Eutrophication of a raw water supply reservoir in Macau, namely Main Storage Reservoir (MSR), was analyzed in this study. The reservoir is a pumped storage reservoir that receives water from the West River of the Pearl River network 20 km away. In recent years, there were reports that the reservoir experienced some problems of algal bloom, and the situation appeared to be worsening. To understand the eutrophication process of the reservoir, totally 18 monthly water parameters including 5 hydrological parameters (precipitation, imported volume, exported volume, water level and hydraulic retention time (HRT)), 4 physical parameters (temperature, pH, turbidity and conductivity), 7 chemical parameters (dissolved oxygen (DO), Ammonium, nitrite, nitrate, total nitrogen (TN), orthophosphate (PO4 3- ) and total phosphorus (TP)), and 2 biological parameters (phytoplankton abundance and chlorophyll-a) were sampled and monitored in 2010. The correlation analysis was conducted to determine the dominant parameters that were correlated with the phytoplankton population. Principle component analysis (PCA) followed by multiple linear regression (MLR) stepwise method was used to simplify the complexity of the relationships among 10 selected parameters, and to predict phytoplankton abundance level using the selected principle component scores (PCS) of the principle components (PCs) with eigen values greater than 1, as independent variables

The results showed that the trophic state indices (TSIs) of MSR were estimated as 58-72 for the whole year, indicating that the reservoir was in eutrophication – hypereutrophic status. There was a seasonal variation on eutrophication level with one peak of the phytoplankton population in the late summer, suggesting that temperature was an important factor that causing the bloom. The phytoplankton abundance was at low level (< 40 million cells/L) from January to June, and dramatically increased to 120 millions cells/L in July, with the linear regression coefficient (R2 , p<0.01) of 0.75 against temperature. Anti-correlations were found between phytoplankton abundance and TN (R2 =0.94, p<0.01), nitrate (R2 =0.94, p<0.01), TN/TP ratio (R2 =0.88, p<0.01) or conductivity (R2 =0.97, p<0.01), implying that the nutrients were consumed in the rapid growth of phytoplankton abundance during the bloom. Combined with the weak correlation of phytoplankton population (R2 <0.21) to TP, It was highly probably that TN, instead of TP, is the cause of blooming. The PCA showed that only three PCs with eigen value over 1 can explain 84.6% of total variation of the water parameters and the MLR model was successful in predicting phytoplankton abundance in the reservoir with the predictive power (R2 , p<0.01) of 0.90.