面向多维度综合评价的自然保护地边界优化方法研究——以环哀牢山国家公园候选区为例

Research on Boundary Optimization Methods for Protected Areas Based on Multi-Dimensional Comprehensive Evaluation: A Case Study of the Proposed Ailaoshan National Park Region

关键词:  风景园林  自然保护地  生态系统服务  生物多样性  系统保护规划  边界优化

Key words:  landscape architecture  protected area  ecosystem service  biodiversity  systematic conservation planning  boundary optimization

摘要:

科学的自然保护地边界划定是确保生态系统完整性和生物多样性有效保护的关键前提。以环哀牢山国家公园候选区为例，构建融合生态系统服务、生物多样性保护与人类活动影响的多维评价体系，提升划定方法的科学性与系统性。通过InVEST模型评估生态系统服务供给能力，结合MaxEnt模型分析关键物种适生区，再设置兼顾人类活动影响、管理可行性等因素的保护成本，运用Marxan系统保护规划模型识别优先保护区与保护空缺。结果表明：1)生态系统服务高价值区集中在哀牢山中南部与无量山南部，关键物种适生区集中在哀牢山东部及南部；2)现状保护地覆盖了23.92%的一级优先保护区，但存在3 410.72 km2的一级保护空缺，集中分布于无量山西南侧与哀牢山东西两侧区域；3)为实现更加有效的保护，建议将国家公园面积扩展至3 664.96 km2，周边保护地总面积提升至2 243.97 km2。研究成果可为其他自然保护地边界优化提供方法参考，并提出了潜在“其他有效的区域保护措施”(OECMs)的识别方法

Abstract:

Scientific boundary delineation of protected areas (PAs) is a crucial prerequisite for ensuring ecosystem integrity and effective biodiversity conservation. To address this need and enhance the scientific rigor and systematicity of delineation methodologies, this study employs the proposed Ailaoshan National Park region in Yunnan Province, China, as a critical case study. This region, characterized by rich biodiversity as a key transition zone between tropical and temperate Asia and serving as the catchment for major transboundary rivers, necessitates precise boundary setting. The study constructed a comprehensive multi-dimensional evaluation framework integrating three core dimensions: ecosystem service (ES) supply capacity, biodiversity conservation value, and human activity impacts. Specifically, the spatial patterns of four key ES (habitat quality, carbon storage, water yield, and soil retention) were quantitatively evaluated using the InVEST model, leveraging high-resolution spatial datasets. Concurrently, the suitable habitats for four flagship/key species endemic to the region - Western Black Crested Gibbon (Nomascus concolor), Green Peafowl (Pavo muticus), Forest Musk Deer (Moschus berezovskii), and Hume's Pheasant (Syrmaticus humiae) - were modeled using the Maximum Entropy (MaxEnt) algorithm, incorporating species occurrence records and critical environmental variables. To identify priority conservation areas and gaps, we implemented systematic conservation planning using the Marxan model. Recognizing the importance of ecological homogeneity and management feasibility, sub-catchments (n=2,947) derived from hydrological analysis were selected as planning units. The eight conservation features comprised the four ES layers and the four species habitat suitability layers. Protection targets were set differentially based on species endangerment, distribution range, and established benchmarks, ranging from 30% for widespread species and ES to 50% for the critically endangered and range-restricted gibbon. Critically, a conservation cost surface was meticulously developed to reflect management feasibility and human pressure. This cost layer integrated four weighted factors via Analytic Hierarchy Process (AHP) and expert scoring: landuse type, Normalized Difference Vegetation Index (NDVI), existing PA management level, and human footprint intensity. Marxan was run iteratively with sensitivity analysis to determine an optimal Boundary Length Modifier (BLM) balancing cost and spatial compactness. Key findings include: 1) High-value ecosystem service areas are concentrated in central-southern Ailaoshan and southern Wuliangshan, while key species habitats cluster in eastern and southern Ailaoshan; 2) Existing protected areas cover 23.92% of primary priority zones, but leave 3,410.72 km2 of primary conservation gaps, mainly in southwestern Wuliangshan and eastern/western Ailaoshan; 3) To achieve more effective conservation and enhance ecological connectivity, especially between the Ailaoshan and Wuliangshan sections currently separated by non-suitable habitat barriers for key species like the gibbon, significant boundary optimization is proposed. We recommend expanding the core national park area from the original proposal to 3,664.96 km2 (an increase of 2,127.59 km2), integrating large, contiguous high-priority patches. Additionally, surrounding PAs should be increased to a total area of 2,243.97 km2 (an increase of 1,061.43 km2). This optimized configuration substantially increases coverage of primary priority areas, raising the overlap with ES priority zones by 38.65% and with biodiversity priority zones by 52.74%. For areas where formal PA expansion faces high socio-economic costs or management feasibility challenges, the study also proposes the identification of potential "Other Effective area-based Conservation Measures" (OECMs) or Conservation Compatible Lands (CCL) as complementary strategies. This study provides a replicable methodological reference for boundary optimization of other protected areas, particularly national parks, by demonstrating the integrated application of InVEST, MaxEnt, and Marxan within a multi-dimensional framework that explicitly incorporates ecosystem services, biodiversity, human impacts, and management costs.