Abstract:
It is valuable for the control of water quality in lakes to grasp the global functional relationship between phytoplankton Biomass (Chlorophyll-a; Chl-a), total phosphorous (TP) and total nitrogen (TN) in lake ecosystems. This paper aims at developing A comprehensive model that explains the relationship between Chl-a, TP and TN in lakes under a wide range of\ Environments. Three regression models, the conventional Ordinary Least Squares (OLS) model, random effect panel model and dynamic panel model are compared. Estimation based on water quality data from 396 lakes in 18 European countries reveals that OLS estimation gives comparable parameters to those of many earlier studies, in which both TP and TN are significant Determinants of Chl-a. Application of the non-conventional estimation methods alters this parameter structure radically. Station-specific effects being controlled, TN/TP is not a significant factor. The inclusion of auto-regressive effects makes TN insignificant. These results suggest that the concentration of Chl-a could be controlled by reducing the concentration Of TP, rather than TN, the effect of which might however be offset by the past concentration of Chl-a. These models were tested by applying them to simulate the relationship for 9 Japanese lakes to show the superior performance of the non-conventional dynamic model.
