Abstract: Ecological space is an important area for providing ecological service functions and protecting national ecological security. Simulation and prediction for future ecological space can provide reference for the formulation of policies for ecological environment protection, the optimization and control of national space. Taking Wanzhou District as the research area, the data of land use, digital elevation model (DEM), roads, rivers, administrative centers, ecological red lines, and natural protected areas of the District in 2000, 2006, 2012 and 2018 were used to define the types of ecological space of the District. According to the data, the types of ecological space of Wanzhou District include forest land, grassland, water areas, and unused land. The characteristics of the variations of the ecological spaces were analyzed to build the ANN-CA-Markov model for the simulated prediction of the ecological space of Wanzhou District in 2024, and the quality of the habitats of the ecological space of was also evaluated. The results show that the ANN-CA-Markov model has relatively high simulation accuracy by using the cellular automata model (CA) as the main body, which is the effective solution to the deficiency of the CA model. As the simulation is able to deal with impact factors and ecological space changes, a complex nonlinear relation forecasting the distribution of ecological space with the model has higher simulation accuracy (precision reaching 0.9 836) towards transferring quantities and probability forecasts of the ecological spaces. The Wanzhou District has shown a trend of continuous increase in fragmentation degree, differentiation degree, and area uniformity degree, while the stability of the ecological space is gradually declining. The habitat quality of the ecological space is presenting a declining trend year by year. Therefore, while ensuring the increase of the quantity and scale of ecological space, both ecological environment protection and ecological space optimization and control should be taken into account to ensure the stability of the ecological space and habitat quality.

Key words: ecological space, predictive, Ann-CA-Markov, InVEST model, habitat quality