Spatio-temporal dynamics of water quality and their linkages with the watershed landscape in highly disturbed headwater watersheds in China.

The water quality of headwater streams is a worldwide concern because of their critical roles in supplying clean water for drinking and other consumptive uses. Here, we evaluate temporal trends and spatial dynamics of the permanganate index (COD), ammonia-nitrogen (AN), and total phosphorus (TP) for 31 sites in headwater watersheds of the Huai River Basin, China. The seasonal Mann-Kendall test and correlation and variance analyses were applied to long-term time series (2003-2010) of water quality data in order to investigate the patterns of water quality trends, as well as their linkages with the watershed landscape. The results indicated that (1) more than 1/3 of headwater monitoring sites have exhibited either significantly increasing or decreasing trends in COD, AN and TP, while only TP increased for most them; (2) obvious increasing concentration gradients were observed for all water quality parameters along the upstream to the downstream continuum. Such spatial patterns can be highly explained by land cover and landscape configuration metrics. The percent of urban land and urban-related landscape metrics (such as the Landscape Division Index) were the primary explanatory variables for AN, while the aggregation metrics of cropland and urban land cover were the main predictors of COD and TP; (3) historical dynamics of COD, AN, and TP were influenced by land cover transitions. The trends of COD and TP may be attributable to the change in the wetland landscape, while the trends of AN were likely related to changes in forestland area as well as environmental management. Overall, our study determined the spatial and temporal dynamics of water quality parameters in the headwater watersheds and interpreted the possible reasons behind their spatio-temporal dynamics, which can have important implications for sustainable landscape planning as well as headwater watershed management.