Abstract: Complex ecological environment has posed ever-increasing obstacles to regional ecological security and protection in recent years. Land-use change has been one of the most important indicators for the landscape structure and ecosystem evolution under accelerated urbanization. It is highly urgent to enhance ecosystem conservation for high quality and stability. Ecological networks with ecological corridors and nodes have been also necessary for the process of territorial space planning and ecological civilization construction for ecosystem protection and restoration. Taking the Poyang Lake Plain area as the study area, a quantitative simulation was carried out for the future land-use change under different scenarios using the remote sensing image, planning, and ecological data in 2000, 2010, and 2020. A CA-Markov model was used to predict the land use distribution in the study area in 2030, under the scenarios of Conventional Development Scenario (CDS), Fast Urban Growth (FUG), and Ecological Conservation Scenario (ECS). An evaluation was made on the landscape connectivity to determine the ecological source under various circumstances. The ecological corridors, "Barrier points", and "Pinch points" were retrieved to construct the comprehensive resistance surface using the circuit theory and the Linkage Mapper Arc10.TBX tool. The ecological network verified the stability under different attack modes using the robustness of network connectivity. The results showed as follows. 1) There was the most change in the area of cultivated and construction land in the future land-use change. Specifically, the area of construction land under the FUG increased the most, up 70.27 % from 2020. The cropland area decreased in the three future scenarios. Forest land area under CDS, FUG and ECS decreased by 6.90 %, 7.23 % and 6.70 % respectively compared with 2020. 2) The ecological sources areas under CDS, FUG and ECS scenarios in 2030 were 2 464.2, 2 666.72 and 3 141.88 km2, respectively, where 169, 255, and 299 ecological corridors were identified, respectively, according to the landscape connectivity assessment. Consequently, the ECS scenarios wad retained more ecological sources and ecological corridors. 3) Among the three future scenarios, the FUG and CDS presented the largest number of "Barrier points" (23), and "Pinch points" (117), respectively, whereas, the ECS presented the least "Barrier points" and "Pinch points", but more stable network structure characteristics and functions. 4) The connectivity robustness of CDS, FUG, and ECS scenarios declined slowly in the "Random attack" mode, where the critical values of ecological network stability were 89%, 90%, and 90% node failure ratio, respectively. Among them, the ECS scenario presented a higher critical value of network stability in the "Intentional attack" mode, particularly with a node failure ratio of 55%. Therefore, the ecological network under ECS achieved the highest stability among the three scenarios. This finding can provide a strong quantitative reference and decision-making on ecological spatial planning, ecosystem stability, and protection management strategies in the Poyang Lake Plain region. The case study can also be offered for ecological network research in similar regions.