Seasonal variation in phytoplankton community assembly in hydrologically isolated urban lakes

As primary producers, phytoplankton are key indicators of lake health and temporal ecosystem dynamics. Understanding their community structure and assembly mechanisms is therefore essential for effective lake management. This study examined the seasonal dynamics, assembly processes and environmental drivers of phytoplankton communities in eight urban lakes in Dongxihu District, Wuhan, based on quarterly sampling from 2022 to 2023. A total of 202 phytoplankton species and varieties from eight phyla were identified. Phytoplankton density showed clear seasonal variation, peaking in summer (mean: 15.71 × 106 cells L−1; range: 3.05–24.57 × 106 cells L−1) and reaching its lowest levels in winter. Cyanobacteria dominated throughout the study. Beta-diversity partitioning revealed that community differences among lakes were driven primarily by species turnover, which accounted for more than 90% of total beta-diversity in all seasons. Checkerboard (C-score) analysis suggested that community assembly in summer, autumn and winter was more consistent with stochastic structuring, whereas stronger environmental filtering was evident in spring. Mantel analysis revealed clear seasonal differences in environmental drivers. In spring, total phosphorus (TP) and Secchi depth (SD) were significantly correlated with β-diversity indices (P < 0.05). In contrast, in summer, autumn and winter, the permanganate index (CODMn) showed significant correlations with β-diversity (P < 0.05). NMDS combined with envfit (stress < 0.10) further indicated that environmental gradients significantly structured species abundance patterns. CODMn was the dominant factor in summer and winter, whereas TP and SD were predominant in spring. Overall, these findings indicate that phytoplankton assembly in these hydrologically isolated urban lakes is shaped by seasonally variable mechanisms, with factors such as organic pollution (reflected by CODMn) and nutrient availability (TP) modulating community structure. Notably, deterministic environmental filtering surpassed stochasticity only in spring.