Succession of phytoplankton functional groups in Macau’s two shallow urban border reservoirs under multiple changing factors

The community diversity of phytoplankton is quite high, and hundreds of phytoplankton species can coexist. Aggregating phytoplankton species into different phytoplankton functional groups (PFGs) can better assess and predict the potential structural and functional effects of aquatic ecosystems. The continuous interaction of key PFGs is the key cause of cyanobacterial blooms in drinking water supply systems, especially in shallow border reservoirs, but the reasons have not been understood. This study was carried out in Macau’s two shallow border reservoirs, the Main Storage Reservoir (MSR) and Sai Pa Van Reservoir (SPVR). Then, phytoplankton samples from these reservoirs were collected monthly over a four-year period (from January 2011 to December 2014) to reveal key factors driving PFGs succession and their synthetic effects on the dominant species. The trophic state index (TSI) of TN and TP of the two reservoirs ranged from 36.25 to 83.95 and 8.32–56.13, respectively, indicating that the SPVR and MSR were between a eutrophic and hypereutrophic state. The phytoplankton analysis results showed that there were 14 dominant PFGs and 8 dominant PFGs in the MSR and SPVR, respectively. The group S1 was always predominant in the SPVR during the succession of PFGs, while the succession of PFGs in the MSR were quite different, the PFGs could mainly be summarized as S1/SN/X1 (2011) → X1/J/B/TB/P (2012, 2013) →S1/X1/J (2014). In addition, redundancy analysis (RDA) through stepwise discriminant analysis was used to analyze the dominant PFGs and abiotic factors in both reservoirs. The results indicated that the turbidity value of the reservoir was the key factor influencing groups S1, Lo and SN in the SPVR. However, the different PFGs were influenced by the turbidity, NH3–N, NO3–N and EC values of the reservoir in the MSR. The relationship between PFGs and environmental factors was described by stepwise regression analysis. The investigation of the different ecological responses between the interconnected reservoirs was useful for targeted water source protection.