Chironomid assemblages collected from seven different streams in South Korea were investigated. The subfamily composition of chironomids was clearly differentiated accross different levels of organic pollution. Species abundance distributions (SADs) of chironomid communities were compared with the total macroinvertebrate communities across different levels of pollution. The number of species with minimal range of abundance was lower in SADs for chironomid communities compared with total communities. The log normal distribution was accepted for both total and chironomid communities, while the geometric series was relatively more suitable for chironomids and the log series were more fitted to total communities. The a values in the log normal distribution increased in chironomid communities across different levels of pollution, while γ values increased at the polluted sites for both chironomid and total communities. In the Power law analysis, the parameter decreased in chironomid communities. The dominance decay model was more fitted to chironomid communities in clean conditions while random fraction and assortment models were more suitable for the polluted sites.

Benthic macroinvertebrates collected at seven different streams displaying different pollution levels were used to investigatespecies abundance patterns in polluted streams. Community response to disturbances in streams was analysed using speciesabundance distribution (SAD) for benthic macroinvertebrates across different levels of pollution. The slopes of rank abundancewere characteristically steeper with decreasing species richness at the polluted sample sites, while the slopes were less steep withhigher species richness at the clean or slightly disturbed sample sites. The SADs were broadly fitted to the log normaldistribution in benthic macroinvertebrate communities across different levels of pollution. A geometric series was partly acceptedfor the communities at the severely polluted sites where new species could not be readily introduced. A power law was applied tothe SADs, and the parameters reflected the states pertaining to the sample sites. The SADs were also efficient in revealingecological state of communities where physico-chemical indicators could not be easily differentiated for stressful conditions instreams.