The application of Bayesian quantile regression to analyse the relationship between nutrients content and phytoplankton abundance in Sutami reservoir

Phytoplankton plays a significant role in aquatic ecosystem as the main feed for another aquatic biota. However, the abundance of phytoplankton must be controlled. This is due to the exaggeration of phytoplankton abundance can lead to eutrophication and the mass mortalities of fish. One of the dependent factors for phytoplankton abundance is the nutrients (nitrate and phosphate) in the relevant aquatic ecosystem. This study aims to analyze the relationship between nutrients and phytoplankton abundance, also to determine the maximum limit of the nutrient content in order to prevent eutrophication in Sutami Reservoir. The relationship usually analyzed by simple linear regression. Unfortunately, the data of phytoplankton abundance and nutrients content in Sutami Reservoir contains outlier according to Cook's distance criteria. It means that simple linear regression cannot be used. Thus, the alternative method which is Bayesian quantile regression considered. The result of analysis indicates that the parameter value of regression model between nutrients content and phytoplankton abundance is varying, which are depended on the analyzed quantile.