High-density urban development has led to increasing fragmentation of blue-green spaces, which constrains the realization of ecological benefits. This study applies machine learning and SHAP analysis to identify the relationships between morphological characteristics of blue-green spaces and ecological benefits, and constructs a morphology-ecology response model to support ecological benefit prediction and morphological optimization within urban planning schemes. The iterative nature of the model makes it applicable to most urban blue-green space planning scenarios. It enables future planning regulation under the condition of coordinated optimization of morphology and ecology, thereby supporting the scientific planning of urban blue-green spaces.