Do the key factors determining phytoplankton growth change with water level in China's largest freshwater lake?

The interpretation of environmental effect on phytoplankton growth is not straightforward, especially in ecosystems with high variation in hydrological conditions due to confounding variables. Lake Poyang is characterized by high fluctuation of water level (WL) due to its free connection to the Yangtze River. High frequency (weekly) and long-term (2009-2018) samplings were conducted in Lake Poyang resulting in 170 samples. Our study aimed to illustrate that whether the key factors determining phytoplankton growth change with WL. Furthermore, the impact of WL on phytoplankton was also detected. Six periods were classified in our study, i.e., dry season, raising period (I and II), wet season, and falling period (I and II) based on WL variation. Environmental characteristics were significantly (P < 0.01) different among these periods with the exception of orthophosphate. Based on the whole data set, water temperature (WT) was the critical parameter affecting phytoplankton growth. However, according to time series analysis, the key factors varied in different periods. Underwater light condition, which was represented by Secchi depth (SD), was the most critical factor controlling phytoplankton growth, especially in the periods with relatively high WL and WT (i.e., raising period II, wet season, and falling period I). The role of water temperature on phytoplankton was more evident in falling periods. In dry season with the lowest WL, total phosphorus limited phytoplankton growth. Regarding WL, its impact on phytoplankton was mainly through change in environmental parameters, such as water flow velocity, water transparency, and nutrients. Furthermore, time series analysis well simulated phytoplankton chlorophyll a in Lake Poyang, with larger R-adj(2) (except raising period II and falling period II) and lower model error in all 6 periods. Our results revealed that the critical factors controlling phytoplankton growth were various with WL. Additionally, time series analysis will benefit local water resource management.