Physical habitat simulation for a fish community using the ANFIS method

Dongkyun Im, Sung Uk Choi, Byungwoong Choi

Research output: Contribution to journalArticlepeer-review

18 Citations (Scopus)


This study presents physical habitat simulations for the fish community in a reach of the Dal River, Korea. The study reach is 2.3 km long, located downstream from the Goesan Dam. The reach is gravel-bed, including a bend. A riffle is present at the apex of the bend, and pools are located before and after the riffle. Field monitoring revealed that five fish species are dominant, namely Zacco platypus, Coreoleuciscus splendidus, Zacco temminckii, Pungfungia herzi, and Acheilognathus yamatsute, and account for 78% of the total fish community. These five fishes, which include both lotic and lentic fishes, were selected as target species. The Center for Computational Hydroscience and Engineering 2D (CCHE2D) model, a 2D shallow water equation solver, and the Adaptive Neuro Fuzzy Inference System (ANFIS) method were used for hydraulic and habitat simulations, respectively. The Mahalanobis distance method was used for identifying outliers in the monitoring data. It was shown that the ANFIS method combined with data quality assessment predicts the Composite Suitability Index (CSI) better than the method without it. The CSI distributions in the study reach are presented for various flows. It was found that the distributions of the combined CSI predicted by the ANFIS method are higher than those predicted by the conventional multiplicative aggregation method with Habitat Suitability Curves (HSCs). In addition, it was revealed that the discharge that yields the maximum Weighted Usable Area (WUA) for the whole fish community is more than twice the discharge yielding the maximum WUA for a single target species. This increased discharge leads to a 58% increase in the maximum WUA for the fish community.

Original languageEnglish
Pages (from-to)73-83
Number of pages11
JournalEcological Informatics
Publication statusPublished - 2018 Jan

Bibliographical note

Publisher Copyright:
© 2017 Elsevier B.V.

All Science Journal Classification (ASJC) codes

  • Ecology, Evolution, Behavior and Systematics
  • Ecology
  • Modelling and Simulation
  • Ecological Modelling
  • Computer Science Applications
  • Computational Theory and Mathematics
  • Applied Mathematics


Dive into the research topics of 'Physical habitat simulation for a fish community using the ANFIS method'. Together they form a unique fingerprint.

Cite this