抄録
The conditions of the rice-growing farmers in Japan become much severer in recent years, so that it is difficult for most small-and medium-sized farmers to continue farming and to properly maintain their farmland in most areas. Establishment of a collective farming system at community level is one of the most hopeful measures to improve their labor productivity and cost saving. Future perspective of the community agriculture is important information to accelerate the establishment of collective farming association.
Applying the Multi-Agent Systems (MAS), we have developed a new simulation model named “Agent-based Simulation Model for Agricultural Planning (ASMAP)” that can properly deal with both farmland lease and contract operations of farming among the farmers in a community. It is a completely-revised version of our previous model (CALL model) that can deal with farmland lease only. The ASMAP can predict future agricultural structure of the community and the area of abandoned farmland under the given conditions. We believe this model can bring about a lot of essential information for agricultural planning at community level that the previous model cannot provide.
We applied the ASMAP to the community of one-hundred farm households in Kanzaki Town, Hyogo Prefecture and verified some behaviors of the model outputs. We found the model has enough accuracy by comparing the simulation results with the actual observations.
We also conducted the analysis on the effects of collective farming association and found it is quite effective to decrease the area of abandoned cultivation. At the last, we pointed out some issues for diffusion of the program package of the model.
