Quantitative Relationship between Chlorophyll a and the Key Controlling Factors in Four Plateau Lakes in Yunnan Province, China
ABSTRACT: Most lakes in China are impaired by eutrophication; therefore it is necessary to take effective measures to restore the water quality. Plateau lakes in Yunnan Province, China are unique due to their distinct natural characteristics. Different dynamic processes contribute to the eutrophication of each lake, making it impossible to implement the same management method to all the lakes. Therefore, it is essential to identify the key driving factors for lake eutrophication to support more reliable and effective decision making. Previously, many water quality management and restoration projects have been implemented to address the water quality problems in lakes in Yunnan Province; however, the achieved water quality improvement is far from expected environmental objectives. Therefore, conducting comparative research among the plateau lakes in Yunnan Province not only can identify the difference in the driving factors and dynamic processes between these plateau lakes, but also can find some similar key processes in eutrophic mechanism. In this study, four typical plateau lakes were selected, including Lake Dianchi, Lake Chenghai, Lake Fuxian and Lake Yilong. An integrated approach of absolute principle components score-multivariate linear regression (APCS-MLR) and structural equation modeling (SEM) method were developed in this study to understand the influence of water chemistry variables on chlorophyll a (Chl a). The SEM results were further validated with the artificial neural networks (ANN). The model results demonstrated that among all factors, the physical and chemical conditions in the lakes have the greatest influence on Chl a. However, different physical and chemical factors make different contribution to the Chl a between the four lakes. The comparative analysis also showed that lake morphology, watershed population and industrial composition, and aquatic ecosystem variations would have great influence on lake eutrophication.
KEYWORDS: Plateau Lakes; Eutrophication; Structural Equation Modeling; Artificial Neural Networks; Absolute Principle Components Score-Multivariate Linear Regression