A case study on environmental sustainability: A study of the trophic changes in fish species as a result of the damming of rivers through clustering analysis

• Published in: Computers & industrial engineering Vol. 135; pp. 1239 - 1252
• Main Authors: Almeida, Ricardo de, Steiner, Maria Teresinha Arns, Coelho, Leandro dos Santos, Francisco, Cláudia Aparecida Cavalheiro, Steiner Neto, Pedro José
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.09.2019
• Subjects: Clustering analysis; Environment sustainability; Reservoir phase; River phase; Trophic categories of fish; River phase; Environment sustainability; Clustering analysis; Reservoir phase; Trophic categories of fish
• ISSN: 0360-8352; 1879-0550
• DOI: 10.1016/j.cie.2018.09.032

•Knowledge of fish assemblage constitutes excellent bio-indicators.•Analysis of the trophic changes in species of fish.•We analyze the data prior and after to the damming.•It was used Clustering Analysis in order to classify the species of fish. The damming of rivers has been long used for electricity generation and is among the most used sources of renewable energy. However, building dams may cause several transformations in the environment, being changes in fish assemblage one important consequence, especially when there are communities that rely on fishing as a source of income. The aim of the present study is to analyze the trophic changes in fish species caused by the damming of rivers. Trophic data (stomach content) on fish from the Corumbá Reservoir in the State of Goiás, Brazil, which was collected prior (River phase) and after (Reservoir phase) the building of the dam, were used to carry out the study using Clustering techniques. The methodology used was composed of data exploratory analysis, followed by the assignment of clusters for the later implementation of knowledge. The definition of the number of clusters, the usage of different types of clustering distances and the use validation indexes are discussed. A modified version of the Teitz & Bart algorithm, originally used for facilities location problems, was introduced for Clustering problems and the results were compared with three well-known Clustering algorithms from literature. The clustering approaches were applied separately in both phases and in both cases, five large clusters of fish were determined: generalists, insectivores, herbivores, piscivores, and detritivores. With this evaluation, could be used by biologists in order to evaluate environmental effects and managers can develop strategies to address the social and economic impacts caused to the communities that depend on fishing.