River Water Quality And Its Responses To Context Characteristics Of Landscapes In Taihu Lake Basin

Water is one of the most important material foundation of human survival and development. Nowadays, water resource shortage and water environmental problem have gone from bad to worse all over the world. Exploring the cause and solution to the dilemma has become a major issue faced by all human beings to be solved.Non-point source pollution mainly caused by human activities was considered to be the principal reason for water environmental problem, and its effect is significantly influenced by natural and geographical conditions. It has become a consensus shared by most of scientists that context characteristics including human activities factors and natural factors determine the regional water environment to some degree across multiple spatial scales.Taihu Lake basin experienced dramatic urbanization and economic growth since the1980's. The water environments have been deteriorating in association with such rapid regional development. Non-point source pollution by land use is recognized as the most important factor influencing water quality. Revealing the spatiotemporal characteristic of water environment and its effect factors will do much help to the pointed protection and treatment of the basin's water environment.To explore the responses of water quality to context characteristics of landscapes at watershed scale, we examined the correlation between9landscape factors characterizing land use, stream density, precipitation and topography and5water chemistry indicators of the rivers in Taihu Lake basin, based on the water quality monitoring data at80invertebrate sites released by Taihu Basin Authority from2006to2010. We analyze the river water quality status of the basin and its spatiotemporal dynamic characteristics using spatial autocorrelation analysis and nonparametric test. Using redundancy analysis, the correlation between the factors and water chemistry indicators was conducted at9spatial scales from0.5km to24km, respectively. The results showed that:(1) From2006-2010, river water quality of the80invertebrate sites involved in the research presented an overall poor status, with most of the water quality of the sites worse than grade V, grade V and grade IV. The indicators with concentration exceeding the national standard included total phosphorus (TP), ammonia nitrogen (AN), biochemical oxygen demand (BOD), chemical oxygen demand (COD) and dissolved oxygen (DO).(2) Water quality varied significantly by location: in the upstream basin, apparent difference was observed between the plain areas with dense stream networks and the forested areas; while in the downstream basin, the rivers flowing out of Taihu Lake showed significant difference between their upper and lower reaches.(3) From the perspective of water quality grades, most of the sites presented an improving trend though the five years. But there was great difference between the change trends of frequencies of the water quality indicators concentration exceeding the national standard, as DO declined annually, while other indicators showed various degree of fluctuation. Compared2006with2010, frequency of TP concentration exceeding national standard declined with a statistical significance (p<0.1), while AN and DO declined extreme significantly (p<0.01), but BOD and COD increased with no statistical significance.(4) The investigated landscape factors had significant effects on water quality of the rivers. The water chemistry indicators showed scale-and location-dependent responses to the landscape factors. Moreover, different indicators had distinct responses. The explanatory power of each landscape factor on the variance in water quality showed apparent scale-dependent features.(5) With increasing spatial scale, the curves of total explanatory power showed a consistent pattern in the upstream and downstream basin, presenting double peaks near the scales of0.5-1km and16km. In the upstream basin, natural wetlands and slope presented the highest exploratory power at the small scale and large scale, respectively. While in the downstream basin, stream density and farmland had the largest effect at small scale and large scale, respectively.(6) All the water quality indicators had different response to the context characteristics of landscapes, scale-depended and varied by location between upstream and downstream basin. Specifically, AN, TP and DO were positively correlated with settlement cover in the upstream basin, while with farmland cover and stream density in the downstream basin. BOD and COD exhibited negative correlations with natural wetlands and positive correlations with artificial wetlands in the upstream basin; while they were negatively correlated with slope and positively correlated with stream density in the downstream basin.In general, water quality of the rivers in Taihu lake basin from2006-2010presented an overall poor status but an improving trend. Water quality was significantly affected by the landscape factors at multiple spatial scales. The effect power and principle scales of the factors, responses of indicators to the factors and their relationship by locations all showed great differences.The result indicates that differential treatment tactics should be developed for different regions and water quality indicators but increasing proportion of forest and natural wetland has a universal meaning for improving the water environment. Further research may focus on the cooperation of the landscape factors and the spatially nested mechanism, while enhancing the connection between results of theoretical research and practical applications is also needed.