Eutrophic Lake Restoration Requires Multidimensional Approaches to Stressor Management

作者：Zhongzhao Duan, Wei Gao, Xiangzhen Kong, Hugh J. MacIsaac, Erik Jeppesen, Sabine Hilt, David P. Hamilton, Steven W. Wilhelm, R. Michael L. McKay, Yuan Zhang, Boqiang Qin, Yong Liu, Haijun Wang, Min Pan, Tiyuan Xia, Fengchang Wu, Hans W Paerl, Xuexiu Chang · 发表于：Environmental Science & Technology · 年份：2026 · DOI：10.1021/acs.est.5c11422 · 被引用次数：1 · 研究领域：Aquatic Ecosystems and Phytoplankton Dynamics、Freshwater macroinvertebrate diversity and ecology、Marine and coastal ecosystems

Cultural eutrophication has triggered lake regime shifts and cyanobacterial blooms, causing losses of key ecosystem services (e.g., water supply) and necessitating costly remediation. While nutrient enrichment generally is a primary driver, insufficient insight into biotic-hydrological stressor interactions may result in an overreliance on nutrient reduction as the sole intervention. Focusing on China's subtropical eutrophic Lake Dianchi, we applied an innovative coupled hydrodynamic-ecological model to identify key drivers of ecosystem degradation. This approach uniquely involved simulating multiple restoration strategies across eight degradation scenarios, 28 climate change scenarios, and over 100 regulatory scenarios, offering a comprehensive analysis of potential solutions. Our results identify the synergistic effects of grass carp introduction and elevated water levels as key drivers of the observed clear-to-turbid regime shift in the 1970s. Subsequently, fluctuations in nutrient loading became dominant drivers of eutrophication and cyanobacterial blooms. We further show that restoring a clear-water state with submerged macrophytes would optimally require a multidimensional strategy: 63% nutrient loading reduction (76% under simulated climate change conditions), 1.5 m water level lowering, and ∼80% benthivorous fish removal over three consecutive summers. Our study demonstrates that combining biotic, hydrological, and nutrient management outperforms single-factor manipul...