The phytoplankton community can be affected by multiple environmental factors such as climate, meteorology, hydrology, nutrients, and grazing. The complex interactive effects of these environmental factors as well as the resilience of phytoplankton communities further make the prediction of phytoplankton communities’ dynamics challenging. In this study, we analyzed multiple environmental factors and their relative importance in predicting both phytoplankton shifting and cyanobacteria abundance in two regulated shallow lakes in central China. Our results indicated that the phytoplankton community in the study areas could be mainly classified into 1. Cryptophyta dominated group, 2. Biologically diverse group, and 3. Cyanobacteria dominated group. The Multinomial Logistic Regression model indicated the Cryptophyta dominated group was sensitive to temperature, while other groups were sensitive to both temperature and nutrients. The interactive effects of temperature and nutrients were synergistic in the cyanobacteria dominated group, while they were antagonistic or minor in other groups. The Negative Binomial Regression model suggested high total phosphorus and low total nitrogen but not temperature were responsible for high cyanobacteria abundance. The conditional plot indicated nutrients affected cyanobacteria abundance more significantly under low wind speeds and lake volume fluctuations, and cyanobacteria abundance in the cyanobacteria dominated group maintained high levels with increasing hydrological dynamics. Our results demonstrated that environmental factors played inconsistently significant roles in different phytoplankton groups, and reducing nutrients could decrease adverse effects of warming and water project constructions. Our models can also be applied to forecast phytoplankton shifting and cyanobacteria abundance in the management of regulated shallow lakes.