Differences in seasonal dynamics, ecological driving factors and assembly mechanisms of planktonic and periphytic algae in the highly urban Fenhe River

Abstract

Background

Algae play important roles in urban river ecosystems and are the cornerstones of most water quality monitoring programs. Thus, a better understanding of algal community dynamics is needed to support sustainable management of water resources in urban rivers.

Results

In this study, we quantified the seasonal variations in planktonic and periphytic algal community structure in the highly urban Fenhe River and identified environmental factors affecting algal community structure and diversity. We monitored planktonic (drifting) and periphytic (attached) algal communities in the Taiyuan section of the Fenhe River over one year. The results indicated that Cyanophyta was the dominant phylum in both communities, followed by Bacillariophyta and Chlorophyta. Significant differences were observed in the composition of the planktonic and periphytic algal communities. In particular, the periphytic algal community was more diverse than the planktonic community. Water temperature and pH were the main environmental factors affecting planktonic and periphytic algal community structure, respectively, while nutrients were the most significant factor affecting planktonic and periphytic algal diversity. Ecological modeling indicated that the variations in the algal communities of the Fenhe River are mainly driven by stochastic processes. A co-occurrence network developed for the communities displayed positive interactions between the planktonic and periphytic algae.

Conclusions

These findings deepen our understanding of the seasonal interaction between planktonic and periphytic algae and the driving factors affecting community structure in the Fenhe River. They also provide a theoretical basis for the managing and protecting water resources in urban river ecosystems.

Background

Algae are an integral part of global biogeochemical cycles and are the main primary producers in river ecosystems (Litchman et al. 2015). Riverine freshwater algae include planktonic (i.e., drifting “phytoplankton”) and periphytic species (i.e., attached to surfaces) (Song et al. 2016). Phytoplankton inhabit the water column and are susceptible to wind and water flow, whereas periphyton attach and grow on substrates, including bottom sediments, aquatic macrophytes, organic litter (e.g., leaf packs), and woody structures (e.g., fallen trees or tree roots) (Huang et al. 2023). Despite that the two algal types occur in different physical habitats, planktonic and periphytic algae coexist in nearly all river ecosystems, being interrelated and competitive with each other while also existing in dynamic equilibrium (Wang et al. 2019; Hu et al. 2022a).

Rivers are significant systems for water cycling, material transport, and social and economic development (Liu et al. 2020). However, urban river environments are gradually deteriorating with rapid urbanization and industrialization. Flows are decreasing in many systems, and rivers are concomitantly experiencing rapid losses of biodiversity and ecosystem integrity (Yang et al. 2021a). Previous studies have shown that the diversity and structure of riverine algal communities vary seasonally and are affected by environmental factors, such as water temperature, light intensity, and water quality (Chen et al. 2023). It has been reported that nitrogen and phosphorus have been identified as the primary factors shaping the variations in algae in urban rivers (Wang et al. 2024). Additionally, studies have shown that the patterns of spatiotemporal variation in algal communities are related to latitude and longitude. For example, Wu et al. (2023) reported that planktonic algal communities are more affected by geographical factors than periphytic algae in interconnected river-lake systems. In addition, the responses of planktonic and periphytic algae to seasonal changes may differ (Huang et al. 2023). While planktonic algae inhabit the water column, they have some capacity to avoid long-term exposure to adverse environments in that they are easily transported to other areas by winds, waves, and river flows. In contrast, periphytic algae are attached to substrates and essentially sessile; thus, they are unable to avoid stressful conditions (Song et al. 2016; Zhao et al. 2022). Previous studies have rarely compared the community structure and diversity patterns of planktonic and periphytic algae within the same river system.

Elucidating the assembly processes of algal communities has always been a central theme in freshwater ecology (Huber et al. 2020; Huang et al. 2023). Two categories of mechanisms—termed “deterministic” and “stochastic”—have proven to be important in driving the assembly of algal communities (Niu et al. 2024). Deterministic processes are often predictable and result from the effects of environmental and/or biological interactions on different species or communities. On the other hand, stochastic processes are often unpredictable and result from natural disasters, random migration, or human activities. Stochasticity also assumes that the species in a community are equivalent and independent of any environmental characteristics that might serve as drivers of community structure (Zhou and Ning 2017). The importance of these two types of processes varies depending on the habitat type and the actual ecological driving factors (Liu et al. 2023). Although previous studies have considered both deterministic and stochastic processes in the context of community structure, the assembly mechanisms regarding algal communities in river environments remain unclear.

Another less studied issue is the species co-existence mechanism, which identifies the potential associations or interactions among taxa through network-based analyses (Liu et al. 2022). Co-occurrence networks have been widely used to understand the relationships among microorganisms (Zhang et al. 2024). The topological structural characteristics of co-occurrence networks shed light on the complexity and stability of these networks (Freilich et al. 2018). At the same time, key groups exhibiting high connectivity within a network play a vital role in maintaining stability of the network (Berry and Widder 2014). However, little is known about the interaction between freshwater algae and the mechanisms of their network stability in river environments.

The Fenhe River is located in Shanxi Province, China, and plays a key role in providing water resources and absorbing urban sewage for the local area (Feng et al. 2021). The river exhibits seasonal variations in this reach, thus, providing an ideal location for examining seasonal differences in algal communities (Yang et al. 2021a). In this study, we monitored planktonic and periphytic algal communities in the Taiyuan City section of the Fenhe River over 1 year. We examined seasonal changes in the algal community structure and identified the main factors driving the variation. Our main questions included: (1) How does algal community structure and diversity vary across seasons? (2) What are the main environmental factors affecting changes in algal community structure and diversity? (3) Are the assembly mechanisms of algal communities driven by stochastic or deterministic factors? and (4) How does the complexity and stability of algal community networks vary seasonally? Results of this study will shed light on the environmental factors that tend to drive algal community structure. Additionally, the formation of algal communities will be studied, thus, providing a scientific basis for managing and restoring urban river ecosystems.

Methods

Study area

The Fenhe River is a highly urban river ecosystem located in Taiyuan City, Shanxi Province in northern China (latitude: 37° 27′–38° 25′N, longitude: 111° 30′–113° 09′E). The region experiences a semi-humid continental monsoon climate with an average annual precipitation of 539 mm. This region of Shanxi Province is known for coal production and various industries (e.g., production of iron, steel, and chemicals), and contains a large urban residential area. Taiyuan City is densely populated and highly urbanized, with a population of about 5.3 million. The urban section of the Fenhe River in Taiyuan City is 43 km long, flowing from north to south through the city. The river receives domestic sewage and industrial wastewater discharged along most of its banks. In this study, 12 sampling sites were selected along the Taiyuan City section of the Fenhe River (Fig. 1). Each site was sampled eight times during the spring (April and May 2023), summer (July and August 2023), autumn (October and November 2023), and winter (December 2023 and January 2024).

Fig. 1

Location of the study area and the sampling sites along the Fenhe River in Taiyuan City

Sample collection

Planktonic algae was sampled from 1-L water samples collected at a water depth of 0.5-m at each sampling point using a standard water collector; the samples were immediately fixed and preserved in 15% Lugol’s solution. All samples were sent to the laboratory and settled for 48 h. A standard glass slide (25.4 × 76.2 mm) served as the artificial substrate for colonizing the periphytic algae samples. The depth of the substrate sampler (20 standard slides were fixed on the slide holder and made into a diatom meter) was usually 20–30 cm, with at least two substrates placed at each sampling point for more than 30 days to ensure colonization of algae and a successful sample collection. After 2–4 weeks of cultivation, the artificial substrates were assumed to represent the “living” algal community on the natural substrate (Vlaičević et al. 2021). Three slides were randomly collected from each location on each sampling date. The slides were scrubbed with a nylon brush into a known water volume and immediately fixed in 15% Lugol’s solution.

Planktonic and periphytic algae were counted using a visual field method under an optical microscope (Olympus BX51, Olympus Corporation, Tokyo). “The freshwater algae of China: systematics, taxonomy and ecology” (Hu and Wei 2006) was used as the main reference to identify the specimens.

Planktonic algae cell density was calculated as follows:

$$N = \frac{{{A}}}{{{{Ac}}}} \times \frac{{{V}}}{{{{Va}}}} \times {{n}}$$

where N is the number of planktonic algae per liter of water (cells/L), Ac is the counting area (mm²), A is the area of the counting frame (400 mm²), Va is the counting frame volume (0.1 mL), V is the volume of water remaining after concentrating of each liter of water sample (30 mL), and n is the planktonic algae cell number obtained from counting. Periphytic algae cell density was calculated as follows:

$$N = {{\left( {V \times n} \right)} \mathord{\left/ {\vphantom {{\left( {V \times n} \right)} {\left( {Va \times S} \right)}}} \right. \kern-0pt} {\left( {Va \times S} \right)}}$$

where N is the number of periphytic algae per unit area (cells/cm²), V is the volume of the sample concentrated for quantitative analysis (mL), n is the number of periphytic algae cells determined from counting, Va is the actual sample water volume for quantitative analysis (mL), and S is the total area of the collected samples (cm²).

Measurement of water quality parameters

Water temperature (WT), conductivity (SPC), pH, water pressure (mmHg), salinity (SAL), total dissolved solids (TDS), oxidation reduction potential (ORP), and dissolved oxygen (DO) were measured using a YSI portable water quality monitor during on-site sampling (Yellow Springs Instruments, Yellow Springs, Ohio, USA). Water transparency (SD) was measured using a Seville transparency disc. Water samples were collected 0.5-m below the water surface at each monitoring point and collected three times in parallel. The 1-L water samples were collected in pre-cleaned polyethylene bottles, transported to the laboratory within 24 h, and immediately stored at 4 °C. According to the Environmental Quality Standards for Surface Water in China (GB3838-2002), all water quality parameters were measured within 48 h (Hu et al. 2022b). These measures included total nitrogen (TN), total phosphorus (TP), chemical oxygen demand (COD), ammonium nitrogen (${\text{NH}}_{4}^{ + }$–N), phosphate (${\text{PO}}_{4}^{3 - }$–P), nitrate nitrogen (${\text{NO}}_{3}^{ - }$–N), and dissolved organic carbon (DOC).

Statistical analysis

Box plots were used to analyze seasonal changes in the environmental variables, with Origin software (version 2018) and to draw the algal abundance and algal density during the seasons. Circos plots were used to plot the seasonal variations in planktonic and periphytic algae at the phylum level (https://www.omicstudio.cn/tool/). Using the online Galaxy workflow framework (https://huttenhower.sph.harvard.edu/galaxy/), LEfSe (LDA effect size) was used to analyze the biomarkers with season-specific significant differences.

The α-diversity of the algal communities was expressed using the Shannon–Wiener diversity index, the Pielou evenness index, and the Margalef richness index. These indices were computed using the “vegan” package in R (version 4.1.2). At the same time, SPSS (version 20.0) was used to conduct one-way analyses of variance (ANOVA) to compare index differences. To assess the similarity of the algal communities among the different groups, non-metric multidimensional scaling (NMDS) and similarity analysis (ANOSIM) were performed using the “ade 4” and “ggplot 2” packages in R (version 4.1.2). To explore the algae-environment relationships, Spearman’s correlation analysis, the Mantel test, and Partial Least Squares Path Modeling (PLS-PM) were conducted using the R packages “pheatmap”, “linkET”, and “plspm”, respectively.

We used the Bray–Curtis distance measure to analyze the attenuation of algal community similarity with geographical distance (Ma et al. 2023). Variation partitioning analysis (VPA) was used to calculate the interpretation rate of geographical distance and environmental screening of the algal community structures. Geographical distance was calculated based on the latitude and longitude of each sampling point to obtain the principal coordinates of the neighbor matrix (PCNM) variables (Ma et al. 2023). Niche breadth of the algal communities was calculated using the R package “spaa”, with the C-score value calculated using the R package “EcoSimR”. The assembly mechanism of the algal communities was analyzed by the normalized stochasticity ratio (NST) model (Ning et al. 2019; Zhou and Ning 2017) and the neutral community model (NCM) (Sloan et al. 2006; Rosindell et al. 2011).

A co-occurrence network of the algal communities was constructed using the Gephi program (version 0.9.2) based on the Spearman’s correlation coefficients between the relative abundances of the genera. All paired Spearman’s correlation coefficients were calculated using the “psych” package in R (version 4.1.2). In general, if Spearman’s correlation coefficient (r) between two genera was >|0.6|, P values were < 0.05. The correlation coefficients from genera pairs meeting this significance criteria were imported into Gephi and visualized using a Fruchterman–Reingold layout (Fruchterman and Reingold 1991). The topological characteristics of the network (e.g., node number, edge number, average degree, degree, and closeness centrality) were calculated using the Network Analyzer plugin in Gephi. The co-occurrence network was composed of nodes and links, where nodes represent species or environmental variables, and links are the edges connecting two nodes that represent positive and negative correlations between species. Nodes with a greater degree and high closeness centrality were defined as network hubs (Xiong et al. 2021). In this study, algae genera with higher degrees (> 25) and higher closeness centralities (> 0.4) were identified as “hub species” in the co-occurrence network following previous studies by Yang et al. (2022) and Xiong et al. (2021).

Results

Environmental factors

Figure S1 shows the seasonal variations in the water quality parameters. In general, there were significant (P < 0.05) seasonal differences in all water physicochemical parameters except TN (P = 0.16). Across the year, WT ranged from 1.4 to 29.9 °C, with the highest average WT recorded during summer (28.55 °C). ORP was the highest in spring (163.35 mV) and lowest in winter (86.69 mV). DO values ranged between 6.23 and 11.12 mg/L, with the highest average recorded during winter (9.22 mg/L). The average SPC, TDS, and SALvalues were 1.23 ms/cm, 0.80 g/L, and 0.62 ppt, respectively, and were highest during autumn. The annual variation in pH ranged from 7.76 to 9.55, indicating largely alkaline water, and the highest average value (8.79) occurred in autumn. SD values ranged from 10–150 cm, with the highest mean value recorded during winter (88.32 cm). Average concentrations of TN, ${\text{NH}}_{4}^{ + }$–N, TP, and ${\text{PO}}_{4}^{3 - }$–P were 3.02 mg/L, 0.42 mg/L, 0.06 mg/L, and 0.03 mg/L, respectively, and were highest during the summer. The average concentrations of ${\text{NO}}_{3}^{ - }$–N was highest during spring (0.53 mg/L) and lowest during winter (0.04 mg/L). The average concentrations of COD was highest during the autumn (28.55 mg/L) and lowest during the winter (16.11 mg/L), whereas the mean DOC concentration was highest in the spring (5.83 mg/L) and lowest in autumn (5.26 mg/L).

Community structure of the planktonic and periphytic algae

In terms of all algae taxa, a total of 7 phyla and 94 genera were collected from the Fenhe River during this study. Regardless of whether planktonic or periphytic algae were considered, Cyanophyta was the dominant phylum. As for planktonic algae (Fig. 2a), Cyanophyta accounted for 69.0% of the total abundance followed by Chlorophyta (18.0%) and Bacillariophyta (10.2%). Among periphytic algae (Fig. 2d), Cyanophyta accounted for 42.7% of the total relative abundance followed by Bacillariophyta (36.7%) and Chlorophyta (19.3%).

Fig. 2

Planktonic algal community composition (a, b), cell density (c) and periphytic algal community composition (d, e), and cell density (f) during different seasons. LEfSe analysis of the planktonic (g) and periphytic algae (h) during different seasons

On a seasonal basis, Chlorophyta dominated the planktonic algae during the spring (38.8%) and winter (37.9%), and Cyanophyta dominated during summer (54.1%) and autumn (85.3%) (Fig. 2b). Among the periphytic algae, Bacillariophyta was dominant (Fig. 2e) during spring (65.3%), summer (39.9%), and winter (48.9%), whereas Cyanophyta dominated during autumn (64.1%). The density of planktonic and periphytic algal cells was greater in autumn than during other seasons, mainly because the density of cyanobacteria had reached its annual maximum at that time. The planktonic algal cell densities levels were in the order of in autumn (46.94 × 10⁸ cells/L) > summer (25.11 × 10⁸ cells/L) > spring (57.60 × 10⁷ cells/L) > winter (18.91 × 10⁷ cells/L) (Fig. 2c). However, periphytic algae exhibited the opposite pattern. In particular, periphytic algal cell densities were in the order of autumn (26.54 × 10⁶ cells/cm²) > winter (17.35 × 10⁶ cells/cm²) > summer (12.54 × 10⁶ cells/cm²) > spring (7.50 × 10⁶ cells/cm²) (Fig. 2f). Winter cell densities of periphytic algae were greater than those during summer and spring, largely due to the higher density of diatoms (8.49 × 10⁶ cells/cm²), which exceeded the spring (4.90 × 10⁶ cells/cm²) and summer (5.00 × 10⁶ cells/cm²) densities by more than 60%.

Among the dominant genera of planktonic and periphytic algae, the ten most abundant genera belonged to Bacillariophyta and Cyanophyta (Fig. S2). At the same time, LEfSe analyses indicated there were 2–5 unique algal groups during each season, with potential biomarkers identified for each season (i.e., LDA score > 4.0) (Fig. 2). The planktonic algae groups were dominated by Synedra in spring, whereas the periphytic algae were dominated by Navicula. During summer, planktonic and periphytic algae were both dominated by Mougeotia, whereas Pseudanabaena dominated both algal communities during the autumn. During winter, planktonic algae were dominated by Chlamydomonas, whereas periphytic algae were dominated by Synedra.

Diversity of the planktonic and periphytic algae

Values of the Shannon–Wiener diversity index, Pielou evenness index, and Margalef richness index were used to evaluate α-diversity in the algal communities across different seasons and habitats (Fig. 3). Results indicated that the Shannon–Wiener diversity values for planktonic algae were greatest during winter while those of periphytic algae were greatest during summer. Similarly, the Pielou evenness values was the highest in winter for planktonic algae and in spring for periphytic algae. The Margalef richness index of planktonic algae was highest in winter and lowest in summer, while that of periphytic algae was highest in autumn and lowest in spring. One-way ANOVA indicated that the α-diversity of planktonic and periphytic algal communities was significantly different among seasons (P < 0.05). Although the α-diversity of the planktonic and periphytic algal communities varied seasonally, values for the periphytic algal community were generally greater than those for planktonic algae (Fig. S3a).

Fig. 3

Alpha diversity analysis of planktonic (a) and periphytic algae (b) across seasons. The NMDS using the Bray–Curtis distance shows the β-diversity of planktonic and periphytic algal communities during the seasons (c). Different letters indicate significant differences between the seasons (P < 0.05)

NMDS analyses suggested significant seasonal differences in the composition of the planktonic and periphytic algal communities (ANOSIM: r = 0.650, P = 0.001; Fig. 3c). The differences between the planktonic and periphytic algae communities were significant during all seasons. In particular, the differences between the planktonic and periphytic communities occurred during the winter (r = 0.888, P = 0.001), spring (r = 0.935, P = 0.001), summer (r = 0.300, P = 0.001), and autumn (r = 0.273, P = 0.001); despite their significance, the lower correlations during summer and autumn indicated more similar communities during those seasons (Fig. S3b). In addition, although seasonal changes in the planktonic and periphytic algae communities were fairly obvious, seasonal changes in the planktonic community (r = 0.703, P = 0.001) tended to be greater than seasonal changes in the periphytic algae community (r = 0.376, P = 0.001). The overall results suggested that there were consistent seasonal differences between the planktonic and periphytic communities in the Fenhe River.

Effects of environmental variables on algal community structure

The relationships between the environmental variables and algal community structure were best illustrated by the Mantel test results (Fig. 4a). Overall, the planktonic algal community was significantly associated with many of the environmental factors, including WT, mmHg, SPC, TDS, SAL, pH, ORP, ${\text{NO}}_{3}^{ - }$–N, ${\text{PO}}_{4}^{3 - }$–P, COD and DOC (Table S1). The periphytic algal community was associated with WT, mmHg, pH, ${\text{NO}}_{3}^{ - }$–N, ${\text{PO}}_{4}^{3 - }$–P and DOC (Table S1). Overall, WT (r = 0.477, P < 0.01) and pH (r = 0.240, P < 0.01) were the main driving factors affecting the planktonic and periphytic algae community structures. In addition, different degrees of association were detected among the water quality parameters. Spearman’s correlations analysis was used to identify the main environmental factors that affected the composition of the algal communities at the phylum level (Fig. S4). WT was most associated with Bacillariophyta, Chlorophyta, Euglenophyta, and Pyrrophyta in the planktonic algae community (P < 0.001), while pH was most associated with Bacillariophyta and Cyanophyta (P < 0.05) within the periphytic algae community.

Fig. 4

a Mantel test demonstrating the correlations between the environmental factors and algal community structure and the correlations between the water quality parameters. b PLS-PM revealed the direct effects of the environmental variables on algal diversity. The blue and orange lines indicated positive and negative effects, respectively (Water properties: pH, DO; Nutrients: TN, TP, ${\text{NO}}_{3}^{ - }$–N, ${\text{NH}}_{4}^{ + }$–N, ${\text{PO}}_{4}^{3 - }$–P; Geographical distance: latitude and longitude; planktonic algae and periphytic algae: planktonic and periphytic algal diversity indices). The goodness-of-fit (GoF) was 0.501 (significance: *P < 0.05, **P < 0.01, ***P < 0.001)

PLS-PM modeling based on WT explained 31% and 67% of the variance in the planktonic and periphytic algal diversities, respectively (Fig. 4b). The environmental factors WT, other water properties, nutrients, and geographical distance were significant factors affecting periphytic algal community diversity (P < 0.001). Nutrients were the most significant factor affecting planktonic algal community diversity (− 0.363, P < 0.05). Similarly, the direct effect of WT on nutrients was 0.354 (P < 0.01), which suggested that WT may also indirectly affect algal diversity by changing the physicochemical properties of the water. The direct and indirect effects of various environmental factors on the communities were different (Fig. S5), which affected the algae diversity. Therefore, the combined effects of environmental factors, such as WT, nutrients, water properties, and geographical distance, led to changes in algal diversity. Among them, nutrients were the most significant environmental factor that affected the diversity of both the planktonic and periphytic algae.

Community assembly mechanisms of the planktonic and periphytic algae

The distance-decay relationship (DDR) model (Fig. 5a) suggested that the similarity in the planktonic and periphytic algal communities was significantly negatively correlated with geographical distance (P < 0.001). This finding indicates that the distribution of the algal communities was affected by diffusion restrictions. VPA analysis further indicated that a series of environmental and spatial variables affected the temporal and spatial dynamics of the planktonic and periphytic algal communities (Table 1). The environmental variables played a more critical role in driving variations in the periphytic community during spring and summer (spring: 34.1%; summer: 21.7%) compared to the planktonic community (spring: 16.2%; summer: 11.9%). Spatial variables played a more pivotal role driving the variation in periphytic community during summer and winter (summer: 31.5%; winter: 12.3%) than in planktonic community (summer: 0.4%; winter: 5.0%). However, 18.2–57.3% of this variation could not be explained by environmental or spatial processes. Thus, stochastic processes may have been responsible for much of the variation in the algal communities given the large amount of unexplained variation from the VPA.

Fig. 5

a The distance attenuation relationship and geographical distance between planktonic and periphytic algae. The role of stochastic and deterministic mechanisms in shaping the algal communities was examined using the (b) normalized stochasticity ratio (NST) model, c niche breadth, d the neutral community model (NCM), and e C-scores for planktonic and periphytic algae

Table 1 Effects of environmental and spatial processes on planktonic and periphytic algal communities across different seasons based on variation partitioning analysis (VPA)

The average NST values of planktonic and periphytic algae were 0.961 and 0.853, respectively (Fig. 5b). Both values exceeded the 0.5 threshold, which indicated that the planktonic and periphytic algal communities were regulated more by stochastic processes. Thus, the NST modeling results were verified using NCM mode and niche breadth. NCM modeling successfully predicted the effect of stochastic processes on algal community assembly (Fig. 5d), with the results suggesting that the diffusion limitation was more pronounced in the periphytic algal community than in the planktonic algal community (r² values for planktonic and periphytic algal communities were 0.11 and 0.46, respectively). The periphytic algal community assembly also contained a larger proportional from stochastic processes compared to the planktonic algal community, which explained 11.2% and 46% of the community variance, respectively. The niche breadth (Fig. 5c) values for periphytic algae (8.10) was significantly greater than that for planktonic algae (6.12). This finding suggested that the periphytic algal community had more resources to use, and thus stronger competitive abilities.

The null model based on the C-score was employed in this study (Fig. 5e). A significant difference was detected between C-score_obs and C-score_sim in the planktonic and periphytic algae communities (P < 0.001). The C-score_obs (planktonic: 40.43; periphytic: 30.86) were greater than the mean value under the zero model (C-score_sim: planktonic: 38.79; periphytic: 29.63), indicating that the symbiotic pattern of planktonic and periphytic algae community structure was non-random. The standardized effect size (SES) value was significantly greater than zero for the planktonic and periphytic algae communities, indicating that the relative contribution of deterministic processes was more important for planktonic algae assembly.

Variable network patterns of the algal communities during the seasons

We established co-occurrence networks to visualize the interactions between planktonic and periphytic algae during the seasons (Fig. 6) and listed the relevant topological characteristics in Table S2. There were 97 junctions and 452 edges (92.0% positive, 8.0% negative) in spring, 107 nodes and 545 edges (92.8% positive, 7.3% negative) in summer, 116 nodes and 777 edges (75.0% positive, 25.0% negative) in autumn, and 115 nodes and 743 edges (87.9% positive, 12.1% negative) in winter. These results indicated that the co-evolution between planktonic and periphytic algae across seasons was the main driving force, with algal communities having more potential for interaction during autumn. Cyanobacteria, Chlorophyta, and Bacillariophyta contained the largest number of nodes in each season. The topological features (e.g., nodes, edges, graph density, and clustering coefficients) in the co-occurrence network reflect community stability (Freilich et al. 2018). The ratio of graph density and the clustering coefficient (D/CC) also indicate network stability. The D/CC value in autumn (0.213) was higher than that in other seasons (winter: 0.191, spring: 0.156, summer: 0.148), indicating that the algal community structure in autumn was more stable and resistant to external environmental pressures. Similarly, the D/CC value of periphytic algae (0.197) was greater than that of planktonic algae (0.143), indicating that the structure of the periphytic algal community was more stable (Table S2).

Fig. 6

a Network analysis demonstrating the interaction between planktonic and periphytic algae across the seasons. The red and green edges indicate positive and negative correlations, respectively. Each node represents a unique genus and was colored by the corresponding phylum. b Comparison of node-level topological features (degree and closeness centrality) among the seasons

Network analysis was used to identify key classification groups within complex network structures. We further defined “hub species” as genera with greater values of degree (> 25) and closeness centrality (> 0.4) within the network (Fig. 6b). Overall, a total of 33 genera were classified as hub species (Table S3). In spring, the hub species limited to Scenedesmus, which is a genus of periphytic algae. During summer, Cyclotella (periphytic), and Cosmarium and Scenedesmus (planktonic) were hub species. Among the hub species in autumn, the genera with the highest degrees and closeness centrality values were the planktonic algae Pseudanabaena and Chroococcus, while the hub species during winter were the periphytic algae Crucigenia and Pseudanabaena (Table S3).

Discussion

Seasonal variations in planktonic and periphytic algae

Algae are a key component mediating the eco-chemical cycles and the stability of aquatic ecosystems (Wu et al. 2023). Therefore, understanding of the seasonal dynamics of different algal communities in urban rivers is essential. Our results demonstrated that the community structures of planktonic and periphytic algae in the highly urbanized Taiyuan City section of the Fenhe River varied significantly across seasons. First, the changes can be explained, in part, by variations in environmental factors and hydrological conditions (Pan et al. 2022; Huang et al. 2023). The environmental variables in the Taiyuan City section of the Fenhe River exhibited significant seasonal differences for all factors (except TN), which may have directly affected the algae community dynamics. Secondly, interactions between algal species, including cooperation and competition mechanisms, likely augmented the seasonal changes (Zhang et al. 2022). Importantly, the seasonal variation in the planktonic algal community was greater than that of the periphytic algal community. This variation may have been related to the different habits of the planktonic and periphytic algal forms (Song et al. 2016). For instance, the attached periphytic algal communities living on stable substrates may have reached equilibrium following periods of long-term erosion and sediment deposition (Wang et al. 2019). Thus, seasonal differences in the periphytic algae community were minor by comparison to planktonic algae and not as obvious.

Overall, Cyanophyta was the dominant phylum, followed by Bacillariophyta and Chlorophyta, in the planktonic and periphytic algal communities. This was consistent with previous results from rivers, whereby Cyanophyta was frequently the dominant phylum (Huang et al. 2023). However, the proportion of planktonic and periphytic algae with the dominant phylum differed seasonally. Previous studies have reported that cyanobacteria dominate during summer and autumn, with the seasonal succession of algal taxa significantly affected by environmental change (Wei et al. 2023). Similarly, diatoms occur in cool and turbulent environments and are more adaptable to nutrient deficiencies than cyanobacteria or green algae (Zhang et al. 2023), particularly during the winter and early spring. The Shannon–Wiener diversity and Pielou evenness values for planktonic and periphytic algae were lowest in autumn, which may have been related to the dominance of a single taxa at that time of year. In this instance, Pseudoanabaena was dominant in the planktonic and periphytic algal communities during autumn (Fig. S2). Sudden and dramatic increases in the abundance of a single algae species can lead to displacement of other species occupying the same living space, and thus, inducing competition for shared environmental resources (Zhang et al. 2024).

Environmental factors affecting algal community structure and diversity

The correlations between changes in algal community structure and diversity, and variations in environmental factors, have been demonstrated in many studies (Zhang et al. 2023, 2024). In this study, different degrees of correlation were observed between the water quality parameters, which may have affected the planktonic and periphytic algae in a synergistic manner. We found that WT and pH were the main factors affecting planktonic and periphytic algae community structure, respectively. Additionally, nutrients were the most significant environmental factor that jointly affected the diversity of planktonic and periphytic algae. The effects of water temperature and nutrients on the algal communities and their diversity were quite pronounced. For example, elevated temperatures typically increase photosynthesis, respiration, and phytoplankton community growth and reproduction (Righetti et al. 2019). Water-borne nutrients, such as nitrogen and phosphorus, are the key driving forces for algal growth (Zhang et al. 2024). Studies have indicated that the structure and diversity of algal communities change with concentrations of nitrogen and phosphorus (Shevchenko et al. 2018). On the one hand, increasing temperatures can strengthen thermal stratification in urban rivers. This characteristic subsequently leads to hypoxia in lower water layers, which accelerates nutrient release into the water column (Farrell et al. 2020; Anderson et al. 2021). However, the change in WT also significantly changes the dynamics related to nitrogen-fixing bacteria, thus, affecting the form and content of nitrogen in the river water (Yang et al. 2021b). The ability of algae to absorb inorganic nutrients from the water and convert them into organic nutrients promotes algal growth (Savichtcheva et al. 2015). The absorption of inorganic carbon by algae during photosynthesis increases the pH of the water, while the CO₂ released by respiration reduces the pH. Studies have shown that weak alkaline environments promote photosynthesis by algae, whereas acidic environments reduce algal growth rates (Jakobsen et al. 2015). In addition, the pH of the water can regulate the outflow of nutrients into sediments and the transformation of nutrients in water. A change in pH alters the form of phosphorus by affecting the precipitation of metals that phosphorus binds (Saha et al. 2022). For instance, Peng et al. (2021) reported that acidic conditions promoted the release of ${\text{NH}}_{4}^{ + }$–N into water sediments, whereas alkaline conditions promoted the conversion of ${\text{NH}}_{4}^{ + }$–N to ${\text{NO}}_{3}^{ - }$–N.

Ecological model analyzing algae assembly mechanisms

Elucidating the driving mechanisms of algal community assembly potentially provides a new concept to be considered in the management of urban rivers. The similarities in the planktonic and periphytic algal communities were significantly negatively correlated with geographical distance based on the DDR model. This finding suggests that the tenets of algal community assembly include multiple processes, such as dispersal limitations and environmental selection (e.g., Wang et al. 2019; He et al. 2022). We further analyzed the effect of environmental and spatial variables on algae community assembly using VPA. In general, the environmental and spatial variables had different seasonal effects on the assembly processes of the planktonic and periphytic algae, with 18–57% of the variation unexplained by environmental or spatial processes. One possible explanation is that we did not measure all environmental and spatial variables that potentially affect algal communities (e.g., no meteorological variables were quantified) (Chen et al. 2019; Wu et al. 2022). However, another explanation is that interspecific relationships may have affected the community dynamics, and the VPA did not quantitatively explain the effect of these interactions (Mo et al. 2021). In the end, the VPA results led us to conclude that stochastic processes had more pronounced effects on the assembly of the algal communities. Additionally, NST modeling, the neutral model, and the niche analysis strongly supported this conclusion, with diffusion limitations being more pronounced in the periphytic algal community.

Other studies have reported comparable results to those reported herein. Sun et al. (2020) concluded that benthic diatoms are more susceptible to a dispersal limitation than planktonic diatoms in the Lancang River, China. This may have been because whereas water flow can rapidly move planktonic algae, attached periphytic algae are sessile and more difficult to move. Thus, a diffusion limitation may have had greater impacts on the community distribution of periphytic algae (Wetzel et al. 2012). Alternatively, greater levels of urbanization (e.g., small dam construction) can lead to reduced river flow rates and deepened river channels. These factors would limit periphytic algae throughout the water body, thus, further limiting the spread of periphytic forms (Wang et al. 2019). At the same time, periphytic algae with a wider niche breadths was more resistant to environmental change, while planktonic algae with a narrower niche breadths was more likely to be dispersed through environmental filtration and interactions between species (He et al. 2022).

Co-occurrence patterns between planktonic and periphytic algae

Co-occurrence networks have been widely used to explore the interactions between microbial communities to identify key taxa and elucidate intricate relationships in a given habitat (Zamkovaya et al. 2021). Our results suggested that the number of algae on the positive edge of different seasons exceeded the negative edge, which indicated that the main interaction between planktonic and periphytic algae in the Taiyuan City section of the Fenhe River was cooperation or of mutual benefit. Similarly, the topological characteristics reflected the stability of the community within the co-occurrence network. In our study, the network structural stability of periphytic algae was greater than that of planktonic algae, suggesting (in this case) that the algal community structure was more stable. The availability of environmental resources was the main factor affecting the structure and stability of the ecological networks (Dai et al. 2022; Huang et al. 2023). Our results indicated that planktonic algae were more susceptible to environmental change. In addition, the niche breadth of the periphytic algal community was wider (Fig. 5c), indicating that the periphytic algal community was more adaptable to environmental change than the planktonic algal community. As a result, the periphytic algae networks were more resistant to seasonal changes (Huang et al. 2023). In the co-occurrence network, the algal community structure in autumn had more edges and nodes, which may have been due to more between-species interrelationships playing a role in buffering environmental interference, thus, increasing their adaptability to harsh environments (Xue et al. 2018). The water COD content was higher during autumn, indicating that the organic pollution levels in the water were higher, and thus, the algal communities could have been more affected (Guo et al. 2017).

The existence of key groups within any community is essential for stability of a network (Coux et al. 2016). Our research indicated that different algal hub species were present during different seasons, and the interactions of the algae differed across seasons. More hub species were observed in the autumn and winter networks, which included Pseudanabaena, Chroococcus, and Crucigenia. Previous studies have shown that hub species contribute to network stability (Fang et al. 2023) and that more hub species in autumn and winter may promote network stability. It is possible that the key groups serving as the core hub of the network only exist in specific environments (Chen et al. 2023).

Conclusions

We studied the seasonal variations in planktonic and periphytic algal community structure in the highly urbanized Fenhe River, China. Seasonal differences were observed in the community structure and diversity of the planktonic and periphytic algae. Additionally, the effects of environmental factors on the structure and diversity of the algae communities were assessed along with the assembly mechanisms and co-occurrence patterns within the algal communities. Notably, the planktonic and periphytic algal communities in the Taiyuan section of the Fenhe River conformed to the distance-decay relationship model, and the dispersal limitation dominated the community assembly of the planktonic and periphytic algae communities. Results indicated that there were significant seasonal differences in all but one parameter during the survey period. Regardless of whether planktonic or periphytic algae, Cyanophyta was always the dominant phylum, though the proportions of Cyanophyta and other common groups (e.g., Chlorophyta and Bacillariophyta) varied seasonally. The periphytic algal community was consistently more diverse than the planktonic algal community. WT and pH were the main factors affecting the structure of the planktonic and periphytic algae communities, respectively. Nutrients were the most significant environmental factor that affected the diversity of the planktonic and periphytic algae. Variations in the algal communities were regulated by stochastic processes. A dispersal limitation was more obvious within the periphytic algal community than the planktonic algae community. The periphytic algae niche was larger than that of planktonic algae, with mutualism being the main interaction between algal communities. The periphytic algae network was more stable during seasonal change than the planktonic algae community. The results of this study will be helpful in understanding the responses of algal communities to seasonal changes while providing a scientific basis for the management and protection of urban river aquatic ecosystems.