Improvement Of Water Quality Assessment Using Benthos By Natural Variable Control And Species DNA Barcoding Approach

Both macroinvertebrates and diatom are widely used as indicators for freshwater ecosystem health and water quality biomonitoring.However,the confounding impact of natural variability and human disturbance and the limit of morphological identification affect the accuracy and precision of bioassement.In addition,understanding the concordance between aquatic assemblages and environmental factors is a fundamental step towards achieving sustainable stream and catchment management.Here we used the Random Forest model(RF)to eliminate the impact of natural environmental factors and developed multimetric indices(MMI)to expore the the impact on the performance of diatom and macroinvertebrate MMI and compare MMI between the two assemblages;and identification to the lowest level possible of macroinvertebrate by the morphological characteristic and DNA barcodes to explore the response relationship between the community and environment factors.In summer 2013,we collected diatom,macroinvertebrate,and environmental data from 31 sites in northwest Chinese arid streams(the Ertix River Basin in Xijiang).We used the RF model to control the natural gradient and developed multimetric indices(MMIs)to expore the changes in the performance of diatom and macroinvertebrate MMIs after controlling the natural variables.We also explored the concordance of MMIs between two assemblages in their response to disturbances.Random Forest(RF)modeling revealed that both diatom and macroinvertebrate metrics were strongly associated with natural environmental heterogeneity.However,diatom metrics responded more strongly to climatic variables(e.g.,precipitation and temperature),and macroinvertebrate metrics were more sensitive to geomorphic variables(e.g.,elevation range and watershed topography).Diatom MMIs after adjusting for natural variance had greater precision and responsiveness than macroinvertebrate MMIs.The response of MMI based on diatoms and macroinvertebrate to human disturbance and the result of assessment the health of stream ecological are not consistent.This supports the effect control of natural variables in building MMIs could improve the responsiveness and precision the assessment.Moreover it also indicates that the use of multiple biological assemblages with differing ecological and biogeographical traits in assessments of biological condition and biological diversity.Whether the lower identification level the more reliability of assessment is still in argument for water quality biomonitoring.We collected and measured benthos and environment variables from 20 sites distributed in streams,rivers and Taihu lake.We firstly identified all benthos to lowest level as possible according morphological eharacteristics and then farther identified chironomids to species level according to DNA barcodes which we worked out in the lab.We aimed to explore the response of biological assemblages to environmental factors by the different taxonomic identification level as well as different methods.We morphologically identified 113 taxa belonging to 12 orders and 4 classes.Of which we morphologically indentified 30 taxa of chironomids.We divided the species based on 5%genetic distance.However,we determined 95 DNA barcoding MOTU of chironomids and most of them were found in streams.CCA analysis showed that there was no obvious increase in explanation of community variance by environmental variables between morphological data and DNA barcoding data.In addition,GLM analysis also showed that there was no difference in modeling four index between morphological data and DNA barcoding data.Our results suggested that 1)Using DNA barcoding can increase the species richness and the sensitivity to the environment.2)Only improve the identification level of chironomids would not improve the interpretation of relationship between community and environmental variables.