Landscape ecological risk is the term used to describe potential negative effects of the relationship between environmental processes and landscape structure as a result of both natural and human interactions. The scientific assessment of past and prediction future ecological risks to landscapes is the basis for achieving sustainable regional development. In this study, we built an ecological risk measurement system characterized by variability in landscape composition to reflect the components of landscape ecological risk. In particular, the CA-Markov model was used to simulate the spatial pattern of land use in the Sanjiang Plain, China, under a wetland protection scenario. The evolution of the spatiotemporal land use trend in the study area was then analyzed for past (2011–2021) and future (2021–2036) periods. The results identified that construction land, farmland, water, and grassland areas exhibited an increasing trend from 2011 to 2021, while forest and wetland areas decreased each year. From 2021 to 2036, the grassland areas exhibited a downward trend, while the other land use types maintained the same previous trend, and all land use types experienced a slowdown in the degree of change. The major landscape ecological risk levels in the research region changed from medium-risk and low-risk levels (2011–2021) to low-risk and lower-risk levels (2021–2036). An evident decrease in risk is observed for farmland and construction land in the mid-region of the research district and forests in the south. In contrast, high-risk and higher-risk levels regions were smaller and distributed in the Songhua River basin in the northeast of the research region, wetland nature reserves, and the Xingkai Lake area, with an overall reduction of 59%. Based on the results, we propose several risk prevention and control management strategies. This research acts as a basis for determining Sanjiang Plain's future rational distribution of land resources and provides an example for the reduction of ecological risks in similar regions.