Previous Article in Journal
Migration of Nanoplastic in Soil: Effects of Polymer Properties and Rainfall Conditions
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Water
Volume 17
Issue 24
10.3390/w17243513
water-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Phytoplankton Size as an Ecological Bioindicator in a Subtropical Fragmented River, China

by
Deyu Sang
1,2,
Jingxin Wei
3,
Caiqin Hu
1,4,5,6,
Qianfu Liu
1,4,5,6,
Jinhui Sun
2,* and
Chao Wang
1,4,5,6,7,*
1
Pearl River Fisheries Research Institute, Chinese Academy of Fishery Sciences, Guangzhou 510380, China
2
Tianjin Key Lab of Aqua-Ecology and Aquaculture, Tianjin Agricultural University, Tianjin 300384, China
3
Department of Biological Science and Technology, School of Life Science and Technology, Jinan University, Guangzhou 510632, China
4
Fishery Ecological Environment Monitoring Center of Pearl River Basin, Ministry of Agriculture and Rural Affairs, Guangzhou 510380, China
5
Guangzhou Scientific Observing and Experimental Station of National Fisheries Resources and Environment, Guangzhou 510380, China
6
Guangdong Provincial Key Laboratory of Aquatic Animal Immunology and Sustainable Aquaculture, Guangzhou 510380, China
7
Key Laboratory of Prevention and Control for Aquatic Invasive Alien Species, Ministry of Agriculture and Rural Affairs, Guangzhou 510000, China
*
Authors to whom correspondence should be addressed.
Water 2025, 17(24), 3513; https://doi.org/10.3390/w17243513 (registering DOI)
Submission received: 30 September 2025 / Revised: 1 December 2025 / Accepted: 4 December 2025 / Published: 12 December 2025
(This article belongs to the Section Biodiversity and Functionality of Aquatic Ecosystems)
Browse Figures
Versions Notes

Abstract

This study investigates the Serial Discontinuity Concept (SDC) by analyzing the size of phytoplankton structures across 13 cascade dams in the fragmented East River, China. The results showed that dam flow-regulation minimized seasonal differences in total chlorophyll-a (Chl-a). Spatially, midstream Chl-a was higher during the dry period, whereas increased wet periods were linked to reservoir effects and nutrient conditions. Nano-phytoplankton dominated during both periods, while micro-phytoplankton declined during wet periods due to higher pH and transparency. Micro-Chl-a increased downstream in dry periods as a result of dissolved oxygen levels and silicate. Self-organizing maps revealed greater size–class variability during dry periods, with pH and conductivity identified as key regulators. Aulacoseira granulata (micro-phytoplankton) and Anabaena oscillarioides (also micro-phytoplankton) were co-dominant. This pattern suggests that the flow regulation and water impoundment by cascade dams during the wet period created localized lentic conditions with enhanced water stability, which favored the proliferation of these species, despite the increased seasonal discharge at the basin scale. These findings support the SDC in that (1) longitudinal Chl-a variations empirically validated SDC, especially during dry periods, and (2) a spatially periodic Chl-a pattern was identified, termed the Cascade Cycle of SDC (CC-SDC).

1. Introduction

Phytoplankton, utilizing light energy for photosynthesis to generate organic matter, serve as the primary producers in aquatic ecosystems; they form the foundation of the aquatic food chain. The concentrations of chlorophyll-a (Chl-a) are widely recognized as a key indicator for measuring phytoplankton biomass and assessing phytoplankton size distribution [1]. Phytoplankton are typically categorized by size into micro-phytoplankton (20–200 μm), nano-phytoplankton (2–20 μm), and pico-phytoplankton (<2 μm) [2]. These size-fragmented phytoplankton have their own important roles in the ecosystem, and are characterized by different growth rates [3], nutrient uptake rates [4,5], energy flows [6], and sedimentation characteristics [7]. Studying the composition of phytoplankton size is crucial for understanding the material cycles and energy flows of ecosystems [8]. Currently, a substantial body of research on phytoplankton size composition exists in marine ecosystems, with numerous studies reported from the Indian Ocean [9], the Atlantic Ocean [10,11], and the Pacific Ocean [12]. In freshwater systems, the importance attributed to the size of phytoplankton structures has also been recognized, particularly in lakes, reservoirs, and estuaries [13,14]. However, studies focusing on river ecosystems, especially those fragmented by multiple cascade dams, remain less common. The unique hydrological and physicochemical gradients created by serial impoundments impose distinct selective pressures on phytoplankton communities. Therefore, investigating the response of phytoplankton structure size in regulated rivers and integrating these findings with theoretical frameworks like the Serial Discontinuity Concept (SDC) is a crucial yet underrepresented area in the current literature [15,16]. This study aims to fill this specific knowledge gap. The Serial Discontinuity Concept (SDC) provides a theoretical framework for understanding and predicting the impacts of dams on the composition of river ecosystems and its functions [17]. In the original SDC [18], dams were regarded as the most common disturbances causing discontinuities in the river. These discontinuities enhance spatial differences between upstream and downstream rivers by causing significant distinctions in abiotic parameters and biological communities. The SDC predicts these effects through two variables: the “discontinuum distance” and “parameter intensity”. “Discontinuum distance” refers to the distance at which the expected value of a physical or biological variable changes upstream or downstream due to discontinuity caused by a dam. “Parametric strength” refers to the absolute change that occurs from regulating rivers, often expressed as a deviation from natural or reference conditions [19]. Stanford and Ward [20] re-evaluated the SDC in nine regulated rivers, and six rivers confirmed the SDC’s prediction. The investigations by Ellis and Jones [21] in the Magpie and Michipicoten Rivers (Canada) also provide support for the SDC. Matrix size, epiphytic biomass, and drift density increased below the dam and quickly recovered within 5 km downstream, aligning with SDC predictions. However, benthic invertebrate richness also recovered rapidly, contrary to SDC expectations.
The East River, a primary tributary of the Pearl River system, begins in the Yajibo Mountain located in Xunwu County, Jiangxi Province [22]. The river flows about 562 km from northeast to southwest, passing through Longchuan County, Huicheng District, Boluo County, and other regions in Guangdong Province, before merging with the Pearl River Delta and discharging into the Pearl River Estuary. According to Li et al. [23], the basin’s catchment area measures 35,340 square kilometers, with an average annual runoff of 25.7 billion cubic meters. The basin’s primary water resources for Guangdong, Hong Kong, and Macao include the Xinfengjiang and Fengshuba Reservoirs, along with the Dongshen Water Supply Project [16]. However, heavily developed cascade dams in the Pearl River Basin have had a significant impact on river habitats and connectivity.
Few studies have utilized phytoplankton to validate SDC predictions, despite phytoplankton species richness being recognized as an effective biological indicator for water bodies in the Pearl River Delta network [24]. In this study, the size composition of phytoplankton was investigated for 13 cascade dams of the East River. We explored the following: (1) whether phytoplankton size composition could be used as a biological indicator to reflect the environmental response of the East River cascade dam, and (2) whether the results support SDC prediction. Specifically, we hypothesize that shifts in the relative proportions of micro-, nano-, and pico-phytoplankton can serve as robust bioindicators for dam-induced alterations in hydrology and water chemistry. Hence, this study initially examined the temporal and spatial distribution and correlation between environmental factors and size-fragmented phytoplankton. Subsequently, distance was used as a spatial predictor to assess the accuracy of SDC predictions. Finally, a prediction model was developed to evaluate the impact of cascade dams on size-fragmented phytoplankton.

2. Materials and Methods

2.1. Study Area

This research focused on the primary course of the East River (Figure 1a,b). The sequence of 13 cascade dams from upstream to downstream includes Fengshu Dam (FSB), Longtan (LT), Renkeng (RK), Luoyingkou (LYK), Sulei Dam (SLB), Zhentouzhai (ZTZ), Liucheng (LC), Lankou (LK), Huangtian (HT), Mujing (MJ), Fengguang (FG), Likou (LIK), and Dongjiang (DJ). The station layout is shown in Figure 1c, with the No.1 station of each cascade dam representing the dam reservoir area; the river section containing two dams included 3–6 sampling stations according to the distance. The study area was segmented into three sections: upstream (stations FSB1 to ZTZ1), midstream (stations ZTZ2 to FG1), and downstream (stations FG2 to DJ6) areas. In addition, two stations, Meiliping (MLP) and Pengkeng (PK), were set up in the upstream section of the first dam as reference points.

2.2. Field Sampling and Data Collection

The survey was conducted during the dry period in November 2018 and the wet period in July 2019. Sampling for each period was performed within 10 days. At each sampling location, 1 L of surface water (collected 0.5 m below the surface) was stored in a polyethylene bottle and kept in a dark refrigerator at 4 °C. In the laboratory, the water sample underwent Chl-a size classification and was filtered through membranes with pore sizes of 20 μm, 2 μm, and 0.2 μm under vacuum. The particle sizes corresponded to micro-, nano-, and pico-sizes. The filtered membrane was cut and placed into the centrifuge tube, which was added to 5 mL of a 90% acetone solution and shaken. The mixed solution was placed in the refrigerator at 4 °C and removed after 22 to 24 h. After centrifugation (3500 rpm; 10 min), the optical density values were measured at wavelengths of 630 nm, 664 nm, 647 nm, and 750 nm with a ultraviolet-visible spectrophotometer (UV-5200, Metash. Inc, Shanghai, China), and were recorded as OD630, OD664, OD647, and OD750, respectively. The Chl-a concentration was calculated as follows:
C (mg/L−1) = [11.85 × (OD664 − OD750) − 1.54 × (OD647 − OD750) − 0.08 × (OD630 − OD750)] × 5/1000
The total Chl-a concentration was the sum of three Chl-a concentration sizes [15]. Water samples were collected for species identification alongside those for chlorophyll-a analysis. A 1 L water sample was immediately fixed with Lugol’s iodine solution, allowed to settle for 48 h, and then concentrated to 30 mL. Species identification was performed using an inverted microscope (Zeiss Axio Observer A1, Carl Zeiss AG, Oberkochen, Germany).
A water flow injection analyzer (Skalar-SA1100, Skalar Inc., Breda, The Netherlands) was used to measure total phosphorus (TP), total nitrogen (TN), and silicate (SiO4). In situ measurements of physical environmental factors such as water temperature, pH, oxidation-reduction potential, salinity, and dissolved oxygen were conducted using a portable YSI meter (YSI6600V2, YSI Environment. Inc., Yellow Springs, OH, USA). A Secchi disk was used to assess transparency, and the river width was measured using a diastimeter (a multifunction laser rangefinder).

2.3. Data Analysis

Phytoplankton identification was performed referring to the method of [16]; Redundancy analysis (RDA) was employed in R 3.4.1 to explore the associations between Chl-a size composition and environmental parameters. RDA was performed using the R add-on package “vegan”, employing forward selection to exclude collinear environmental variables from the constrained ordination model. Parameters significantly influencing phytoplankton size composition (p < 0.05, determined by 1000 permutation tests) were included in the model.
To analyze the correlation between distance and size composition, we used Google Earth (v7.1.7.2606) to collect distance data based on the latitude and longitude of the sampling points, and both large-scale data (from upstream to downstream) and data from each cascade dam were collected. Large-scale data referred to the distance of each sampling site (excluding the first one) from the first site. Data obtained from each cascade dam include the distance of each sampling site from the dam in that section. Since the distance data were the same for each season, the relationship with phytoplankton size composition was analyzed for each period using Origin 8.5, respectively.
The samples were categorized using a self-organizing map (SOM) based on morphological parameters, which is an effective method for automatic classification through unsupervised learning. Samples with similar morphological parameters were grouped into the same or adjacent neurons based on their similarity. The SOM produced 56 neurons, arranged in a 9 × 8 hexagonal lattice. The map size was determined using the formula 5 × (number of samples)1/2 [25], and optimization was performed based on the smallest quantification and topographic errors. Using Ward’s linkage method, the map’s cells were grouped based on the similarity of the neurons’ weighted vectors. Group numbers were primarily determined by the dissimilarity levels of each SOM cell in the hierarchical clustering. The U-matrix [26] and the Davies–Bouldin index [27] were utilized to enhance group definitions, and analyses were conducted using MATLAB 2010 [28] with the SOM toolbox [29].
To assess the effectiveness of hierarchical clustering on the SOM units, the cophenetic correlation coefficient [30] was calculated in R [31]. The contributions of each input component with respect to cluster compositions were obtained from the weighted SOM vectors and then visualized in a boxplot [32]. We used the ANOVA test to compare differences in morphological parameters among the clusters in R.
Linear discriminant analysis (LDA) was carried out to determine which environmental variables differentiated between the clusters previously defined by the SOM procedure. Standardized coefficients represented the contribution of each variable to the discrimination between clusters. A random Monte Carlo test with 1000 permutations was used to reveal the significance of environmental variables among clusters. The Kruskal–Wallis test was then carried out to identify the differences between environmental variables among clusters. Cohen’s Kappa test was conducted to determine the significance of the prediction using the “irr” package.
The percentage map showing the total chlorophyll-a concentration and particle size classification for various groups was completed using Origin 8.5 software.

3. Results

3.1. Water Environment Characteristics

The river width, transparency, WT, pH, dissolved oxygen (DO), TP, and SiO4 concentration were the same for environmental factors as those monitored by [16]. There were clear temporal differences in transparency, WT, pH, and conductivity. The range of transparency was 5–130 cm with an average of 26 ± 16.51 cm during the wet period; this differed during the dry period with transparency at 40–250 cm and an average of 94 ± 57.22 cm. During the dry period, the transparency was higher upstream than in the mid- and downstream. The WT was notably higher in the wet period (29.03 ± 1.37 °C) compared to the dry period (22.89 ± 1.61 °C). Higher WT was recorded upstream during the dry period. The pH, transparency, and conductivity in the dry period were significantly higher than in the wet period. The pH varied from 7.26 to 9.18 with an average value of (7.73 ± 0.33) in the wet period, while it ranged from 7.99 to 8.51 with an average value of (8.20 ± 0.11) in the dry period. There was a gradually decreasing trend in the pH from upstream to downstream during the wet period. Conductivity varied temporally, ranging from 19.91 to 155.57 μS/cm during the wet period with an average of (84.42 ± 20.40) μS/cm. The conductivity during the dry period ranged from 97.51 to 176.65 μS/cm, with an average value of (134.27 ± 24.96) μS/cm. During the wet period, higher conductivity was observed downstream (Figure 2).
Moreover, no notable temporal or spatial characteristics were observed in TP and SiO4 concentration. TP ranged from 0.028 to 0.835 mg/L during the wet period with an average value of (0.235 ± 0.15) mg/L, and ranged from 0.064 to 2.635 mg/L during the dry period with the average value of (0.476 ± 0.46) mg/L. The SiO4 concentration in the wet period ranged from 0.996 to 19.020 mg/L, with an average of (15.471 ± 2.90) mg/L, and the SiO4 concentration ranged from 0.483 to 17.029 mg/L during the dry period, with an average value of (12.412 ± 3.42) mg/L (Figure 2).

3.2. Seasonal Characteristics of the Total Chl-a Concentration and Particle Size Composition

There were no significant differences (one-way ANOVA, p > 0.05) in the total Chl-a concentration and size composition of phytoplankton during the two periods (Figure 3a). Micro-Chl-a and nano-Chl-a were the main contributors during the dry and wet periods. Nano-Chl-a was the most dominant in both the dry period and wet period, accounting for 59.3% and 66.31%, respectively. Micro-Chl-a was present at more than 35% during the two periods, second to nano-Chl-a. The proportion of pico-Chl-a was 6.28% in the dry period and 5.00% in the wet period (Figure 3b).

3.3. Spatial Characteristics of the Total Chl-a Concentration

The total Chl-a concentration in the midstream was higher than that in the upstream and downstream during the dry season (one-way ANOVA, p < 0.05), and the total Chl-a concentration showed a fluctuating trend from upstream to downstream during the wet period. Specific to the characteristics of each cascade, over 50% of the extreme values of total Chl-a concentration were observed at stations adjacent to the dams (Figure 4).

3.4. Spatial Characteristics of the Percentage Composition of Chl-a Concentration in Different Particle Sizes

The proportion of micro-phytoplankton was higher downstream compared to the midstream and upstream during the dry period (one-way ANOVA, p < 0.05). The proportion of micro-phytoplankton in the midstream was higher than that of the upstream and downstream during the wet period (one-way ANOVA, p < 0.05). On the contrary, the pattern of nano-phytoplankton contrasted that of micro-phytoplankton, with higher amounts in the upstream and downstream and lower amounts in the midstream during the wet period (one-way ANOVA, p < 0.05) (Figure 5).

3.5. Degree of Redundancy Analysis

In the dry period, the relationship between the size-fractionated Chl-a and environmental factors was analyzed using RDA along with 5 key factors: DO, salinity, conductivity, transparency, and pH (Figure 6a). The results of the ANOVA test showed that Axis 1 (p = 0.011) and Axis 2 (p = 0.018) had significant effects on correlation (R2 = 0.27) (Figure 6a). The eigenvalues of Axis 1 and Axis 2 were 0.091 and 0.066, respectively, which could explain 57.37% and 41.79% of the above environmental factors. The results based on Axis 1 show that pico-Chl-a was positively correlated with transparency, and the micro- and nano-Chl-a were distributed in opposite directions to Axis 2. Micro-Chl-a was positively correlated with DO and negatively correlated with conductivity, while nano-Chl-a was the opposite (Figure 6a).
In the wet period, the relationship between the size composition of phytoplankton and environmental factors was also analyzed using RDA along with five key factors: pH, SiO4, TP, width, and transparency (Figure 6b). The results of the ANOVA test showed that Axis 1 (p = 0.001) and Axis 2 (p = 0.207) had significant effects on correlation (R2 = 0.41) (Figure 6b). The eigenvalues of Axis 1 and Axis 2 were 0.582 and 0.071, respectively, which could explain 87.09% and 10.66% of the above environmental factors, respectively. The results based on Axis 1 show that the proportion of micro- and pico-Chl-a was distributed on the opposite side of Axis 1. The proportion of pico-Chl-a was positively correlated with transparency, while the proportion of micro-Chl-a had a negative correlation (Figure 6b).

3.6. Relationship Between Chl-a Particle Size Composition and Distance of Phytoplankton

The Pearson correlation coefficients between size-fractionated Chl-a and distance indicated that the proportion of micro- and nano-phytoplankton was correlated significantly (p < 0.01) with distance during the dry period (Table 1). The proportion of micro-phytoplankton was positively correlated with distance, and the proportion of nano-phytoplankton was negatively correlated. During the wet period, total Chl-a and nano-phytoplankton concentrations were significantly positively correlated (p < 0.01) with distance (Figure 7).

3.7. Predictive Analysis

SOM analysis was conducted based on the similarity of the composition of phytoplankton size in the dry period; two taxa were initially identified and then subdivided into three subgroups (Figure 8a). D1 represents the downstream samples (Figure 8b). This cluster was characterized by a higher total Chl-a concentration and a higher proportion of micro-Chl-a compared to the other two taxa (Figure 8c,d). D2 represents the majority of upstream and midstream samples, and D3 represents the upstream, midstream, and downstream stations (Figure 7b). No significant differences were identified for the total Chl-a concentration and phytoplankton size composition between these two taxa (Figure 8c,d).
According to Cohen’s Kappa test (p < 0.001), the five environmental parameters selected accurately predicted all SOM clusters. Environmental parameters explained 56.7% of the variation in phytoplankton size composition during the dry period. The forecast rates for taxa were 65.2% (D1), 52.4% (D2), and 50% (D3), respectively. D1 and D2 were the most closely related to the first sorting axis (Figure 9a), with pH and conductivity identified as significant influencing factors (Figure 9b). D3 was most closely related to the second sort axis (Figure 9a), with transparency as an important predictor (Figure 9b).
SOM analysis was also conducted based on the similarity of the phytoplankton size composition in the wet period. Two clusters were initially identified and then subdivided into three subgroups (Figure 10a). W1 represents the two midstream cascades of Liucheng and Lankou (Figure 10b), and the total Chl-a concentration and proportion of micro-Chl-a were significantly higher compared to the other two taxa (Figure 10c,d). W2 included the highest number of upstream, midstream, and downstream sites, representing the general characteristics of most stations during the wet period (Figure 10b); finally, W3 mainly included a small number of upstream stations, and this cluster was characterized by a significantly higher percentage of micro-Chl-a than the other two clusters (Figure 10d).
According to Cohen’s Kappa test (p < 0.001), the five environmental parameters selected accurately predicted all SOM clusters. Environmental constraints explain the 80.6% difference in the relative percentage composition of different size-fractionated Chl-a concentrations during the wet period. The clustering prediction rates were 77.8% (W1), 91.3% (W2), and 14.3% (W3), respectively. Clusters W1 and W2 were the most closely related to the first sort axis (Figure 11a), with transparency and total phosphorus identified as significant influencing factors (Figure 11b); by contrast, W3 was most closely related to the second axis (Figure 11a), with pH, river width, and silicate identified as important predictors (Figure 11b).

3.8. Relationship Between Dominant Species and Size-Fractionated Chl-a

The relationship between dominant species and size-fractionated Chl-a is shown in Table 2. Micro-Chl-a and nano-Chl-a were dominant during the two periods. Specifically, regardless of sampling period, Aulacoseira granulata was dominant, with diameters ranging between 2 and 20 µm and higher. Notably, the short-chain-forming diatom Aulacoseira granulata was identified as a key nano-phytoplankton dominant taxon.

3.9. Relationship Between Dominant Taxa and Environmental Factors

The relationship between the biomass of dominant taxa and environmental factors was assessed via RDA (Figure 12). The results of the ANOVA test demonstrated that Axis 1 (p = 0.001) and Axis 2 (p = 0.001) had significant effects on correlation (R2 = 0.34), with eigenvalues of 0.006 and 0.003, respectively. This explains 66.80% and 32.01% of the above environmental factors, respectively. The results based on Axis 1 show that the biomass of MVS was positively correlated with TP and conductivity. The results based on Axis 2 indicate that the biomass of MVS was positively correlated with pH, DO, width, and transparency.

4. Discussion

4.1. Spatio-Temporal Characteristics of the Total Chl-a Concentration

There were no significant differences in total Chl-a concentration between the dry and wet periods (Figure 3). This was probably due to the regulating function of the dams [33]. The cascade dams open to release water during the dry period and close to store water during the wet period, modifying the water level and reducing the seasonal discrepancy of Chl-a concentration. Therefore, its inter-seasonal water level difference was significantly smaller than that of the natural river. Such modifying effects on the water level reduce the seasonal discrepancy of Chl-a concentration [34,35]. Our finding of minimal seasonal variations in total Chl-a aligns with the flow-stabilizing role of dams.
Spatially, the total Chl-a concentration in the midstream was higher than that of the upstream and downstream during the dry period. The total Chl-a concentration showed an increasing trend from upstream to downstream during the wet period (Figure 4). First, the cumulative effect of cascade dams enhances the stagnant time, thereby increasing the Chl-a concentration gradually from upstream to downstream during the wet period [36]. Second, the water temperature is an important factor influencing the total Chl-a concentration. Cascade dams are typically closed to retain water during the wet period, which can significantly influence the water temperature. Research by Kuriqi et al. [37] demonstrated that such water retention may induce hyporheic heat advection by decreasing the water temperature upstream and notably increasing it downstream by up to 5 °C above normal conditions. A study of the Nadong River in South Korea by Kim et al. [38] found that Chl-a concentration increases significantly with increasing water temperature, which is consistent with our results. Kimambo et al. [39] found that the Chl-a concentration in water is positively correlated with temperature and solar radiation, and the results proved that there is a correlation and causal relationship between Chl-a concentration and meteorological parameters such as temperature, solar radiation, and water level.
Variation in nutrients is another factor that influences the distribution of Chl-a concentration during the wet period. Although the size-fractionation results of Chl-a indicate that nano-phytoplankton were the primary contributors to total biomass, an analysis of species composition (Table 2) revealed that Bacillariophyceae (diatoms), particularly Aulacoseira granulata, dominated the community structure. This apparent paradox likely stems from the fact that the cell sizes of these dominant diatom species often fall within the nano-size category; thus, their biomass was accounted for in the nano-sized Chl-a. Consequently, diatoms, especially species with flexibility in size classification, are the key groups driving the dynamics of the phytoplankton community in the East River. As a result, the Chl-a concentration increases from upstream to downstream during the wet period. Ahmed et al. [40] found that Chl-a concentrations located behind Kotri Barrage increase from upstream to downstream and are positively correlated with silicate in the Indus River, resulting in the dominance of the diatom phylum, which agrees with our results. The sharp increase in Chl-a concentration in the midstream during the dry period may be related to its lower total phosphorus-to-total nitrogen ratio (Figure 6). Al-Taani et al. [41] found that fluctuations in Chl-a concentrations correspond to lower TP and TN ratios in the Al-Wehda Dam in Northern Jordan, which is consistent with our results. In contrast to the established literature, this study revealed for the first time that the total concentration of chlorophyll-a exhibits a clear spatially periodic pattern (Figure 4), which is defined as the Cascade Cycle of Serial Discontinuity Concept (CC-SDC) [21,42]. Considering cascade dams as a system, the higher total Chl-a concentration downstream supports the SDC prediction that the negative effect of FSB on total Chl-a concentration during the wet period gradually recovers with increasing river length. This is consistent with the results of Xie et al. [42].
Furthermore, this study primarily focused on longitudinal patterns along the mainstem cascade. It should be noted that the operation of large tributary reservoirs within the East River basin, such as the Xinfengjiang Reservoir, may exert significant external influences on the studied reach. With its longer hydraulic residence time, this reservoir likely fosters stable communities favorable for micro-phytoplankton. During the wet period, regulated releases from the Xinfengjiang Reservoir may introduce substantial pre-developed phytoplankton assemblages and their associated water chemical conditions into the East River mainstem. This input of ‘ecologically conditioned water’ could be an important external driver for the observed patches of high micro-Chl-a concentration in some midstream reservoirs during the wet period (Figure 5b). Future research integrating hydrological operation and water quality data from such large tributary reservoirs is needed to more comprehensively disentangle the cascading processes of material transport and ecological effects within complex reservoir networks.

4.2. Spatio-Temporal Characteristics of Phytoplankton Size Composition

The dominance of nano-phytoplankton in both dry and wet periods highlights their role as an indicator of stabilized flow conditions created by dam regulation (Figure 3). The results show that the spatial difference in nano-phytoplankton is consistent with the SDC prediction. Specifically, water discharge was important to the contribution of nano-phytoplankton. Dams regulate the flow between the two periods by changing the condition of the sluice gate, causing an insignificant difference in nano-Chla between the two periods. This reinforces the utility of this species as an indicator of hydrological modification. The relative proportion of micro-phytoplankton decreased in the wet period. This decline serves as a clear indicator of light limitations in turbid, high-flow conditions. Previous studies have shown a significant positive correlation between the relative proportion of micro-phytoplankton and pH [43]. Therefore, the relative proportion of micro-Chl-a decreased in the wet period. Notably, water transparency is a critical environmental regulator governing micro-phytoplankton dominance. It has been found that the low transparency of water bodies may restrict the light available to micro-phytoplankton. Results indicate that low transparency limits light availability, thus influencing micro-phytoplankton growth and abundance.
The spatial characteristics of different size-fragmented Chl-a demonstrated that the proportion of micro-Chl-a in the dry period had a clear increasing trend from upstream to downstream (Figure 5), related to the DO in the water. The RDA results showed that the proportion of micro-Chl-a was positively correlated with the DO concentration in the water (Figure 6), and the spatial variation in the DO concentration during the dry period reflected a fluctuating upward trend (Figure 2).

4.3. Phytoplankton Size Structure as an Ecological Indicator in Fragmented Rivers

Our findings demonstrate that the composition of phytoplankton size provides a multi-faceted and sensitive indicator of ecological changes in a dam-fragmented river [44]. The consistent dominance of nano-phytoplankton across seasons and locations underscores its adaptability to the moderated flow regimes characteristic of regulated rivers, making it a general indicator of the lentic conditions imposed by dams [45].
Conversely, the spatial and temporal shifts in micro-phytoplankton offer more specific diagnostic value. The increase in the proportion of micro-phytoplankton downstream during the dry period was strongly associated with elevated dissolved oxygen and silicate concentrations [46]. This pattern positions micro-phytoplankton, particularly diatoms like Aulacoseira granulata, as a positive indicator for recovery zones where longer flow paths and potential nutrient processing downstream of dams create favorable growing conditions [47]. The positive correlation between micro-phytoplankton and distance during the dry period (Table 1) further supports the utility of these species in tracing longitudinal recovery patterns as predicted by the SDC [17].
The predictive models (SOM and LDA) quantitatively confirmed the indicator value of phytoplankton structure size. During the dry period, pH and conductivity were the primary discriminators of community clusters, and in the wet period, transparency and total phosphorus were key. The high explanatory power of these models (56.7% in the dry period; 80.6% in the wet period) validates that phytoplankton size fractions encapsulate meaningful environmental information and can predictably respond to the altered state of the river ecosystem due to cascade damming.

4.4. Predictive Analysis of Spatiotemporal Variation

According to the results of the SOM classification for the two seasons, the spatial difference in phytoplankton size composition in the dry period was more significant than in the wet period (Figure 10). The main reason for this was that the interception of the dam in the river reduced the spatial heterogeneity between river sections. Li et al. [6] studied Hanfeng Lake Dam in the upper reaches of the Three Gorges Reservoir and revealed significant decoupling between dam-regulated physicochemical variables and Chl-a concentration. The interception of dams leads to the fragmentation of rivers into several environmentally similar reservoirs, reducing their spatial heterogeneity as a whole [48,49]. Bai et al. [50] studied the Yulin River in the Three Gorges Reservoir Area and found significant differences in CO2 and CH4 concentrations in upstream and downstream water bodies during storage and drainage periods. The upstream CO2 concentration was significantly lower than the downstream concentration during the drainage period of the reservoir area. These phenomena confirm that the reservoir area reduces the spatial heterogeneity of the river. Sabater-Liesa et al. [45] found that the regulation of the Ebro River by dams has resulted in spatially distributed water bodies in the reservoir area replacing spatially heterogeneous river sections as well as reductions in spatial heterogeneity. In addition, Sabater-Liesa et al. [45] found that, although phytoplankton are closely related to environmental variables, the responses of different environmental parameters to river regulation are inconsistent, which matches our results.
Analyzing the characteristics of various species composition after SOM classification, it was observed that the taxa representing most stations in the dry period and the wet period changed (Figure 8d and Figure 10d). Most stations only had one dominant species; Aulacoseira granulata was observed consistently in most stations during both dry and wet periods (Table 2). In addition to Aulacoseira granulate, Anabaena oscillarioides was dominant in the wet period. The analysis indicated that this was mainly due to the water storage of dams during the wet period, which slows down the water flow rate and promotes algal growth [51]. Another study demonstrated that filamentous diatoms grow in highly transparent ponds, ditches, shallow lakes, and slow-flowing streams [52]. It is speculated that the slowdown of the water flow caused by dam interception inhibits the growth of Aulacoseira granulata, which is suited to flowing water. This eventually led to Anabaena oscillarioides mutations during the wet period, which was another dominant species in most stations.
It should be noted that a limitation of this study is the lack of detailed daily hydrological operational data for the cascade reservoir system (e.g., inflow/outflow rates, actual storage changes, and calculated hydraulic residence times for individual reservoirs). Therefore, our inference regarding ‘localized lentic conditions’ during the wet period is based primarily on the known flood retention function of the dams, the observed phytoplankton community shifts (such as the emergence of Anabaena oscillarioides), and correlative environmental parameters. Future studies that integrate precise hydrological dispatch data with ecological monitoring will be able to more quantitatively reveal the fragmentation effects of dam operation on the hydrological regime of the river continuum and its resulting ecological impacts.

5. Conclusions

Our research revealed that the phytoplankton Chl-a concentration exhibited an increasing trend from upstream to downstream sites. Furthermore, the spatial pattern demonstrated periodic characteristics as the number of dams increased. During the dry period, the Chl-a concentration at the FG Dam reached a local minimum followed by restorative growth. However, a second peak failed to materialize, probably resulting from the fact that the current configuration of 13 cascade dams was insufficient to complete the full manifestation of periodic cycles. It is theoretically plausible that subsequent periodic cycles (third and fourth cycles) would likely emerge with an increasing number of dams, thereby completing the cyclical pattern predicted by the modified SDC hypothesis. Cascade dams had a greater impact on phytoplankton during the dry season, which could lead to guidance for opening and releasing gates, water quality testing, and regulation.
Future investigations will systematically characterize phytoplankton response signatures across cascade dam systems through the integration of multiple indicators. Particular focus will be given to validating the dynamics of the Cascade Cycle of Serial Discontinuity Concept (CC-SDC) and deciphering the hydro-ecological coupling mechanisms governing successional patterns of phytoplankton.

Author Contributions

D.S. and J.W. wrote the main manuscript text and prepared Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7, Figure 8, Figure 9, Figure 10, Figure 11 and Figure 12. C.W., J.S. and C.H. revised the manuscript. Q.L. collected the data. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Guangdong Basic and Applied Basic Research Foundation (NO.2021A1515012079 and NO.2019A1515011589), and the Project of Financial Funds of Ministry of Agriculture and Rural Affairs: Investigation of Fishery Resources and Habitat in the Pearl River Basin.

Data Availability Statement

The original contributions presented in this study are included in the article material. Further inquiries can be directed to the corresponding authors.

Acknowledgments

We are very grateful to all staff members of our team for their assistance during the field work.

Conflicts of Interest

The authors declare that there are no competing interests.

References

  1. Alvarez-Fernandez, S.; Riegman, R. Chlorophyll in North Sea Coastal and Offshore Waters Does Not Reflect Long Term Trends of Phytoplankton Biomass. J. Sea Res. 2014, 91, 35–44. [Google Scholar] [CrossRef]
  2. Guerrero, F.; Castro, M.C. Chlorophylla of Size-Fractionated Phytoplankton at a Temporary Hypersaline Lake. Int. J. Salt Lake Res. 1996, 5, 253–260. [Google Scholar] [CrossRef]
  3. Stelmakh, L.; Kovrigina, N.; Gorbunova, T. Phytoplankton Seasonal Dynamics under Conditions of Climate Change and Anthropogenic Pollution in the Western Coastal Waters of the Black Sea (Sevastopol Region). J. Mar. Sci. Eng. 2023, 11, 569. [Google Scholar] [CrossRef]
  4. Aksnes, D.L.; Egge, J.K. A Theoretical Model for Nutrient Uptake in Phytoplankton. Mar. Ecol. Prog. Ser. 1991, 70, 65–72. [Google Scholar] [CrossRef]
  5. Litchman, E.; Klausmeier, C.A. Trait-Based Community Ecology of Phytoplankton. Annu. Rev. Ecol. Evol. Syst. 2008, 39, 615–639. [Google Scholar] [CrossRef]
  6. Liu, H.; Xie, Y.; Qiu, Y.; Wang, L.; Wang, F.; Laws, E.A.; Huang, B. Examining the Size-Specific Photosynthesis-Irradiance Parameters and Relationship with Phytoplankton Types in a Subtropical Marginal Sea. Ecol. Indic. 2021, 130, 108094. [Google Scholar] [CrossRef]
  7. Padisák, J.; Soróczki-Pintér, É.; Rezner, Z. Sinking Properties of Some Phytoplankton Shapes and the Relation of Form Resistance to Morphological Diversity of Plankton—An Experimental Study. Hydrobiologia 2003, 500, 243–257. [Google Scholar] [CrossRef]
  8. Pauly, D.; Christensen, V. Primary Production Required to Sustain Global Fisheries. Nature 1995, 376, 279. [Google Scholar] [CrossRef]
  9. Chitari, R.R.; Anil, A.C. Estimation of Diatom and Dinoflagellate Cell Volumes from Surface Waters of the Northern Indian Ocean. Oceanologia 2017, 59, 389–395. [Google Scholar] [CrossRef]
  10. Flynn, R.F.; Haraguchi, L.; Mcquaid, J.; Burger, J.M.; Lunga, P.M.; Stirnimann, L.; Samanta, S.; Roychoudhury, A.N.; Fawcett, S.E. Nanoplankton: The Dominant Vector for Carbon Export Across the Atlantic Southern Ocean in Spring. Sci. Adv. 2023, 9, eadi3059. [Google Scholar] [CrossRef]
  11. Meyer, M.G.; Brzezinski, M.A.; Cohn, M.R.; Kramer, S.J.; Paul, N.; Sharpe, G.; Niebergall, A.K.; Gifford, S.; Cassar, N.; Marchetti, A. Size-Fractionated Primary Production Dynamics During the Decline Phase of the North Atlantic Spring Bloom. Glob. Biogeochem. Cycles 2024, 38, e2023GB008019. [Google Scholar] [CrossRef]
  12. Yan, W.; Chen, Z.; Zhang, L.; Wang, F.; Zhang, G.; Sun, J. Nanophytoplankton and Microphytoplankton in the Western Tropical Pacific Ocean: Its Community Structure, Cell Size and Carbon Biomass. Front. Mar. Sci. 2023, 10, 1147271. [Google Scholar] [CrossRef]
  13. Hardikar, R.; Ram, A.; Parthipan, V. Distribution of Size-Fractionated Phytoplankton Biomass from the Anthropogenically Stressed Tropical Creek (Thane Creek, India). Reg. Stud. Mar. Sci. 2021, 41, 101577. [Google Scholar] [CrossRef]
  14. Rajaneesh, K.M.; Mitbavkar, S.; Anil, A.C. Dynamics of Size-Fractionated Phytoplankton Biomass in a Monsoonal Estuary: Patterns and Drivers for Seasonal and Spatial Variability. Estuar. Coast. Shelf Sci. 2018, 207, 325–337. [Google Scholar] [CrossRef]
  15. Sun, J.; Li, W.; Wu, N.; Lai, Z.; Du, W.; Jia, H.; Ge, D.; Wang, C. Spatio-Temporal Patterns and Predictions of Size-Fractionated Chlorophyll a in a Large Subtropical River, China. J. Freshw. Ecol. 2020, 35, 1–12. [Google Scholar] [CrossRef]
  16. Wei, J.; Li, W.; Yang, W.; Zeng, Y.; Liu, Q.; Gao, Y.; Li, H.; Wang, C. Phytoplankton Species Richness as an Ecological Indicator in a Subtropical, Human-Regulated, Fragmented River. River Res. Appl. 2023, 39, 718–733. [Google Scholar] [CrossRef]
  17. Ward, J.V.; Stanford, J.A. The Serial Discontinuity Concept: Extending the Model to Floodplain Rivers. Regul. Rivers Res. Manag. 1995, 10, 159–168. [Google Scholar] [CrossRef]
  18. Ward, J.; Stanford, J. The Serial Discontinuity Concept Of Lotic Ecosystems. Dyn. Lotic Ecosyst. 1983, 29–42. [Google Scholar]
  19. Zhang, S.-L.; Feng, P. Serial Discontinuity Concept and Its Development Status in River Ecosystem Research. Adv. Water Sci. 2005, 16, 758–762. [Google Scholar]
  20. Stanford, J.A.; Ward, J.V. Revisiting the Serial Discontinuity Concept. Regul. Rivers Res. Manag. 2001, 17, 303–310. [Google Scholar] [CrossRef]
  21. Ellis, L.E.; Jones, N.E. A Test of the Serial Discontinuity Concept: Longitudinal Trends of Benthic Invertebrates in Regulated and Natural Rivers of Northern Canada. River Res. Appl. 2016, 32, 462–472. [Google Scholar] [CrossRef]
  22. Yasong, T. Variation Analysis Method for Hydrologic Annual Distribution Homogeneity Based on Gini Coefficient. A Case Study of Runoff Series at Longchun Station in Dongjiang River Basin. J. Hydroelectr. Eng. 2012, 31, 7–13. [Google Scholar]
  23. Li, H.; Ge, D.; Yang, W.; Zeng, Y.; Liu, Q.; Gao, Y.; Mai, Y.; Wang, C. Taxa Composition and Ecological Traits of Filamentous Diatoms in a Regulated Fragmented River, China. Fundam. Appl. Limnol. 2022, 195, 305–323. [Google Scholar] [CrossRef]
  24. Jia, H.J.; Lai, Z.N.; Wang, C. Temporal and Spatial Patterns of Phytoplankton Species Richness in the Pearl River Delta. Acta Ecol. Sin. 2019, 39, 3816–3827. [Google Scholar]
  25. Toolbox, S.; Vesanto, J. Neural Network Tool for Data Mining: SOM Toolbox. In Proceedings of the Symposium on Tool Environments and Development Methods for Intelligent Systems (TOOLMET2000), Oulu, Finland, 13–14 April 2000. [Google Scholar]
  26. Ultsch, A. Self-Organizing Neural Networks for Visualisation and Classification. In Information and Classification; Opitz, O., Lausen, B., Klar, R., Eds.; Springer: Berlin/Heidelberg, Germany, 1993; pp. 307–313. [Google Scholar] [CrossRef]
  27. Davies, D.L.; Bouldin, D.W. A Cluster Separation Measure. IEEE Trans. Pattern Anal. Mach. Intell. 1979, PAMI-1, 224–227. [Google Scholar] [CrossRef]
  28. Demuth, H.; Beale, M.; Hagan, M. Neural Network Toolbox, For Use with MATLAB; The MathWorks, Inc.: Natick, MA, USA, 2000; p. 415. [Google Scholar]
  29. Alhoniemi, E.; Himberg, J.; Parhankangas, J.; Vesanto, J. SOM Toolbox. Online Documentation. 2000. Available online: http://www.cis.hut.fi/projects/somtoolbox (accessed on 3 December 2025).
  30. Sneath, P.H.A.; Sokal, R.R. Numerical Taxonomy. Nature 1962, 193, 855–860. [Google Scholar] [CrossRef] [PubMed]
  31. Ihaka, R.; Gentleman, R. R: A Language for Data Analysis and Graphics. J. Comput. Graph. Stat. 1996, 5, 299–314. [Google Scholar] [CrossRef]
  32. Huete-Ortega, M.; Calvo-Díaz, A.; Graña, R.; Mouriño-Carballido, B.; Marañón, E. Effect of Environmental Forcing on the Biomass, Production and Growth Rate of Size-Fractionated Phytoplankton in the Central Atlantic Ocean. J. Mar. Syst. 2011, 88, 203–213. [Google Scholar] [CrossRef]
  33. Poff, N.L.; Olden, J.D.; Merritt, D.M.; Pepin, D.M. Homogenization of Regional River Dynamics by Dams and Global Biodiversity Implications. Proc. Natl. Acad. Sci. USA 2007, 104, 5732–5737. [Google Scholar] [CrossRef]
  34. Pringle, C.M.; Freeman, M.C.; Freeman, B.J. Regional Effects of Hydrologic Alterations on Riverine Macrobiota in the New World: Tropical-Temperate Comparisons: The massive scope of large dams and other hydrologic modifications in the temperate New World has resulted in distinct regional trends of biotic impoverishment. While neotropical rivers have fewer dams and limited data upon which to make regional generalizations, they are ecologically vulnerable to increasing hydropower development and biotic patterns are emerging. BioScience 2000, 50, 807–823. [Google Scholar] [CrossRef]
  35. Carlisle, D.M.; Wolock, D.M.; Meador, M.R. Alteration of Streamflow Magnitudes and Potential Ecological Consequences: A Multiregional Assessment. Front. Ecol. Environ. 2011, 9, 264–270. [Google Scholar] [CrossRef]
  36. Engel, F.; Attermeyer, K.; Ayala, A.I.; Fischer, H.; Kirchesch, V.; Pierson, D.C.; Weyhenmeyer, G.A. Phytoplankton Gross Primary Production Increases along Cascading Impoundments in a Temperate, Low-Discharge River: Insights from High Frequency Water Quality Monitoring. Sci. Rep. 2019, 9, 6701. [Google Scholar] [CrossRef]
  37. Kuriqi, A.; Pinheiro, A.N.; Sordo-Ward, A.; Bejarano, M.D.; Garrote, L. Ecological Impacts of Run-of-River Hydropower Plants—Current Status and Future Prospects on the Brink of Energy Transition. Renew. Sustain. Energy Rev. 2021, 142, 110833. [Google Scholar] [CrossRef]
  38. Kim, K.B.; Jung, M.-K.; Tsang, Y.F.; Kwon, H.-H. Stochastic Modeling of Chlorophyll-a for Probabilistic Assessment and Monitoring of Algae Blooms in the Lower Nakdong River, South Korea. J. Hazard. Mater. 2020, 400, 123066. [Google Scholar] [CrossRef] [PubMed]
  39. Kimambo, O.N.; Chikoore, H.; Gumbo, J.R.; Msagati, T.A.M. Retrospective Analysis of Chlorophyll-a and Its Correlation with Climate and Hydrological Variations in Mindu Dam, Morogoro, Tanzania. Heliyon 2019, 5, e02834. [Google Scholar] [CrossRef] [PubMed]
  40. Ahmed, W.; Wu, Y.; Kidwai, S.; Li, X.; Zhang, G.; Zhang, J. Spatial and Temporal Variations of Nutrients and Chlorophyll a in the Indus River and Its Deltaic Creeks and Coastal Waters (Northwest Indian Ocean, Pakistan). J. Mar. Syst. 2021, 218, 103525. [Google Scholar] [CrossRef]
  41. Al-Taani, A.A. Trend Analysis in Water Quality of Al-Wehda Dam, North of Jordan. Environ. Monit. Assess. 2014, 186, 6223–6239. [Google Scholar] [CrossRef]
  42. Xie, C.; Hu, C.; Yang, W.; Wu, N.; Liu, Q.; Wei, J.; Wang, C. Phytoplankton Species Composition as Bioindicator in the Largest Fragmented Channel of the Pearl River, China. Environ. Monit. Assess. 2024, 196, 389. [Google Scholar] [CrossRef]
  43. Wang, C.; Lai, Z.; Li, Y.; Li, X.; Lek, S.; Hong, Y.; Tan, X.; Li, J. Population Ecology of Aulacoseira Granulata in Xijiang River. Acta Ecol. Sin. 2012, 32, 4793–4802. [Google Scholar] [CrossRef]
  44. de Castro, L.S.; de Souza Lopes, A.A.; Colares, L.; Palheta, L.; de Souza Menezes, M.; Fernandes, L.M.; Dunck, B. Dam Promotes Downriver Functional Homogenization of Phytoplankton in a Transitional River-Reservoir System in Amazon. Limnology 2021, 22, 245–257. [Google Scholar] [CrossRef]
  45. Sabater-Liesa, L.; Ginebreda, A.; Barceló, D. Shifts of Environmental and Phytoplankton Variables in a Regulated River: A Spatial-Driven Analysis. Sci. Total Environ. 2018, 642, 968–978. [Google Scholar] [CrossRef] [PubMed]
  46. Shi, S.; Xu, Y.; Li, W.; Ge, J. Long-Term Response of an Estuarine Ecosystem to Drastic Nutrients Changes in the Changjiang River during the Last 59 Years: A Modeling Perspective. Front. Mar. Sci. 2022, 9, 1012127. [Google Scholar] [CrossRef]
  47. Mohanty, T.R.; Tiwari, N.K.; Kumari, S.; Ray, A.; Manna, R.K.; Bayen, S.; Roy, S.; Das Gupta, S.; Ramteke, M.H.; Swain, H.S.; et al. Variation of Aulacoseira Granulata as an Eco-Pollution Indicator in Subtropical Large River Ganga in India: A Multivariate Analytical Approach. Environ. Sci. Pollut. Res. 2022, 29, 37498–37512. [Google Scholar] [CrossRef]
  48. Parasiewicz, P.; Belka, K.; Łapińska, M.; Ławniczak, K.; Prus, P.; Adamczyk, M.; Buras, P.; Szlakowski, J.; Kaczkowski, Z.; Krauze, K.; et al. Over 200,000 Kilometers of Free-Flowing River Habitat in Europe Is Altered Due to Impoundments. Nat. Commun. 2023, 14, 6289. [Google Scholar] [CrossRef] [PubMed]
  49. Anderson, E.P.; Jenkins, C.N.; Heilpern, S.; Maldonado-Ocampo, J.A.; Carvajal-Vallejos, F.M.; Encalada, A.C.; Francisco Rivadeneira, J.; Hidalgo, M.; Canas, C.M.; Ortega, H.; et al. Fragmentation of Andes-to-Amazon Connectivity by Hydropower Dams. Sci. Adv. 2018, 4, eaao1642. [Google Scholar] [CrossRef]
  50. Bai, X.; He, Q.; Li, H.; Xu, Q.; Cheng, C. Response of CO2 and CH4 Transport to Damming: A Case Study of Yulin River in the Three Gorges Reservoir, China. Environ. Res. 2022, 208, 112733. [Google Scholar] [CrossRef]
  51. Ha, K.; Jang, M.H.; Joo, G.J. Spatial and Temporal Dynamics of Phytoplankton Communities along a Regulated River System, the Nakdong River, Korea. Hydrobiologia 2002, 470, 235–245. [Google Scholar] [CrossRef]
  52. Wang, C.; Wu, N.; Li, W.; Liu, Q.; Lai, Z.; Fohrer, N. Curved Filaments of Aulacoseira Complex as Ecological Indicators in the Pearl River, China. Ecol. Indic. 2020, 118, 106722. [Google Scholar] [CrossRef]
Water 17 03513 g001
Figure 1. The study area and sampling locations: (a) study area; (b) dam locations along the East River’s main channel; and (c) sampling sites situated between the connecting dams. A and B represent two adjacent cascade dams. The first station of each dam represents its reservoir area, while the river reach between the two dams has 3 to 6 sampling stations set up at intervals.
Figure 1. The study area and sampling locations: (a) study area; (b) dam locations along the East River’s main channel; and (c) sampling sites situated between the connecting dams. A and B represent two adjacent cascade dams. The first station of each dam represents its reservoir area, while the river reach between the two dams has 3 to 6 sampling stations set up at intervals.
Water 17 03513 g001
Water 17 03513 g002
Figure 2. The temporal and spatial characteristics of environmental factors.
Figure 2. The temporal and spatial characteristics of environmental factors.
Water 17 03513 g002
Water 17 03513 g003
Figure 3. The seasonal characteristics of (a) the total and (b) the concentration percentage of Chl-a.
Figure 3. The seasonal characteristics of (a) the total and (b) the concentration percentage of Chl-a.
Water 17 03513 g003
Water 17 03513 g004
Figure 4. The spatial characteristics of the total Chl-a concentration.
Figure 4. The spatial characteristics of the total Chl-a concentration.
Water 17 03513 g004
Water 17 03513 g005
Figure 5. The spatial characteristics of the Chl-a concentration percentage in the (a) dry period and (b) wet period.
Figure 5. The spatial characteristics of the Chl-a concentration percentage in the (a) dry period and (b) wet period.
Water 17 03513 g005
Water 17 03513 g006
Figure 6. The relationship between the proportions of size-fractionated Chl-a and key environmental parameters based on redundancy analysis in (a) dry and (b) wet period.
Figure 6. The relationship between the proportions of size-fractionated Chl-a and key environmental parameters based on redundancy analysis in (a) dry and (b) wet period.
Water 17 03513 g006
Water 17 03513 g007
Figure 7. Relationship between Chl-a size composition and distance. ((a,b) indicate the dry period; (c,d) indicate the wet period).
Figure 7. Relationship between Chl-a size composition and distance. ((a,b) indicate the dry period; (c,d) indicate the wet period).
Water 17 03513 g007
Water 17 03513 g008
Figure 8. SOM according to the proportion of size-fractionated Chl-a in the dry period. (a) The relationship between clusters. (b) The characteristics for each cluster. (c) Total Chl-a concentration of each cluster. (d) The proportion of size-fractionated Chl-a for each cluster.
Figure 8. SOM according to the proportion of size-fractionated Chl-a in the dry period. (a) The relationship between clusters. (b) The characteristics for each cluster. (c) Total Chl-a concentration of each cluster. (d) The proportion of size-fractionated Chl-a for each cluster.
Water 17 03513 g008
Water 17 03513 g009
Figure 9. LDA for SOM groups (D1, 2, and 3) and environmental variables in the dry period. (a) The distribution and overlap of clusters in the F1 and F2 dimensions. (b) The distribution of water quality parameter variation related to F1 and F2.
Figure 9. LDA for SOM groups (D1, 2, and 3) and environmental variables in the dry period. (a) The distribution and overlap of clusters in the F1 and F2 dimensions. (b) The distribution of water quality parameter variation related to F1 and F2.
Water 17 03513 g009
Water 17 03513 g010
Figure 10. SOM according to the proportion of size-fractionated Chl-a during the wet period. (a) The relationship between clusters. (b) The characteristics for each cluster. (c) Total Chl-a concentration of each cluster. (d) The proportion of size-fractionated Chl-a for each cluster.
Figure 10. SOM according to the proportion of size-fractionated Chl-a during the wet period. (a) The relationship between clusters. (b) The characteristics for each cluster. (c) Total Chl-a concentration of each cluster. (d) The proportion of size-fractionated Chl-a for each cluster.
Water 17 03513 g010
Water 17 03513 g011
Figure 11. LDA for SOM groups (W1, 2, and 3) and environmental variables during the wet period. (a) The distribution and overlap of clusters in the F1 and F2 dimensions. (b) The distribution of water quality parameter variation related to F1 and F2.
Figure 11. LDA for SOM groups (W1, 2, and 3) and environmental variables during the wet period. (a) The distribution and overlap of clusters in the F1 and F2 dimensions. (b) The distribution of water quality parameter variation related to F1 and F2.
Water 17 03513 g011
Water 17 03513 g012
Figure 12. Relationship between the biomass of dominant taxa and environmental factors based on redundancy analysis. AGA: Aulacoseira granulata; AFA: Aulacoseira fennoscandica; MVS: Melosira varians; CMA: Cyclotella meneghiniana; AOS: Anabaena oscillarioides.
Figure 12. Relationship between the biomass of dominant taxa and environmental factors based on redundancy analysis. AGA: Aulacoseira granulata; AFA: Aulacoseira fennoscandica; MVS: Melosira varians; CMA: Cyclotella meneghiniana; AOS: Anabaena oscillarioides.
Water 17 03513 g012
Table 1. Pearson correlation coefficient between distance and factors.
Table 1. Pearson correlation coefficient between distance and factors.
Distance
Total concentrationdry period0.112
wet period0.546 **
Micro concentrationdry period0.192
wet period0.181
Nano concentrationdry period0.024
wet period0.666 **
Pico concentrationdry period−0.057
wet period0.087
Micro proportiondry period0.468 **
wet period−0.027
Nano proportiondry period−0.357 **
wet period0.119
Pico proportiondry period−0.228
wet period−0.219
Note(s): “**” represents p < 0.01.
Table 2. Relationship between dominant species and size-fractionated Chl-a.
Table 2. Relationship between dominant species and size-fractionated Chl-a.
TimeSize-Fractionated Chl-aCategoryAlgal TaxaBiomass Proportion
2018.11nano- and micro-Chl-aBacillariophyceaeAulacoseira granulate (AGA)66%
micro-Chl-aBacillariophyceaeAulacoseira fennoscandica (AFA)1%
nano- and micro-Chl-aBacillariophyceaeMelosira varians (MVS)8%
nano- and micro-Chl-aBacillariophyceaeCyclotella meneghiniana (CMA)1%
2019.7nano- and micro-Chl-aBacillariophyceaeAulacoseira granulata25%
micro-Chl-aCyanophytaAnabaena oscillarioides (AOS)11%
nano- and micro-Chl-aBacillariophyceaeCyclotella meneghiniana1%
nano- and micro-Chl-aBacillariophyceaeMelosira varians8%
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Sang, D.; Wei, J.; Hu, C.; Liu, Q.; Sun, J.; Wang, C. Phytoplankton Size as an Ecological Bioindicator in a Subtropical Fragmented River, China. Water 2025, 17, 3513. https://doi.org/10.3390/w17243513

AMA Style

Sang D, Wei J, Hu C, Liu Q, Sun J, Wang C. Phytoplankton Size as an Ecological Bioindicator in a Subtropical Fragmented River, China. Water. 2025; 17(24):3513. https://doi.org/10.3390/w17243513

Chicago/Turabian Style

Sang, Deyu, Jingxin Wei, Caiqin Hu, Qianfu Liu, Jinhui Sun, and Chao Wang. 2025. "Phytoplankton Size as an Ecological Bioindicator in a Subtropical Fragmented River, China" Water 17, no. 24: 3513. https://doi.org/10.3390/w17243513

APA Style

Sang, D., Wei, J., Hu, C., Liu, Q., Sun, J., & Wang, C. (2025). Phytoplankton Size as an Ecological Bioindicator in a Subtropical Fragmented River, China. Water, 17(24), 3513. https://doi.org/10.3390/w17243513

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.
Back to TopTop