Abstract
There are great impacts on the streams in western Taiwan（and these impacts are） due to excessive anthropogenic disturbances. Many Permanent monitoring stations were set for detecting the changes of water quality, but the scale was limited to points and could not scale up to line (stream section) and area (drainage). We could not develop reasonable stream management without considering biological and ecological sustainabilities. It’s necessary for us to establish biological indicators. The purpose of this research is to classify stream landscape and establish biological indicators for detecting water pollution in streams. In the paper, we use water quality data and biological data of Chungkan River, Kehyea River and Nankann River during Sep. 2002 to Sep. 2003. Water temperature, dissolved oxygen, biochemical oxygen demand, chemical oxygen demand, ammonia nitrogen, phosphorus, chloride, suspended solids, total organic carbon, heavy metals, pH, coliform, organic chlorine and turbidity were included in water quality data. Biological data were inclusive of algaes, benthos, crabs, shrimps, fish and birds of three streams. Integrating the landscape factors with Schumm stream classification method, we could classify a stream into several longitudinal zones. On the basis of classifying zones, chemio-physical water quality indicators –RPI and biological indicators could be applied, and then the seriousness of polluted streams showed. Finally, we get a further conclusion about correlations between pollution and species community. We suggest that the classification system of stream and biological indicator could be used in the streams of Taiwan.