, the Spearman correlation analyses indicated that the concentrations of TP had nonsignificant relationships with watershed land cover (Figure 5). While, TN was significantly negatively correlated with the urban and farmland area, and had a significant positive correlation with the forest area. Nevertheless, the relationship between TN and watershed land cover is a spurious correlation as it is inconsistent with the common relationship observed in the TGR region, where watershed with high farmland and urban development tends to have more exports of TN and TP (Ye et al., 2009; Zhang et al., 2019; Huang et al., 2022).

The lack of a reasonable correlation between the concentrations of TN and TP and watershed land cover in the tributary bays of TGR (Figure 5) indicates that these chemical water quality indices may be susceptible to be affected by other factors, such as water level fluctuation in the TGR. Recent studies from the field observation (Xiang et al., 2021) and the hydrodynamic model (Luo et al., 2022) have supported this point, showing that water quality in the backwater regions of TGR’s tributaries is mainly dominated by the backwater from the mainstream of the TGR. In contrast to traditional water quality indices, the Q index developed from phytoplankton functional composition (Padisak et al., 2006), can provide a more comprehensive and integrated assessment of ecological water quality in aquatic ecosystems because the biological assessment can measure the long-term effects of environmental changes on ecosystems (Prasse et al., 2015; Gecheva et al., 2023). Our study underscores the utility of the Q index in assessing ecological water quality in reservoir ecosystems.

Our study has several implications for the management of water quality in reservoirs. One essential implication for reservoir management is that our study revealed that appropriate watershed management can improve the ecological water quality in the backwater regions of TGR’s tributaries. Specifically, the control the development of farmland and urban and increase forest area can improve the ecological water quality in the backwater regions of TGR’s tributaries (Figure 4). The Three Gorges Reservoir Region is a key area for the socio-economic development of the Yangtze River Economic Belt. However, this region is still undergoing high-speed development. According to the public bulletins from the government (China National Environmental Monitoring Centre, 1998-2016), the total yearly urban sewage discharge in the TGR Region ranged from 4.04 × 10⁹ t to 12.12 × 10⁹ t from the year 2003–2016 (Figure 6). Meanwhile, this region is also the key area for ecological and environmental protection for the backwater regions of TGR’s tributaries. In light of these facts, it is essential to implement appropriate watershed management measures to improve the ecological water quality in the backwater regions of TGR’s tributaries. Our study found that watershed management is an effective approach to enhance the ecological water quality in the backwater regions of TGR’s tributaries, but given the flat slope (Figure 4), this is likely to be a long-term process. Therefore, our study recommends exploring both short and middle-term approaches to improve the ecological water quality of TGR, in addition to the long-term watershed management measures.

Another implication is that the reservoir management practices should take into account the potential environmental and ecological impacts associated with fluctuations in the water levels of the reservoir. Our study showed that the land cover in the tributary watershed can significantly affect the ecological water quality in the tributary bay (Figure 4); however, the explanatory power of land cover changes to the ecological water quality appears to be limited, suggesting backwater from the mainstream of TGR would affect the ecological water quality in the tributary bays. This point was also supported by Xiang et al. (2021), which reported that the water quality (e.g., TN, TP) in the tributaries of TGR was mainly affected by the backwater movement from the mainstream of TGR, which was driven by the water level fluctuations. Meanwhile, research also showed that the water level fluctuation in reservoirs will drive the development of phytoplankton blooms in the backwater region in the Xaingxi Bay of TGR (Ye et al., 2022), which might pose a series of environmental and ecological problems (Anderson et al., 2012; Amorim and Moura, 2021). Given all the above concerns, our study suggests that effective reservoir management should also consider the impacts of water level fluctuations.

Finally, our study highlights the practicality of biological assessment in effective reservoir management. Compared to the classical water chemical indexes (e.g., TN, TP), we found that the phytoplankton functional group-based Q index can better indicate the watershed land cover changes. This result presents the advantages of biological assessment, which is sensitive to environmental changes and measures long-term effects (Prasse et al., 2015; Gecheva et al., 2023). In light of the above facts, the biotic index has been suggested as a fundamental measurement in assessing the ecological water quality by the Water Framework Directive of Europe (European Environment Agency, 2000) and the US Environmental Protection Agency (Barbour et al., 1999). And our study indicates Q index is a useful biotic index in assessing reservoir ecological water quality and has important value in water resources management.