Evaluation of effect of sediment on aquatic ecosystems using decision tree forest and group method of data handling

Abstract

In this study, the performances of the decision tree forest and group method of data handling for evaluation scale of the severity (SEV) of ill effect for fishes were investigated. The independent variables were concentration of suspended sediment (SS), species, life stage, and duration of exposure. This study is based on 198 data of aquatic ecosystem quality over a wide range of sediment concentrations (1–500,000 mg SS/L) and durations of exposure (1–35,000 h). Results showed that exposure duration is the most important factor on SEV, and based on the results, this alternative approach is better than traditional regression models with a higher recognition rate, forecast accuracy, and strong practical value.