Cascading effects of human activities and ENSO on the water quality of Poyang Lake in China

The aquatic environment in lake ecosystems is greatly affected by human activities and global climate change, while studies on the cascading effects on water environments using a holistic approach are scarce. We employed generalized least squares (GLS) modeling to assess the annual trends in water quality of Lake Poyang from 1983 to 2018 and found that total nitrogen (TN), ammonia nitrogen (NH₄), and the chemical oxygen demand (COD_Mn) increased, while total phosphorus (TP) showed no significant changes. Moreover, Cross-correlation function analyses demonstrated that following the Three Gorges Dam (TGD) operation, the influence of human activities, such as grain yield per unit area (GYP) and urban population (Upop), on water quality became more pronounced, while the role of regional meteorological factors like the monthly maximum value of daily minimum temperature (TNX) decreased. Generalized multilevel path models (GMPMs) revealed that human activities (GPY, Upop, fertilizer application) as well as climate (El Niño-Southern Oscillation (ENSO), meteorology) affected the water quality variables directly or indirectly via the hydrology (sediment discharge, water level). Thus, hydrology dominated the changes in TP (31.6 %) and TN (25.2 %), while human activities controlled the changes in NH₄ (17.9 %) to a higher extent and meteorology the changes in COD_Mn (21.3 %). By contrast, ENSO exerted a relatively weak control on the water quality variables. Our results highlighted that regional meteorology as well as hydrology strongly modified the cascading effects of ENSO and human activities on water quality.