Development of a bacteria-based index of biotic integrity

Download PDF
5 downloads 0 Views 2MB Size Report
Comment
A bacteria-based index of biotic integ- rity was developed in Three Gorges Res- ervoir. • Samples were collected in the low water level, impoundment and ...
Science of the Total Environment 640–641 (2018) 255–263

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Development of a bacteria-based index of biotic integrity (Ba-IBI) for assessing ecological health of the Three Gorges Reservoir in different operation periods Yi Li a, Nan Yang a, Bao Qian b, Zhengjian Yang c, Defu Liu c, Lihua Niu a, Wenlong Zhang a,⁎ a b c

Key Laboratory of Integrated Regulation and Resource Development on Shallow Lakes, Ministry of Education, College of Environment, Hohai University, Nanjing 210098, PR China Hydrology Bureau of Changjiang Water Resources Commission, Wuhan, Hubei 430010, PR China Hubei Key Laboratory of Ecological Restoration of River-lakes and Algal Utilization, Hubei University of Technology, Wuhan 430068, Hubei Province, PR China

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• A bacteria-based index of biotic integrity was developed in Three Gorges Reservoir. • Samples were collected in the low water level, impoundment and sluicing periods. • Hydromorphology and anthropogenic disturbances were used to define reference sites. • Five core metrics discriminate between reference and impaired conditions. • The novel index assessed the spatial and temporal variation of the ecological status.

a r t i c l e

i n f o

Article history: Received 2 April 2018 Received in revised form 23 May 2018 Accepted 23 May 2018 Available online xxxx Editor: J Jay Gan Keywords: Bacterial communities Index of biotic integrity The Three Gorges Reservoir Reservoir operation Quantitative assessment

a b s t r a c t It is urgently needed to quantitatively assess ecological health of the Three Gorges Reservoir (TGR) when considering its special environmental conditions and temporal variations caused by reservoir operation. This study developed a bacteria-based index of biotic integrity (Ba-IBI) based on sediment samples collected along the TGR in low water level period, impoundment period and sluicing period, respectively. Reference conditions were defined using 8 ecological variables describing the hydromorphology and anthropogenic disturbances around the sites. Five core metrics, including % Acidobacteria, % Gemmatimonadetes, % Geobacter, Methanotroph and Phototroph, were selected after the screening processes. The developed index could clearly discriminate reference and impaired conditions and exhibited significant relationship with environmental parameters according to the redundancy (p b 0.01) and multivariable linear regression analysis (R2 = 0.76). By implementing Ba-IBI in the TGR, the ecological health of the sampling sites was defined as “Excellent” (25%), “Good” (50%) and “Fair” (25%) separately. The spatial variation of biotic integrity was closely associated with environmental and ecological changes, especially the increase of nutrient concentrations. This study revealed a significant tendency that the ecological health in the low water level and sluicing periods was better than that in the impoundment period, which could be attributed to the hydrodynamic changes due to water level fluctuation. This study provides a comprehensive understanding of ecological health of the TGR in different operation periods and the index offers a guideline for the reservoir regulation in the similar areas. © 2018 Elsevier B.V. All rights reserved.

⁎ Corresponding author at: College of Environment, Hohai University, Xikang Road #1, Nanjing 210098, PR China. E-mail address: [email protected] (W. Zhang).

https://doi.org/10.1016/j.scitotenv.2018.05.291 0048-9697/© 2018 Elsevier B.V. All rights reserved.

256

Y. Li et al. / Science of the Total Environment 640–641 (2018) 255–263

1. Introduction With the development of national economy and society, the aquatic environments and their biota were constantly affected by anthropogenic activities such as water abstraction, sewage discharge, over exploitation of land and water project construction (Marzin et al., 2012). Rivers have attracted flocks of scholars and researchers due to their functions of maintaining natural ecological balance and serving society development. The Yangtze River, the longest river in Asia and the third-longest river in the world, flows a total length of 6300 km and a drainage area of 1,800,000 km2 (Chen et al., 2001). Built in the middle reaches of the Yangtze River, the Three Gorges Dam (TGD) is currently the world's largest dam and results in the formation of a deep water reservoir, the Three Gorges Reservoir (TGR). The TGR has fully operated since 2010 with water release in flood season (May to September) and water storage in drought season (November to March of the following year), giving rise to the fluctuation of water level from 145 m to 175 m (Han et al., 2017). The annual operation of the reservoir has resulted in several ecological problems such as algal bloom in the tributaries, sedimentation, debris flow and a series of environment changes in the water level fluctuation zone (Wang et al., 2013). Wilcox et al. (2002) proposed that variation from water level fluctuation could be much greater than variation from anthropogenic stress, causing unpredictable responses. Consequently, obvious spatial and temporal variations of the aquatic ecosystem are presented within the TGR. As the key channel connecting the southwest and the east, the TGR and its annual operation not only provides drinking water for the Chongqing southwestern economic center and the Hubei province of China, but also impacts the water security of cities in the lower regions of the Yangtze River (Gao et al., 2017). Therefore, studies on the water quality monitoring and aquatic ecosystem assessment of the TGR have been increasing over the years (Li et al., 2013). Several studies reported water quality data from the Yangtze River mostly focusing on organic or inorganic parameters, the mechanism of pollutants, the pollution load and source (L. Huang et al., 2015; Ma et al., 2011). Numerous studies monitored water quality in the TGR using multiple analysis approaches such as multi-objective environmental reservoir operation methodology (Hu et al., 2016), risky grade index (RGI) and coverage area index (Zhou et al., 2017). Temporal variations of the nutrients, biochemical indexes and heavy metals during the stable operation of the TGR were also assessed (Gao et al., 2016). Although some indices based on physicochemical parameters have been chosen to evaluate the environmental condition of the TGR, there is still need to modify the more narrowly defined chemical criteria and to develop biological criteria based on ecological principles. Changes in environmental parameters, hydromorphology and ecosystem health have already been linked to the variation of the biotic indices based on the taxonomic and/or functional characteristics of biological aggregates (Villeneuve et al., 2018). Such environmental assessments were often undertaken by using the index of biotic integrity (IBI), which was first introduced by Karr (1981) and widely used to assess ecological health (Karr et al., 1986). Organisms such as macroinvertebrates (Klemm et al., 2003), fish (Mercado-Silva et al., 2002), algae (Zalack et al., 2010) and plankton (Wu et al., 2012) were commonly used as indicators to evaluate the health of aquatic environments. However, the construction of the TGD could alter the hydrologic regimes and connectivities, resulting in transformation of the Yangtze River from a riverine system into a lacustrine system. Due to the migration restrictions, spawning grounds destructions, habitats and food web changes, the behavior and functions of most macro-organisms were affected (Franssen, 2012). Thus, the traditional index of biotic integrity could not be used to assess ecological health of the TGR effectively and a new indicator is needed to be proposed. In aquatic systems, sediment bacteria are the foundation of biogeochemical cycles, participating in a great number of biochemical processes such as phosphorous cycling and nitrogen transformation. As

the most numerous and active organisms in the basal trophic level of stream food web, bacteria have short life cycles and are sensitive to environmental changes (Lau et al., 2015). Bacterial communities have been proved to be better indicator of ecological health compared with macro-organisms due to their sensitivity to the external influence. In our previous research, bacterial communities have been included in the process of biotic integrity assessment and the sensitivity and reliability of the index were validated (J. Li et al., 2017). Bacterial communities also exhibited temporal variations with seasonal or temporal environmental changes (Z. Li et al., 2017) and were believed to be sensitive to water level fluctuation caused by reservoir operation (Weise et al., 2016). Previous research about bacterial communities in the TGR mainly focused on their spatial distribution (Yan et al., 2015), yet failed to illustrate bacterial response to the operation of the TGR. Additionally, bacteria used in water quality monitoring of the TGR only involved a few forms such as fecal indicator bacteria (Wang et al., 2015). Thus, a quantitative assessment based on the bacterial communities was expected to illustrate ecological health of the TGR. Considering the above factors, the main objective of our study was to assess ecological health of the Three Gorges Reservoir in different operation periods by developing a bacteria-based index of biotic integrity (Ba-IBI). Water and sediment samples were collected from 12 sampling sites across the TGR during three sampling campaigns. Redundancy and multivariable linear regression analysis were carried out for clarifying the correlation between the Ba-IBI and environmental parameters. Temporal variation of biotic integrity associated with water level fluctuation caused by reservoir operation was also investigated. This study provides a better understanding on ecological health of the TGR and illustrates the potential effects of reservoir operation on bacteria-based biotic integrity. 2. Materials and methods 2.1. Study area and data collection The Three Gorges Reservoir (TGR), downstream from Yichang of Hubei Province up to Jiangjin of Chongqing municipality (105°50′111°40′ E, 28°31′-31°44′ N), is currently the largest artificial reservoir of the world. The TGR area covers approximately 55,000 km2 and includes 20 country-level administrative districts. Belonging to the subtropical climate, the average temperature of the area is 17.8 °C and the average annual rainfall is 1100 mm. The northern and eastern parts of the TGR are dominated by high mountains and the western parts are low plains (Teng et al., 2017). In this study, 12 sampling sites were selected from upstream to downstream of the TGR as shown in Fig. 1. The monthly average of hydrology data such as water level, flow velocity and precipitation, were obtained from the Hydrology bureau of Changjiang Water Resources Commission, as well as the websites of “The Yangtze River Hydrology” (http://www.cjh.com.cn) and “The National Hydrological and Rainfall Regime” (http://xxfb.hydroinfo.gov. cn). The meteorological factors within the study area, such as air temperature, wind speed, relative humidity, were collected from the National Meteorological Information Centre of China (NMIC) (http:// data.cma.cn). Soil type data as well as the Land Use and Land Cover (LULC) data were provided by the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (RESDC) (http:// www.resdc.cn). The administrative divisions and population information were obtained from the website of “Information Management Center for the Ecological and Environment Monitoring of the Three Gorges Project” (http://www.tgenviron.org). 2.2. Sampling and sample analysis We carried out 3 sampling campaigns in the low water level period (September 2016), impounding period (December 2016) and sluicing period (April 2017), respectively. During each sampling campaign, samples

Y. Li et al. / Science of the Total Environment 640–641 (2018) 255–263

257

Fig. 1. Locations of sampling sites along the Three Gorges Reservoir (TGR), including Zhutuo (ZT), inlet of the TGR, Tongguanyi (TGY), Cuntan (CT), Changshou (CS), Qingxichang (QXC), Wanxian (WX), Xiaojianghekou (XJH), Fengjie (FJ), Daninghekou (DNH), Guandukou (GDK), Xiangxihekou (XXH) and Miaohe (MH) in outlet of the TGR.

of surface water and sediment in the center of the Yangtze River were collected in triplicate. At each site, water samples were collected 50 cm below the surface and composited into a 1-L polyethylene bottle and then preserved at 4 °C for further analysis. Sediment samples were collected from 10-cm depth of the streambed using a hand-held coring device. All sediment samples were preserved in transit using a liquid nitrogen jar (YDS20-80F) and stored at −80 °C in the freezer until DNA extraction. Simultaneous with sampling, pH, water temperature (T, °C) and dissolved oxygen (DO, mg/L) were measured in situ by a HACH HQ30d portable meter (HACH Company, Loveland, CO, USA). At each site, water samples were also collected for further laboratory analysis including suspended solids (SS, mg/L), ammonium nitrogen (NH3-N, mg/L), total nitrogen (TN, mg/L), total phosphorus (TP, mg/L), biochemical dissolved oxygen demand (BOD5, mg/L), chemical oxygen demand (COD, mg/L) and Escherichia coli (E. coli, A/L). All these parameters were measured according to the Environmental Quality Standards for Surface Water in China. DNA was extracted from each sediment sample using the PowerSoil DNA Isolation Kit (Mo Bio Laboratories Inc., Carlsbad, CA, United States) following the manufacturer's instructions, and then used for the standard PCR assays of bacterial abundance. For high-throughput sequencing analysis of bacterial communities, all the DNA samples were sequenced by Illumina Miseq Sequencing at Shanghai Majorbio Biopharm Technology Co., Ltd., China. The operational taxonomic units (OTUs) were defined by 97% sequence similarity cut-off. The relative abundance comparison and alpha-diversity analysis of the OTUs were conducted using packages in R (v.3.12, http://www.r-project.org). 2.3. Development of Ba-IBI 2.3.1. Sites classification For developing the Ba-IBI, study sites need to be classified into reference sites and impaired sites. The reference sites of the TGR were distinguished based on the concept of least-disturbed conditions with reference of the criteria in the study of Matono et al. (2012). As shown in Table 1, land use, river connectivity, soil erosion, sediment load, riparian zone, biodiversity, urbanization and pollution were included in the evaluation. Each variable was scored from 1 (maximum disturbed) to 5 (minimum disturbed). An ecological index, calculated by the sum of the scores, was used to indicate the ecological conditions as well as the external disturbances of each site.

2.3.2. Metrics selection and screening 36 samples collected from 12 sampling sites were used to develop the Ba-IBI. We compiled a large pool of candidate metrics, which belonged to diversity, composition, tolerance and function metrics. Diversity metrics, indicating how many different types were included in a community, included multiple alpha diversity indices at phylum and genus levels, respectively. The calculation was conducted using the Paleontological Statistics (PAST) software (Hammer and Harper, 2009) and vegan package in R (v.3.12, http://www.r-project.org) as described by Oksanen et al. (2018) in the Ecological Diversity Indices part. For the composition metrics, not only the dominant phyla, classes, orders, families and genera, but also the ratios of some representative bacterial groups were included. For the candidate metrics involved in environment tolerance, pollution-tolerance bacterial taxa were selected according to the process detailed in the study of Y. Li et al. (2017). Function metrics were automatically generated from the taxonomic input data using METAGENassist (http://www.metagenassist.ca) (Arndt et al., 2012). Additionally, in order to reduce the influence of the peak value and to ensure the data approach normality, all the proportional values were arcsine and square root transformed according to Jia et al. (2013). Since not all the metrics could effectively indicate ecological health of the TGR, metrics used for the development of Ba-IBI were chosen using the following step-wise screening procedure (Stoddard et al., 2008). (1) Range test: metrics with insufficient ranges or poor discrimination between reference and impaired sites were discarded. (2) Responsiveness test: metrics were selected using non-parametric MannWhitney U tests (p b 0.05) to examine the responsiveness of the remaining candidate metrics discriminating the minimally and the most impaired sites. (3) Discrimination power test: metrics were measured using box and whisker plots tests (Barbour et al., 1996). The degree of inter-quartile (IQ) overlap in the box plots was used to define the separation power of each metric between reference and impaired conditions. Only metrics with IQ N 2 were considered to have high discrimination power and were maintained to the next step. (4) Redundancy test: to avoid redundancy, robust statistical tool was used to determine the most efficient metrics among redundant metrics (Casatti et al., 2009). Redundant metrics within the reference sites were identified using Spearman correlation analysis. Pairs of the metrics with strong correlations (r N 0.75, p b 0.01) were considered to be redundant and the redundant metrics were finally selected based on their

258

Y. Li et al. / Science of the Total Environment 640–641 (2018) 255–263

Table 1 Description and scoring criteria of 8 variables used to classify reference conditions. Variables

Description

Criteria

Score

Land use

Agriculture/forest land coverage

River connectivity

River connectivity and surface runoff

Soil erosion

Bank stability and erodibility

Sediment load

Sediment deposited on the riverbed

Riparian zone

Revetment types and distribution

Biodiversity

Community composition and distribution

Natural forest; N70% unutilized land Bush or/and grass land Gardens and pastures Horticultural crops and grazing Irrigated crops; N70% utilized land No barriers of water flow Occasional connectivity cut off; Existence of passages for most species Frequent connectivity cut off or runoff change; Existence of passages for certain species Long-term connectivity cut off; Existence of passages only in short periods Permanent river connectivity cut off and artificial barrier Little soil erosion; Stable river bank Occasional soil erosion; Little destroyed river bank Moderate soil erosion; b30% destroyed river bank Frequent soil erosion; 30–60% destroyed river bank Intensive soil erosion; N60% destroyed river bank b5% of coarse particles of the stream bed are covered with fine sediments (sand, silt, clay) 5–25% of coarse particles of the stream bed are covered with fine sediments (sand, silt, clay) 25–50% of coarse particles of the stream bed are covered with fine sediments (sand, silt, clay) 50–75% of coarse particles of the stream bed are covered with fine sediments (sand, silt, clay) N75% of coarse particles of the stream bed are covered with fine sediments (sand, silt, clay) All the riverside are covered by reinforced natural revetment Riverside are covered by ecological revetment made by natural materials Riverside are covered by revetment made by compound materials Riverside are covered by flinty revetment made by concrete or rebar Most of the revetments are destroyed Great variety of plants, birds and fishes; Existence of all the natural habitats Existence of most species; Existence of natural habitats including aquatic vegetation, plant litter, in-stream wood and rock Existence of many types of species; Existence of natural habitats including riverside vegetation, wood and rock Existence of certain types of species; Existence of artificial habitats Only few types of species; Only one or two kinds of habitats Few settlements along the rivers Straggling village Distant suburb Large-scale town Economically developed cities No industrial, agricultural or domestic pollutions Few non-point source pollutions mainly from human activities along the rivers Widespread pollution mainly caused by agriculture cultivation Frequent drainage of industrial waste water and domestic sewage Severe point source pollution and non-point source pollution

5 4 3 2 1 5 4 3 2 1 5 4 3 2 1 5 4 3 2 1 5 4 3 2 1 5 4

Urbanization

Urban area and development degree

Pollution

External contaminants disturbance

applicability to represent ecological conditions of the sampling sites. Mann-Whitney U tests and Spearman correlation analysis were all conducted using IBM SPSS Statistics 20.0. 2.3.3. Index development This study used the 5th and 95th percentiles scaling system to calculate the final score of each metric (Abdelkefi et al., 2013). For metrics that increased with impairment, the lower threshold was calculated based on the 5th percentile of metric values and the final Ba-IBI value = (Maximum-Site value) / (Maximum-Lower threshold). For metrics that decreased with impairment, the upper threshold was calculated based on the 95th percentile of metric values and the final Ba-IBI value = Site value / Upper threshold. Then the final Ba-IBI scores were classified into three scales based on the 75th and 25th percentile to assess ecological health of the sampling sites. 2.4. Testing of the Ba-IBI The performances of the Ba-IBI were evaluated using responsiveness test and relevancy test. The responsiveness of the Ba-IBI was tested by calculating the discrimination between the reference and impaired conditions using box and whisker plots tests (Barbour et al., 1996). The larger the difference between the values, the more responsive the BaIBI was. The correlation between Ba-IBI and the ecological index was also analyzed to demonstrate the applicability of the index in the study area. The relevancy of the Ba-IBI was evaluated by analyzing the correlation between the Ba-IBI and the environmental parameters.

3 2 1 5 4 3 2 1 5 4 3 2 1

First, redundancy analysis (RDA), a multivariate ordination technique for direct gradient analysis, was chosen to test the bacterial relationship with environment and estimate the amount of variation in the species data that is explained by measured environmental variables. Meanwhile, Ba-IBI and its core metrics were also set as variables to investigate their relationship with physicochemical parameters and bacterial assemblages. RDA was conducted using Canoco 5.0 (http://www. canoco5.com). Second, a multivariable linear regression model was used to analyze the correlation between the Ba-IBI and the physicochemical characteristics in R (v.3.12, http://www.r-project.org). The multivariate linear regression model was as follows: Ba‐IBI ¼ β0 þ β1 x1 þ β2 x2 þ … þ β11 x11 þ ε where βi is the slop coefficient of explanatory variable xi, xi is the environmental variable and ε is the model error or residual. This model was also used to predict the scores of Ba-IBI. The predicted and observed Ba-IBI were compared and validated. 3. Results 3.1. Environmental and bacterial data The measured water level, velocity, DO, pH, SS, NH3-N, TN, TP, BOD5, COD and E. coli at sampling sites in the three operation periods are listed in Table 2. The velocity, pH, SS, TP, BOD5 and COD at the impaired sites exhibited higher values than the reference sites. In particular, the water

Y. Li et al. / Science of the Total Environment 640–641 (2018) 255–263

level showed obvious variability at the impaired sites compared with the reference sites. Comparing the temporal variation during the operation periods, the highest values of water level were observed in the impoundment period, accompanied by the lowest flow velocity and suspended solids. The nutrient factors varied moderately at different sampling sites. Samples collected in the sluicing period were generally characterized by higher TN, TP and BOD5. Correlation analysis was conducted on the eleven variables and the results are shown in Table S1. After the analysis, water level, velocity, DO, TN and E. coli were found to be relatively independent metrics characterizing the environmental conditions of the TGR. The other variables were also considered in the further analysis. In this study, Illumina Miseq Sequencing of 16S rRNA genes generated an average of 49,052 sequence reads from the 36 DNA samples. With the library coverage of each sample ≥ 90.23%, the reads were grouped into 67,875 OTUs with an average of 3146 OTUs per sample. 61 phyla, 186 classes, 328 orders, 625 families, 1320 genera and 2415 species were assigned from the reads. 3.2. Development of the Ba-IBI The ecological indexes of the sampling sites were calculated based on the 8 variables shown in Table 1. The scores varied from 18 to 36 and sites with scores N30 were selected as the reference sites, i.e. MH, GDK and FJ. The distribution of the reference sites and impaired sites are shown in Fig. 1. A total of 136 candidate metrics (Table S2), assigned to four categories were examined for possible inclusion in the Ba-IBI. Diversity metrics included the alpha-diversity indexes such as Simpson, Shannon, Equitability and Fisher alpha indices. Composition metrics were composed of the top 10 phyla, 15 classes, 20 orders, 19 families and 18 genera, as well as the percentage of Firmicutes, Chloroflexi and Acidobacteria (% FCA), the ratio of Bacteroidetes, Firmicutes and Gammaproteobacteria to Alphaproteobacteria (BFG/A), the ratio of Bacteroidetes, Firmicutes and Gammaproteobacteria to Betaproteobacteria (BGN/B) and the ratio of Bacilli, Bacteroidetes and Clostridia to Alphaproteobacteria (BBC/A). Tolerance metrics contained the response of genera taxa to the five independent physicochemical parameters described above, including water level, velocity, DO, TN and E. coli. Function metrics included bacterial phenotypes (e.g. Gram-positive and Gram-negative), metabolism (e.g. Ammonia Oxidizer, Nitrite reducer and Sulfide oxidizer) and functional pathways (e.g. Aerobic, Heterotroph and Phototroph). The box plots of the metrics in the discrimination power test are shown in Fig. S1 and the results of the redundancy test are shown in Table S3. After the screening tests, five core metrics were finally selected for the development of Ba-IBI (Table 3). Three composition metrics, i.e. % Acidobacteria (M24), % Gemmatimonadetes (M29) and % Geobacter (M105), and two function metrics, i.e. Methanotroph (M132) and Phototroph (M135) showed abilities to indicate ecological health of

259

the TGR. Methanotroph decreased with the disturbance while the other four metrics increased. The final Ba-IBI scores were the sum of the core metric values calculated using formula listed in Table 3, ranging from 2.170 to 4.153 with an average value of 3.265. 3.3. Application of the Ba-IBI The variation of the Ba-IBI along the TGR at different sampling sites in the three operation periods are shown in Fig. 2. The variation trend in the low water level period was similar to that in the sluicing period, while the impoundment period showed significant difference. The reference sites, FJ, GDK and MH, represented the highest Ba-IBI during each period. The final Ba-IBI scores were classified into three scales based on the 75th and 25th percentile of the Ba-IBI scores (4.061 and 2.605, respectively): “Excellent” (4.061–4.753), “Good” (2.605–4.067) and “Fair” (2.170–2.605). The classifications of the sampling sites are shown in Fig. 3. 50% of the sampling sites were in “Good” conditions, 25% in “Excellent” conditions and 25% in “Fair” conditions. Most of the “Excellent” and “Good” categories of samples were collected in the low water level period and sluicing period, while most of the “Fair” samples were collected in the impoundment period. Fig. 3(a) and (c) showed that almost all the sampling sites of the TGR in the low water level and sluicing periods were in at least “Good” conditions except TGY and CT. However, in the impoundment period, 41.67% of the sites were in “Fair” conditions, especially the middle and upper reaches of the TGR. 3.4. Testing of the Ba-IBI The boxplot in the responsiveness test was plotted in Fig. S2, showing the power of calculated Ba-IBI to discriminate impaired and reference conditions. Furthermore, the Ba-IBI exhibited significant correlation (R2 = 0.71) with the ecological index calculated in the selection of reference sites as shown in Fig. S3. The biplot of the RDA results indicated that physicochemical parameters made significant contributions to the variation of bacterial communities (p b 0.01) (Fig. 4). The total variation was 2166.205 and the explanatory variables accounted for 60.4%. The first two axes explained up to 34.24% of RDA 1 and 8.35% of RDA 2 of the total variation in bacterial community structure. There was a positive association of RDA axis 1 with TP and NH3-N, and a negative association of RDA axis 1 with E. coli, BOD5, DO and water level. RDA axis 2 was positive associated with TN and negative associated with COD, velocity, pH and SS. The Ba-IBI and the core metrics showed obviously positive and negative relationship with the first two RDA axes. Almost all the metrics increased with more positive RDA axis 2 and more negative RDA axis 1. Using multivariate linear regression, the relationship between the Ba-IBI and the environmental factors were represented assuming a linear relationship. The coefficients of the multivariate linear regression model were estimated and statistical tests were performed to ensure

Table 2 Median and standard error of physicochemical parameters over the sampling periods. Physicochemical parameters

Water level (m) Velocity (m/s) Dissolved oxygen (mg/L) pH Suspended solids (mg/L) Ammonium nitrogen (mg/L) Total nitrogen (mg/L) Total phosphorus (mg/L) Biochemical dissolved oxygen demand (mg/L) Chemical oxygen demand (mg/L) Escherichia coli (A/L)

Low water level period

Impoundment period

Sluicing period

Reference

Impaired

Reference

Impaired

Reference

Impaired

160.20 ± 0.54 0.70 ± 0.10 7.48 ± 0.18 7.45 ± 0.12 50.00 ± 56.66 0.125 ± 0.015 2.10 ± 0.12 0.13 ± 0.05 0.90 ± 0.25 1.24 ± 0.38 13,998 ± 918

160.43 ± 77.53 1.00 ± 0.17 7.14 ± 0.14 7.63 ± 0.10 143.00 ± 39.94 0.136 ± 0.009 1.89 ± 0.14 0.15 ± 0.03 1.36 ± 0.34 1.84 ± 0.27 12,000 ± 2026

174.56 ± 0.36 0.50 ± 0.15 7.55 ± 0.15 7.44 ± 0.03 43.00 ± 53.43 0.130 ± 0.001 1.92 ± 0.29 0.11 ± 0.03 1.02 ± 0.30 1.30 ± 0.31 12,899 ± 1161

174.00 ± 79.55 0.80 ± 0.16 7.25 ± 0.13 7.76 ± 0.13 119.00 ± 31.09 0.132 ± 0.013 1.86 ± 0.16 0.14 ± 0.02 1.27 ± 0.33 1.89 ± 0.23 11,036 ± 2244

150.30 ± 0.89 0.70 ± 0.15 7.42 ± 0.22 7.43 ± 0.08 62.00 ± 59.19 0.128 ± 0.025 2.30 ± 0.24 0.14 ± 0.04 1.08 ± 0.15 1.27 ± 0.34 13,026 ± 2083

149.48 ± 68.62 1.20 ± 0.15 7.18 ± 0.15 7.55 ± 0.10 150.00 ± 42.90 0.126 ± 0.013 2.01 ± 0.16 0.17 ± 0.03 1.37 ± 0.36 1.88 ± 0.29 12,778 ± 2667

260

Y. Li et al. / Science of the Total Environment 640–641 (2018) 255–263

Table 3 The extreme values, anchors and formulas calculation for the selected metrics making up the bacteria-based index of biotic integrity (Ba-IBI). Number

Metrics

Minimum

Maximum

Lower anchor

Upper anchor

Response

Formula

M24 M29 M105 M132 M135

% Acidobacteria % Gemmatimonadetes % Geobacter Methanotroph Phototroph

0.195 0.079 0.022 0.071 0.071

0.486 0.252 0.116 0.152 0.221

0.236 0.098 0.028 0.084 0.084

0.453 0.221 0.113 0.152 0.192

Increase Increase Increase Decrease Increase

(0.486 − M24) / (0.486 − 0.236) (0.252 − M29) / (0.252 − 0.098) (0.116 − M105) / (0.116 − 0.028) M132 / 0.152 (0.221 − M135) / (0.221 − 0.084)

that the residuals were normally distributed. The correlation coefficients and statistical parameters are shown in Table S4. SS, TN, E. coli and COD contributed a lot for predicting the Ba-IBI with p b 0.05, followed by water level with p b 0.10. The results of Ba-IBI prediction based on the multivariate linear regression model (R2 = 0.76) using the physicochemical parameters are shown in Fig. 5. 4. Discussion 4.1. Development of the bacteria-based index of biotic integrity In the development of IBI, selection of reference sites is a critical step. Reference conditions refer to the sites with little or no influence from human society. However, all the sampling sites in this study were located in the artificial reservoir, which may be significant influenced by human activities. In such cases, a reference condition is predicted for comparable environmental condition, rather than quantitatively defined (Hughes and R.M, 1994). Therefore, our selection of reference sites was based on the relative ecological conditions, including natural environments and human disturbances along the TGR. Similar methods have been used in earlier studies (J. Huang et al., 2015; Matono et al., 2012). The ecological index indicated the environmental conditions of the selected reference sites were most closed to the original conditions and least impaired by human activities. The Ba-IBI development was considered not only statistical point of view, but also an ecological context, thus providing a quantitative and comprehensive investigation of bacterial community. In order to contain as much bacterial community information as possible, we constructed a pool of candidate metrics, which could reflect the biotic and ecological conditions in the TGR. The distinct bacterial community composition and function result in the variation of candidate metrics in different studies. After the step-by-step screening, five independent metrics were selected. Acidobacteria, widespread distributed in soil,

fresh water, acid mine stream, and aquatic sediments (Zimmermann et al., 2012), are believed to play an important role in biogeochemical processes (Barns et al., 1999) and could be an indicator of the environmental variations. Gemmatimonadetes phylum, making up approximately 2% of the total soil bacteria as one of the top nine phyla found in soils, are demonstrated to be unique in low soil moisture and are sensitive to environment changes (Debruyn et al., 2011). Geobacter, widely distributed in soils and aquatic sediment, was the first organism to have the ability to oxidize organic compounds and metals, and was demonstrated to be the predominant Fe(III)-reducing bacteria in most environments (Childers et al., 2002). Methanotrophs, consuming most of the methane in streams, provide an essential service for the aquatic environments (Deng et al., 2016). Due to the emission of natural methane in the TGR, Methanotrophs are believed to be an indispensable index for the development of Ba-IBI. Phototrophs, obtaining energy from light to carry out various cellular metabolic processes, play an important role in the processes of metabolism, degradation, and the use in current risk assessments (Hand et al., 2001). Comparing the obtained core metrics from this study with the previous studies based on bacterial communities, the discrimination could be attributed to the distinct natural environmental conditions, land use types and human interference degrees among the study areas. The focus of assessment on Qinhuai River (J. Li et al., 2017), a typical urban river, is the organic pollution and eutrophication induced by human disturbances. Thus, metrics such as proportion of Paenibacillus, Nitrosomonas, and OTUs tolerant to organic pollution were selected due to their crucial role in the process of bioremediation and nitrogen cycling, and their potential to indicate organic pollution levels. In the assessment of Taihu Basin (Y. Li et al., 2017), one of the most developed and highly populated areas in China, the intensive anthropogenic disturbances such as point and non-point source pollution have become the focus of research. Therefore, the selected metrics were largely associated with antibiotic resistance, nitrification, ammoxidation and

Fig. 2. Variations of the calculated bacteria-based index of biotic integrity (Ba-IBI) at 12 sampling sites along the Three Gorges Reservoir in low water level period, impoundment period and sluicing period, respectively. The yellow bars outline the reference sites (FJ, GDK, MH) where the Ba-IBI are relative high. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Y. Li et al. / Science of the Total Environment 640–641 (2018) 255–263

261

Fig. 3. Classification of the 12 sampling sites based on the results of the bacteria-based index of biotic integrity in (a) low water level period, (b) impoundment period and (c) sluicing period.

pathogen control, contributing to the ecological protection in the basin. While in this study, the selected metrics were more applicable in the deep water reservoirs and exhibited higher potential for reflecting

ecological health of the TGR and predicting future environmental variations. 4.2. Ecological assessment of the TGR in different operation periods Rivers are open systems, and the community composition, structure and function are closely related to external disturbances. The TGR exhibited an obvious spatial heterogeneity in bacterial communities as a result of the common features of the deep water reservoirs, including

Fig. 4. Redundancy analysis of the relationships among species assemblages, environmental parameters, the core metrics and the scores of bacteria-based index of biotic integrity (Ba-IBI).

Fig. 5. Correlation analysis of the predicted and observed values of bacteria-based index of biotic integrity (Ba-IBI). The solid line is the idea line where the model perfectly fit the data set. The dashed line is the linear fit of predicted Ba-IBI and observed Ba-IBI.

262

Y. Li et al. / Science of the Total Environment 640–641 (2018) 255–263

the gradients of flow velocity, hydraulic retention time, light and nutrient availability (Galat, 1990). Significant increase in the Ba-IBI was observed from lower to upper reaches of the TGR. High variability of bacterial communities and biotic integrity were mainly associated with environmental and ecological changes, especially the increase of nutrient concentrations, such as SS, TN, BOD5 and COD, which is consistent with previous studies (Gao et al., 2017; Z. Li et al., 2017). Anthropogenic activities mainly contributed to the nutrient parameters variation. High nutrient concentrations were found in catchments with high urban land cover, indicating the possible contributions of urban runoff and domestic sewage to nutrient pollution. For example, most of the upstream sites, such as ZT, TGY, CT and CS, were located in the economically advanced centers with large population and diverse industrial enterprise. Urbanization has triggered a shift from agriculture to industry such as steel or automobile manufacturing, pulp and paper industry, electronics industry and new energy industry. These industries, accompanied by sewage discharge during human activities, are the main source of pollutants in the TGR. The agricultural runoff is generally also an important source of high nutrient concentrations, causing the significant environmental discrimination among the sampling sites. Most of the downstream sites maintained a better ecological health, except some sites (XXH and DNH) where the tributaries flow into the Yangtze River, carrying multiple disturbances from the upstream of the tributaries. This study revealed a significant tendency that the ecological health in the low water level and sluicing periods was better than that in the impoundment period, especially at the impaired sites in the lower reaches of the TGR. The profound discrepancy in the impoundment period compared with the other two periods could probably be attributed to the reservoir operation. The hydrological factors, including water level and flow velocity, were controlled by the dam regulations and runoff changes (Wang et al., 2012). The water level increased from 145 m to 175 m generated significant hydrological disturbances. And water flow was considered as one of the most important factors influencing the microbial communities within fluvial biofilms (Li et al., 2015), decreased significantly by reservoir operation (Bier et al., 2015). In the impoundment period, high water level and low flow conditions directly affected the interchange of materials in water body. Particularly under impaired conditions, low flow caused the intensification of eutrophication due to the limitation of nutrients dilution and acceleration of primary production (Matono et al., 2012). These hydrodynamic and physicochemical changes had potential impacts on the bacterial community structure and function (Weise et al., 2016), resulting in the generally decrease of biotic integrity or ecological health. The results were in accordance with previous studies, demonstrating that reservoir operation could affect environmental parameters, phytoplankton and bacterioplankton compositions (Ruiz-Gonzalez et al., 2013; Yan et al., 2015). Overall, this study may be among the first to quantitatively assess ecological health of the TGR based on bacterial communities and to evaluate the potential effects of reservoir operation on biotic integrity. 4.3. Evaluation and future application of the assessment method Aquatic ecosystems are characterized by complex interaction among hydrological, physicochemical and biological processes, and are impaired by increasing human activities. External disturbances may alter the ecosystem and thus modify the resident biological communities. Biological criteria are widely used in ecological health assessment due to their ability to integrate the impacts of measured and unmeasured factors, thereby providing a better evaluation of environment conditions than water chemistry alone (Colin et al., 2016). They do not replace physicochemical methods, but they do increase the probability of assessment on detecting degradation due to anthropogenic influences (Karr, 1991). In this study, the developed index offered a quantitative measure of biotic integrity based on bacterial diversity, composition and function. The Ba-IBI could also reflect the ecological health due to

its significant correlation with environmental conditions. Besides the good performance of Ba-IBI in discriminating the reference and impaired sites (Fig. S2), the strong correlation between Ba-IBI and the ecological index (Fig. S3) was also observed. Thus, the local habitat variables were the ones most relevant for bacterial communities and the capability of Ba-IBI to apply an effective bioassessment of ecological health in the TGR was demonstrated. As an evaluation method, it is worth mentioning the weaknesses of the developed Ba-IBI and further applications to other environments. Firstly, since we only used data at 12 sampling sites in three operation periods, the sensitivity of the metrics to discriminate ecological conditions can be further enhanced by expanding to a larger catchment. Secondly, a more comprehensive IBI including more kinds of microorganism such as archaea and fungi, was needed for further assessing ecological health. To increase the universality and applicability of this quantitative assessment method for larger spatial scales, the index still needs improvements such as the selection of reference conditions (Elias et al., 2016), calibration for natural variance (Pereira et al., 2016) and probabilistic sampling designs (Hughes and Peck, 2008). After being applied to other areas, this method could provide guidance for reservoir operation, which is the central link to the regulation of hydropower projects. The regulation of reservoirs should consider not only safety and economic factors, but also the significant contributions of both anthropogenic activities and hydrodynamic conditions. 5. Conclusions This study presented an assessment method to quantify ecological health in the largest artificial reservoir, the Three Gorges Reservoir. A bacteria-based index of biotic integrity (Ba-IBI) was developed and validated. This index was capable of discriminating reference and impaired conditions and reflecting ecological health of the TGR effectively. We identified the spatial variability in biotic integrity of the TGR in different operation periods, explored the relationship between Ba-IBI and physicochemical parameters, and qualified the contributions of hydrodynamic conditions and anthropogenic disturbances on ecological health. The results showed a better condition in the lower reaches of the TGR than that in the upper reaches, and indicated a decrease of biotic integrity in the impoundment period compared with the low water level and sluicing periods. This ecological assessment method could be used as reference for the regulation of the TGR, and could also be used in the bioassessment of other similar areas. Acknowledgements This study was supported by the National Natural Science Foundation of China (No. 91547105 and 51779076); the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (No. 51421006); the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD); the Six Talent Peaks Project in Jiangsu Province (2016-JNHB-007); the fifth 333 High Level Talents training Project of Jiangsu Province. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2018.05.291. References Abdelkefi, A., Rui, V., Nayfeh, A.H., Hajj, M.R., 2013. Linking land use, in-stream stressors, and biological condition to infer causes of regional ecological impairment in streams. Freshw. Sci. 32 (3), 801–820. Arndt, D., Xia, J.G., Liu, Y.F., Zhou, Y., Guo, A.C., Cruz, J.A., Sinelnikov, I., Budwill, K., Nesbo, C.L., Wishart, D.S., 2012. METAGENassist: a comprehensive web server for comparative metagenomics. Nucleic Acids Res. 40 (W1), W88–W95.

Y. Li et al. / Science of the Total Environment 640–641 (2018) 255–263 Barbour, M.T., Gerritsen, J., Griffith, G.E., Frydenborg, R., Mccarron, E., White, J.S., Bastian, M.L., 1996. A framework for biological criteria for Florida streams using benthic macroinvertebrates. J. N. Am. Benthol. Soc. 15 (2), 185–211. Barns, S.M., Takala, S.L., Kuske, C.R., 1999. Wide distribution and diversity of members of the bacterial kingdom acidobacterium in the environment. Appl. Environ. Microbiol. 65 (4), 1731–1737. Bier, R.L., Voss, K.A., Bernhardt, E.S., 2015. Bacterial community responses to a gradient of alkaline mountaintop mine drainage in central Appalachian streams. ISME J. 9 (6), 1378–1390. Casatti, L., Ferreira, C.P., Langeani, F., 2009. A fish-based biotic integrity index for assessment of lowland streams in southeastern Brazil. Hydrobiologia 623 (1), 173–189. Chen, Z., Li, J., Shen, H., Wang, Z., 2001. Yangtze River of China: historical analysis of discharge variability and sediment flux. Geomorphology 41 (2), 77–91. Childers, S.E., Ciufo, S., Lovley, D.R., 2002. Geobacter metallireducens accesses Fe(III) oxide by chemotaxis. Nature 416 (6882), 767–769. Colin, N., Maceda-Veiga, A., Flor-Arnau, N., Mora, J., Fortuã, O.P., Vieira, C., Prat, N., Cambra, J., De, S.A., 2016. Ecological impact and recovery of a Mediterranean river after receiving the effluent from a textile dyeing industry. Ecotoxicol. Environ. Saf. 132, 295–303. Debruyn, J.M., Nixon, L.T., Fawaz, M.N., Johnson, A.M., Radosevich, M., 2011. Global biogeography and quantitative seasonal dynamics of gemmatimonadetes in soil. Appl. Environ. Microbiol. 77 (17), 6295. Deng, Y., Liu, Y., Dumont, M., Conrad, R., 2016. Salinity affects the composition of the aerobic methanotroph community in alkaline lake sediments from the Tibetan Plateau. Microb. Ecol. 73 (1), 101–110. Elias, C.L., Calapez, A.R., Almeida, S.F.P., Chessman, B., Simões, N., Feio, M.J., 2016. Predicting reference conditions for river bioassessment by incorporating boosted trees in the environmental filters method. Ecol. Indic. 69, 239–251. Franssen, N.R., 2012. Genetic structure of a native cyprinid in a reservoir-altered stream network. Freshw. Biol. 57 (1), 155–165. Galat, D., 1990. Reservoir limnology: ecological perspectives. Trans. Am. Fish. Soc. 121 (5), 696–698. Gao, Q., Li, Y., Cheng, Q., Yu, M., Hu, B., Wang, Z., Yu, Z., 2016. Analysis and assessment of the nutrients, biochemical indexes and heavy metals in the three gorges reservoir, China, from 2008 to 2013. Water Res. 92, 262. Gao, J.M., Wu, L., Chen, Y.P., Zhou, B., Guo, J.S., Zhang, K., Ouyang, W.J., 2017. Spatiotemporal distribution and risk assessment of organotins in the surface water of the Three Gorges Reservoir Region, China. Chemosphere 171, 405–414. Hammer, Ø., Harper, D.A.T., 2009. Past: Paleontological Statistics Software Package for Educaton and Data Anlysis. Han, C., Zheng, B., Qin, Y., Ma, Y., Yang, C., Liu, Z., Cao, W., Chi, M., 2017. Impact of upstream river inputs and reservoir operation on phosphorus fractions in waterparticulate phases in the Three Gorges Reservoir. Sci. Total Environ. 610, 1546–1556. Hand, L.H., Kuet, S.F., Lane, M.C., Maund, S.J., Warinton, J.S., Hill, I.R., 2001. Influences of aquatic plants on the fate of the pyrethroid insecticide lambda-cyhalothrin in aquatic environments. Environ. Toxicol. Chem. 20 (8), 1740–1745. Hu, M., Huang, G.H., Sun, W., Ding, X., Li, Y., Fan, B., 2016. Optimization and evaluation of environmental operations for three gorges reservoir. Water Resour. Manag. 30 (10), 3553–3576. Huang, L., Fang, H., Reible, D., 2015. Mathematical model for interactions and transport of phosphorus and sediment in the Three Gorges Reservoir. Water Res. 85, 393–403. Huang, Q., Gao, J., Cai, Y., Yin, H., Gao, Y., Zhao, J., Liu, L., Huang, J., 2015. Development and application of benthic macroinvertebrate-based multimetric indices for the assessment of streams and rivers in the Taihu Basin, China. Ecol. Indic. 48 (1), 649–659. Hughes and R.M, 1994. Defining biological status by comparing with reference conditions (chapter 4). Book Chapter. Hughes, R.M., Peck, D.V., 2008. Acquiring data for large aquatic resource surveys: the art of compromise among science, logistics, and reality. J. N. Am. Benthol. Soc. 27 (4), 837–859. Jia, Y., Sui, X., Chen, Y., 2013. Development of a fish-based index of biotic integrity for wadeable streams in Southern China. Environ. Manag. 52 (4), 995–1008. Karr, J.R., 1981. Assessment of biotic integrity using fish communities. Fisheries 6 (6), 21–27. Karr, J.R., 1991. Biological integrity: a long-neglected aspect of water resource management. Ecol. Appl. 1 (1), 66–84. Karr, J.R., Fausch, K.D., Angermeier, P.L., Yant, P.R., Schlosser, I.J., 1986. Assessing biological integrity in running water: a method and its rationale. Illinois Natural History Survey Special Publication. 1–28 No. 5, p. 5. Klemm, D.J., Blocksom, K.A., Fulk, F.A., Herlihy, A.T., Hughes, R.M., Kaufmann, P.R., Peck, D.V., Stoddard, J.L., Thoeny, W.T., Griffith, M.B., Davis, W.S., 2003. Development and evaluation of a macroinvertebrate biotic integrity index (MBII) for regionally assessing mid-Atlantic highlands streams. Environ. Manag. 31 (5), 656–669. Lau, K.E.M., Washington, V.J., Fan, V., Neale, M.W., Lear, G., Curran, J., Lewis, G.D., 2015. A novel bacterial community index to assess stream ecological health. Freshw. Biol. 60 (10), 1988–2002. Li, K., Zhu, C., Wu, L., Huang, L., 2013. Problems caused by the Three Gorges Dam construction in the Yangtze River basin: a review. Environ. Rev. 21 (3), 127–135.

263

Li, Y., Wang, C., Zhang, W., Wang, P., Niu, L., Hou, J., Wang, J., Wang, L., 2015. Modelling the effects of hydrodynamic regimes on microbial communities within fluvial biofilms: combining deterministic and stochastic processes. Environ. Sci. Technol. 49 (21), 12869–12878. Li, J., Li, Y., Qian, B., Niu, L., Zhang, W., Cai, W., Wu, H., Wang, P., Wang, C., 2017. Development and validation of a bacteria-based index of biotic integrity for assessing the ecological status of urban rivers: a case study of Qinhuai River basin in Nanjing, China. J. Environ. Manag. 196, 161. Li, Y., Niu, L., Wang, P., Zhang, W., Wang, C., Li, J., Wu, H., 2017. Development of a microbial community-based index of biotic integrity (MC-IBI) for the assessment of ecological status of rivers in the Taihu Basin, China. Ecol. Indic. 85. Li, Z., Lu, L.H., Guo, J.S., Yang, J.X., Zhang, J.C., He, B., Xu, L.L., 2017. Responses of spatialtemporal dynamics of bacterioplankton community to large-scale reservoir operation: a case study in the Three Gorges Reservoir, China. Sci. Rep. 7. Ma, X., Li, Y., Zhang, M., Zheng, F., Du, S., 2011. Assessment and analysis of non-point source nitrogen and phosphorus loads in the Three Gorges Reservoir area of Hubei Province, China. Sci. Total Environ. 412 (412–413), 154–161. Marzin, A., Archaimbault, V., Belliard, J., Chauvin, C., Delmas, F., Pont, D., 2012. Ecological assessment of running waters: do macrophytes, macroinvertebrates, diatoms and fish show similar responses to human pressures? Ecol. Indic. 23 (4), 56–65. Matono, P., Bernardo, J.M., Oberdorff, T., Ilhéu, M., 2012. Effects of natural hydrological variability on fish assemblages in small Mediterranean streams: implications for ecological assessment. Ecol. Indic. 23 (23), 467–481. Mercado-Silva, N., Lyons, J.D., Maldonado, G.S., Nava, M.M., 2002. Validation of a fishbased index of biotic integrity for streams and rivers of central Mexico. Rev. Fish Biol. Fish. 12 (2), 179–191. Oksanen, J., FBlanchet, G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P., BO'Hara, R., Simpson, G., Solymos, P., et al., 2018. vegan: community ecology package. Ordination methods, diversity analysis and other functions for community and vegetation ecologists. Version 2.4–6 https://CRAN.R-project.org/package=vegan. Pereira, P.S., Souza, N.F., Baptista, D.F., Oliveira, J.L.M., Buss, D.F., 2016. Incorporating natural variability in the bioassessment of stream condition in the Atlantic Forest biome, Brazil. Ecol. Indic. 69, 606–616. Ruiz-Gonzalez, C., Proia, L., Ferrera, I., Gasol, J.M., Sabater, S., 2013. Effects of large river dam regulation on bacterioplankton community structure. FEMS Microbiol. Ecol. 84 (2), 316–331. Stoddard, J., Herlihy, A., Peck, D., Hughes, R., Whittier, T., Tarquinio, E., 2008. A process for creating multimetric indices for large-scale aquatic surveys. Freshw. Sci. 27 (4), 878–891. Teng, M.J., Zeng, L.X., Xiao, W.F., Huang, Z.L., Zhou, Z.X., Yan, Z.G., Wang, P.C., 2017. Spatial variability of soil organic carbon in three gorges reservoir area, China. Sci. Total Environ. 599, 1308–1316. Villeneuve, B., Piffady, J., Valette, L., Souchon, Y., Usseglio-Polatera, P., 2018. Direct and indirect effects of multiple stressors on stream invertebrates across watershed, reach and site scales: a structural equation modelling better informing on hydromorphological impacts. Sci. Total Environ. 612, 660–671. Wang, Y., Xia, Z., Wang, D., 2012. A transitional region concept for assessing the effects of reservoirs on river habitats: a case of Yangtze River, China. Ecohydrology 5 (1), 28–35. Wang, J.N., Dong, Z.R., Liao, W.G., Li, C., Feng, S.X., Luo, H.H., Peng, Q.D., 2013. An environmental flow assessment method based on the relationships between flow and ecological response: a case study of the Three Gorges Reservoir and its downstream reach. Sci. China Technol. Sci. 56 (6), 1471–1484. Wang, Z.D., Xiao, G.S., Zhou, N., Qi, W.H., Han, L., Ruan, Y., Guo, D.Q., Zhou, H., 2015. Comparison of two methods for detection of fecal indicator bacteria used in water quality monitoring of the Three Gorges Reservoir. J. Environ. Sci. 38, 42–51. Weise, L., Ulrich, A., Moreano, M., Gessler, A., Kayler, Z.E., Steger, K., Zeller, B., Rudolph, K., Knezevic-Jaric, J., Premke, K., 2016. Water level changes affect carbon turnover and microbial community composition in lake sediments. FEMS Microbiol. Ecol. 92 (5). Wilcox, D.A., Meeker, J.E., Hudson, P.L., Armitage, B.J., Black, M.G., Uzarski, D.G., 2002. Hydrologic variability and the application of index of biotic integrity metrics to wetlands: a great lakes evaluation. Wetlands 22 (3), 588–615. Wu, N.C., Schmalz, B., Fohrer, N., 2012. Development and testing of a phytoplankton index of biotic integrity (P-IBI) for a German lowland river. Ecol. Indic. 13 (1), 158–167. Yan, Q., Bi, Y., Deng, Y., He, Z.L., Wu, L.Y., Van Nostrand, J.D., Shi, Z., Li, J.J., Wang, X., Hu, Z.Y., Yu, Y.H., Zhou, J.H., 2015. Impacts of the Three Gorges Dam on microbial structure and potential function. Sci. Rep. 5. Zalack, J.T., Smucker, N.J., Vis, M.L., 2010. Development of a diatom index of biotic integrity for acid mine drainage impacted streams. Ecol. Indic. 10 (2), 287–295. Zhou, B., Shang, M., Wang, G., Li, F., Shan, K., Liu, X., Ling, W., Zhang, X., 2017. Remote estimation of cyanobacterial blooms using the risky grade index (RGI) and coverage area index (CAI): a case study in the Three Gorges Reservoir, China. Environ. Sci. Pollut. Res. Int. 24 (23), 19044–19056. Zimmermann, J., Portillo, M.C., Serrano, L., Ludwig, W., Gonzalez, J.M., 2012. Acidobacteria in freshwater ponds at Doñana National Park, Spain. Microb. Ecol. 63 (4), 844–855.