Abstract: Optimization in spatial structure of land use is an important way to effectively manage ecosystem services and coordinate the conflicts between human and land. Current available optimization programs are mostly from the perspective of whole region, ignoring the regional differentiation of ecosystem services, which would underestimate optimization benefits or be low in feasibility. Taking Bayannur City of Inner Mongolia Autonomous Region as an example, we used ecosystem service cluster and genetic algorithm to implement zoning land use optimization based on SAORES platform. The results showed that compared with current land use pattern, the overall optimization of land use layout functionally improved ecological benefits by 16.8% and reduced the service of food production by 6.4%. The zoning optimization of land use layout was more cost-effective, with 27.3% improvement in ecological benefits and only 5.6% reduction in food production service. The landscape disturbance in the pattern showed a general decrease after optimization. The edge effect was significantly reduced, with no significant difference between the overall and zoning optimization schemes. We proposed an optimized scheme for the spatial configuration of ecological reserves conforming to the long-term development of Bayannur City.

Key words: arid area, ecosystem services, territorial space optimization, genetic algorithm, SAORES.