高寒湿地自然保护区多核心动态分区研究——以四川若尔盖湿地国家级自然保护区为例

中国拥有92466 Km2的各类高原湿地，具有湿地退化、过度放牧等相似特征，保护与利用矛盾突出。高寒湿地保护区尽管在制度上以核心区、缓冲区来约束当地的放牧等外来干扰行为，但在实际管理中却不能起到应有的作用。 本研究以四川若尔盖湿地国家级自然保护区为例，应用3S技术，建立保护区多功能动态分区工作流模型，通过不同植被类型的识别和空间特征分析、不同动物类群在上述植被生境中的时空分布特征分析、保护区主要干扰因素的时空分布特征分析，突出对保护区主要保护对象（湿地生态系统）的保护，对保护区进行管理分区，依据野生动物利用特征和植被生长特征对核心区进行年周期动态利用，缓解保护与发展的矛盾，促进保护区的优化管理。 应用归一化植被指数（NDVI）与植被盖度的相关性，将归一化植被指数（NDVI）转化为植被盖度指数（MDVI），结合保护区牧场划分和时空利用特征专家经验，结果表明，MDVI值在1-139之间主要代表着水体、裸地、沙地等；MDVI140-256为草地和高山灌丛；MDVI210是当地夏牧场和秋冬牧场的划分区间值。 合理的区划需要资金、技术和政策的支持，为保证保护区多核心动态分区的实施，本研究提出了生态工程、牧业发展方式转变、湿地特色产业发展、湿地政策、社区参与和科技支撑等六大保障措施。 In China, 92466 Km2 highland or frigid wetlands are (were) facing major management problems, such as wetland degradation and overgazing. Conflict between conservation and utilization on those wetlands can be found anywhere today. Although many nature reserves have been setup for protection of frigid wetland, and core and buffer zone has been declared to forbid any kinds of disturbance, local farmers still use these areas for grazing. As an example by Sichuan Roige Wetlands National Nature Reserve(SRWNRR), we set up a 3S flow model to analyze the character of year-round distribution patters of vegetation, wildlife, and grazing. Combined and overlapped these characters together, we select multi-core zone and buffer zone, then define a dynamic management period in different zone to optimize protection wetland in the reserve. Normalized Difference Vegetation Index（NDVI）is highly related with coverage of vegetation. When convert NDVI to MDVI (coverage index, 1-256), index 139 and 210 can be as inflexion to distinguish among water/sand/bared land, summer pasture, and autumn / winter pasture. We use these to select different layers and analyze grazing pattern. To be more realistic, we put forward some strategies to support our multi-core and dynamic management of wetland in Roige, including ecological restoration engineering, changing of stock raising industry, changing of wetland policy, community based management and technology renovation supports.