Identification of a Regime Shift in a Shallow Lake Based on Remote Sensing of Submerged Vegetation and Insights for Management

In shallow lakes, clear water with submerged vegetation can shift to turbid water dominated by algae due to disturbances. Identifying factors causing these changes is crucial. Submerged vegetation type, area, and coverage are key indicators. Field investigations are labor-intensive, so remote-sensing methods are used to analyze long-term data. This study used Landsat images to create an NDVI-based decision tree model for Yilong Lake (1991–2020) with 90.09% accuracy and a kappa coefficient of 0.8392. The Mann- Kendall trend test showed ecosystem degradation from 1991 to 2020 (β=-0.0011, P>0.1), especially from 2012 to 2020 (β=- 0.0282, P<0.01). A regime shift was identified, dividing the ecological history into three stages: 1991–1995, 1996–2017, and 2018–2020, with submerged plant coverage of 11.23%, 24.53%, and 5.12%. To prevent eutrophication and protect biodiversity, restoration, load reduction, and water level control are needed.