Abstract
Lake‐water quality is highly dependent on the landscape characteristics in its respective watershed. In this study, we investigated the relationships between lake‐water quality and landscape composition and configuration within the watershed in the Yangtze River basin of China. Water quality variables, including pH, electrical conductivity (EC), dissolved oxygen (DO), Secchi depth (SD), NO~2~^−^, NO~3~^−^, NH~4~^+^, TN, TP, chemical oxygen demand (COD~Mn~), chlorophyll‐a (Chl‐a), and trophic state index (TSI), were collected from 16 lakes during the period of 2001–2003. Landscape composition (i.e. the percentage of vegetation, agriculture, water, urban, and bare land) and landscape configuration metrics, including number of patches (NP), patch density (PD), largest patch index (LPI), edge density (ED), mean patch area (MPA), mean shape index (MSI), contagion (CONTAG), patch cohesion index (COHESION), Shannon's diversity index (SHDI), and aggregation index (AI), were calculated for each lake's watershed. Results revealed that the percentage of agriculture was negatively related to NO~2~^−^, TN, TP, Chl‐a concentrations, and TSI, while the percentage of urban was significantly correlated with EC, NH~4~^+^, and COD~Mn~ concentrations. Among landscape‐level configuration metrics, only ED showed significant relationships with TN, TP concentrations, and TSI. However, at the class level, the PD, LPI, ED, and AI of agriculture and urban land uses were significantly correlated with two or more water quality variables. This study suggests that, for a given total area, large and clustered agricultural or urban patches in the watershed may have a greater impact on lake‐water quality than small and scattered ones. Copyright © 2011 John Wiley & Sons, Ltd.