Abstract: Based on the sum-up and abstraction of the relationships of rapeseed growth characters with ecological environment, cultivar type, production condition and yield target, a dynamic knowl-edge model was developed by using knowledge engineering and system modeling method, which could be used for designing a suitable sowing and transplanting scheme of different rapeseed varieties under different spatial and temporal environments. Case studies on the knowledge model with the data sets of three different sites, nine different variety types, and two different sowing styles indicated a good performance of the model system in decision-making, explanation, and wide applicability.

Key words: Pasture and agriculture transitional zone, System productivity, Ecological reconstruction, Grass/Livestock industries