Quantitative Tolerance Values For Common Stream Benthic Macroinvertebrate And Water Quality Classification Use Of Biotic Index(Bi)in Yangtze River Delta

It is widely recognized that tolerance values (TV) of macroinvertebrate are effectively used in water quality bioassessments. Most original tolerance values were subjectively determined, but recently tolerance values have been objectively developed via statistical analyses of the relationships among biological and environmental data. Based on macroinvertebrate and environmental data from286sampling sites collected from streams in Yangtze River Delta, we used principal component analysis (PCA) to create a synthetic gradient of four variables (total nitrogen, total phosphorus, dissolved oxygen,%silt). For taxa found at=20sites, we calculated their initial tolerance values as their weighted first principal component (PCI) scores at the sites with their relative abundances as the weights, and then scaled from0to10to assign the final tolerance value of each taxon. We derived tolerance values for163taxa quantitatively and values for51taxa, occurred less than20sites, were subjectively designated. All the seven TV related metrics:Hilsenhoff Biotic Index (HBI), intolerant taxa richness, percent of intolerant taxa, intolerant EPT taxa richness, percent of intolerant taxa, tolerant taxa richness and percent of tolerant taxa, our TVr developed in this study had significantly realtionships to total nitrogen, total phosphorus, dissolved oxygen,%silt,%agricultural land use and%urban land use, and a relative higher descriment efficiency than those using unrevised TVr. Our results highlights the importance of deriving regional tolerance value from the objective method for benthic macroinvertebrates and its potential use as independent indicators for water quality bioassessment and management in China.We used the following method to provide bioassessment classification system in water quality (stream, river, and lake) based on Biotic Index (BI), that is the the5th percentile of the BI from all sites was determined as the benchmark of best expected value. Any values below the5th percentile were considered as "excellent", and the values of the5th percentile value to maximum value were evenly distributed into four parts, represented different levels of water quality. Bioassessment classification system of BI for stream water quality was preliminary determined,<2.73Excellent,2.73-4.55Good,4.56-6.38Good-Fair,6.39-8.21Fair,>8.21Poor. Bioassessment classification system of BI for river water quality was preliminary determined,<4.82Excellent,4.82-6.07Good,6.08-7.33Good-Fair,7.34-8.59Fair,>8.59Poor. Bioassessment classification system of BI for lake water quality was preliminary determined,<5.40Excellent,5.40-6.43Good,6.44-7.47Good-Fair,7.48-8.51Fair,>8.51Poor. The accuracy of BI in water quality bioassessment was examined using data from stream sites.The correlation analysis between BI and chemical factors showed that BI had high correlation with DO(r=-0.474), TN(r=0.485), TP(r=0.394) and%silt(r=0.568). Comparing with the value of physical and chemical classification in the national standard GB3838-2002, results showed that the mean value of TP of different water quality levels based on BI classification system match the national standards, and the mean value of DO was generally higher than the maximum value of GB3838-2002, and the mean value of TN was much higher than the maximum value of GB3838-2002, reasons should to be future research. The applications on41samples of the middle and upstream in TiaoXi River and12samples of LinAn, results showed that the formulation of BI criteria in this study was reliable.