Controlling Sustainability of Aquatic Ecosystems:Application of Genetic Programming

Maintaining a harmonic balance between both human and ecosystem water demands is one of the ultimate goals of water resources managers. The first step to meet that goal is to construct a quantitative connection between ecology and water quantity-quality model. In this study, a data mining approach: Genetic Programming (GP) was utilized to generate a quantitative relationship between an ecological target index based on benthic macroinvertebrates and physicochemical water quality indicators. The dataset used in this study included seasonal water quality parameters and number of benthic macroinvertebrates per square meter of Jajrood River located in North-East of Tehran, Iran, from September 2011 to May 2013. Among different forms of equations obtained from GP, the one with the most compatibility of the computed versus observed ecological index was selected. Using GUI-HDMR software, a variance based sensitivity analysis was implemented in order to rank the importance of input parameters and to explore the influence of parameter interactions. This approach can help water resources managers to improve the downstream benthic macroinvertebrates habitat as a reflection of ecosystem health just based on the available information of physicochemical water quality parameters.

Keywords:

Jajrood River , Ecosystem Health , Genetic Programming , Benthic Macroinvertebrates