基于空间生态集合覆盖模型的自学习禁忌搜索算法在保护区空间选址中的应用——以戴云山为例

自然保护区规划是保护生物多样性的有效方式.传统保护区规划方法只能识别物种保护的重点区域,无法科学确定保护区的适宜面积.地块选择方法基于数学模型,从规划区域中选择部分地块组成自然保护区,保护特定物种或生态系统,是缓解生态保护与开发利用矛盾的重要手段.现有地块选择法未考虑各单元生态差异,且最优化算法存在计算效率的瓶颈.本文首先构建适用于森林生态系统的生态值赋分评价体系,据此计算戴云山生态值并绘制其分布图;然后,结合生态值建立生态集合覆盖模型(ESCP),并基于ESCP嵌入空间紧凑性提出空间生态集合覆盖模型(SSCP);最后,利用寻优性能良好的自学习禁忌搜索算法(STS)搜索各保护目标下的近似最优选址方案,给出福建省戴云山现有建成区优化方案.结果表明: 戴云山生态值计算结果在空间分布上存在明显差异;ESCP比原集合覆盖模型(SCP)能产生生态值更高的选址方案;SSCP在ESCP基础上对生态值较高区域有聚集作用,且周长权重越大,聚集效果越明显;建议现保护区可向外拓展136 km²,并将西北向分布长苞铁杉的地块纳入保护区范围.研究结果为实现戴云山保护区可持续发展及土地资源优化配置提供了优化方案,也可为我国森林生态系统类型的自然保护区设计提供新思路.

Site selection of nature reserve based on the self-learning tabu search algorithm with space-ecology set covering problem: An example from Daiyun Mountain,Southeast China

Designing the nature reserves is an effective approach to protecting biodiversity. The traditional approaches to designing the nature reserves could only identify the core area for protecting the species without specifying an appropriate land area of the nature reserve. The site selection approaches, which are based on mathematical model, can select part of the land from the planning area to compose the nature reserve and to protect specific species or ecosystem. They are useful approaches to alleviating the contradiction between ecological protection and development. The existing site selection methods do not consider the ecological differences between each unit and has the bottleneck of computational efficiency in optimization algorithm. In this study, we first constructed the ecological value assessment system which was appropriated for forest ecosystem and that was used for calculating ecological value of Daiyun Mountain and for drawing its distribution map. Then, the Ecological Set Covering Problem (ESCP) was established by integrating the ecological values and then the Space-ecology Set Covering Problem (SSCP) was generated based on the spatial compactness of ESCP. Finally, the STS algorithm which possessed good optimizing performance was utilized to search the approximate optimal solution under diverse protection targets, and the optimization solution of the built-up area of Daiyun Mountain was proposed. According to the experimental results, the difference of ecological values in the spatial distribution was obvious. The ecological va-lue of selected sites of ESCP was higher than that of SCP. SSCP could aggregate the sites with high ecological value based on ESCP. From the results, the level of the aggregation increased with the weight of the perimeter. We suggested that the range of the existing reserve could be expanded for about 136 km² and the site of Tsuga longibracteata should be included, which was located in the northwest of the study area. Our research aimed at providing an optimization scheme for the sustai-nable development of Daiyun Mountain nature reserve and the optimal allocation of land resource, and a novel idea for designing the nature reserve of forest ecosystem in China.