Ecological informatics approach to screening of integrity metrics based on benthic macroinvertebrates in streams

¹ Department of Biological Sciences, Pusan National University, Busan 609-735, Republic of Korea
² Department of Biology, Kyung Hee University, Seoul 130-701, Republic of Korea
³ Water Environment Research Department, The National Institute of Environmental Research, Incheon 404-170, Republic of Korea

^* Corresponding author: tschon@pusan.ac.kr

Received: 30 August 2010
Accepted: 6 March 2011

Abstract

Benthic macroinvertebrates are considered as a representative taxon that indicates the ecological status of freshwater ecosystems. Numerous indices derived from community data have been proposed to estimate either biological water quality or ecosystem health. In this study, metrics based on benthic macroinvertebrates at the family level were screened using ecological informatics to provide a multi-metric measurement that would be suitable for presenting ecological integrity across different levels of environmental impact. Benthic macroinvertebrates were collected at a total of 720 sample sites from river basins and streams in Korea in 2009. Based on four categories of community status (i.e., diversity, richness, tolerance, and composition), 37 metrics were selected as initial candidates according to the literature. The candidate metrics were evaluated according to parameters including discriminatory power, redundancy, and responsiveness to stressors. Self-organizing map was utilized to assist the screening procedure by providing ordination, clustering, and visualization of metric and environmental data. Six metrics were finally selected as a multi-metric and were compared with conventional indicators for presenting the ecological integrity of streams.