Ecological Evolution and Health Assessment of Dongting Lake Wetland Influenced by Dual Coercion of Nature and Society

In order to explore the trend of Dongting wetland ecological pattern evolution, six representative Landsat images from before and after the completion of the Three Gorges Project on the Yangtze River were selected for classification using support vector machine (SVM), maximum likelihood (ML), and CART decision tree (DT) algorithms. The dynamic degree and transition matrix models were utilized to analyze the dynamic evolution characteristics of wetland patterns, leading to the construction of the pressure-state-response (PSR) model for ecological health assessment. The results showed that the CART decision tree algorithm exhibited high accuracy and stability. Dongting Lake wetland landscape was characterized by complex dynamics, increased fragmentation, and enhanced heterogeneity. The water area increased from 680.59 to 913.47 km², but due to the water storage of the Three Gorges Dam and socio-economic activities like vegetation planting, the area sharply decreased between 2001 and 2004. Sand mining in the lake area led to a reduction in mudflat area at a rate of −8.8 km²/a. The sedge and reed areas increase and decrease alternately, and the woodland area significantly decreased after peaking in 2016. The ecological health values in 1995, 2001, 2004, 2010, 2016, and 2022 were 0.345, 0.635, 0.444, 0.661, 0.425, and 0.335, respectively, showing an overall fluctuation from “sub-healthy” to “very healthy” and back to “sub-healthy.” Ecological health is under dual pressure from socio-economic factors (weight 0.262) and natural changes (weight 0.205), especially with recent extreme climate changes impacting wetland health by as much as 0.281. Considering the balance and coordination between socio-economic development and the natural environment comprehensively, it is necessary to optimize water resource management and increase investment in wetland conservation.