QQ客服 官方微博 反馈留言

题名	横断山区自然保护区网络保护成效评价研究
作者	叶鑫
学位类别	博士
答辩日期	2016-05
授予单位	中国科学院研究生院
授予地点	北京
导师	刘国华
关键词	protected area effectiveness，biodiversity，threat assessment，forest loss， functional connectivity 保护有效性，生物多样性，威胁评价，森林退化，功能连接性
其他题名	Assessing the effectiveness of PA network in the Hengduan Mountains of southwest China
中文摘要	保护区是进行生物多样性保护活动的重要基石，因为它提供了多数种类物种赖以生存所需要的安全的庇护所。到2012 年为止，保护区已经覆盖了全球陆地表面13%的面积，这个数字相比较上世纪70 年代增长了约4 倍。同时，保护区在中国西南地区过去的30 年内也经历了爆发式的增长，但是对保护区的保护成效的认识较为缺乏。在保护区周围快速的人类土地利用活动的增加逐渐降低了保护区保护物种，维持生态过程及提供服务的能力。其中，生境丧失和生境隔离是两种最直接的威胁方式。因此，当出现保护区未能减少生境的丧失，保护区生境网络未能建立起连接的廊道，关键的核心区域保护比例低下等问题时，保护区有效性较低。评价保护区的保护成效是识别保护现状和空缺的重要方式，并且可以为采取合理的保护应对措施提供理论依据，为区域生态安全格局的维持提供基础。然而，在西南横断山区生物多样性热点区的大部分研究主要是基于单个保护区来开展的，缺乏关于保护区网络及其周围环境基质的综合研究。因此，我们从如下四个方面对横断山区保护区进行了保护成效的分析和评价。       （1）采用1990、2000、2005、2010 四期土地利用数据，及区域内159 个自然保护区的空间矢量数据进行分析。在保护区边界以外分别建立5km、10km、20km 三个缓冲区梯度。分析土地利用变化特征时，主要考虑研究区主要生境类型，如森林、灌木等的退化，恢复，以及净效应，并分析主要人类干扰地类的变化特征。通过比较1990-2000 与2000-2010 年这两个时段，发现保护区10km 以内，随时间变化森林退化和净退化面积在增加，保护区10km 以外，森林退化和净退化面积在减少。单位面积的森林退化率上，1990-2000 年森林退化和净变化的最高值出现在保护区10-20km 的区域内，然而到了2000-2010 年，退化比例最高值向靠近保护区的0-5km 区域逼近，保护区所受到的森林退化的威胁在逐渐加大。随着缓冲区梯度的增加，2000-2005 和2005-2010 年两个时段均发生森林灌木退化的保护区比例在增加，其中0-5km 范围内增加的幅度较大。后五年相比前五年，退化面积在加剧的保护区个数在增多。同时，森林灌木退化面积也更为严重，特别是在保护区的0-10km 缓冲区内。2000-2010 年间，保护区内，0-5km缓冲区和5-10km 缓冲区出现净森林灌木退化和净草地退化的保护区比例均在0-5km 范围最高，因此是生态保护重点值得关注的区域。        （2）使用协变量匹配的方法对1990 年前建立的保护区，1990-1999 年间建立的保护区，2000-2004 年间建立的保护区进行减缓森林丧失的有效性分析。Gen匹配相比PSM 匹配的效果更好，对于1990 年前建立的保护区，匹配后减缓的森林丧失率为0.0617%，只相当于未进行匹配分析之前的21.7%。对于1990-1999年间建立的保护区，保护区减缓森林丧失的比也为0.0617%，相当于未进行匹配分析之前的47.5%。对于2000-2004 年间建立的保护区，保护区减缓森林丧失的比例为0.0337%，相当于未进行匹配分析之前的59.6%。年均减缓森林丧失的比例上可以得出，1990 年前建立的保护区要小于1990-1999 年建立的保护区和2000-2004 年建立的保护区。年均减缓森林丧失作用最好的是1990 年前建立的市县级自然保护区，达到-0.01329%/年，其次是1990-1999 年间建立的国家级自然保护区，达到-0.01182%/年。省级自然保护区保护的有效性相对较低。但是，国家级保护区在减缓森林丧失的总面积上最多，起到了积极的作用。省级保护区在年均比率及总面积上，对于减缓森林丧失的有效性上都不如其它两种级别的保护区。       （3）基于高精度的土地利用数据和详细的道路数据，结合主要威胁因子采用多尺度的方法定量评估了区域上受威胁程度的大小及空间格局。采用了当地区域威胁指数和综合威胁指数的方法来定量保护区及周围环境的威胁大小及关系，评价了不同级别、不同类型、不同缓冲区大小外所受到威胁的程度。结果显示，国家级保护区的威胁值低于省级保护区，并且显著低于市县级保护区(p<0.05)，表明国家级保护区具有相对较低的威胁水平，以及相对较高的保护和管理水平。区域内，保护区具有明显的边缘效应，较低威胁水平所占的比例在保护区边界外5km 范围内急剧下降。然而，国家级保护区的边缘效应更为明显，在边界外5km范围内下降了58.3%，高于省级保护区的34.8%以及市县级保护区的33.4%。此外，保护区的面积大小和保护区内及缓冲区内综合威胁指数中较低威胁等级所占的比例呈极显著正相关关系。2000 年关键核心区植被生境受到保护的比例均达到了10%以上，其中核心区湿地受保护的比例最高。灌木、草地和湿地的关键核心区受到的保护主要来自于国家级保护区。        （4）区域上基于县域尺度利用dECA 指数分析了1990，2010 两个时段景观连接性的变化。从1990 年到2010 年间，连接性的变化在整个区域上是处于相对比较稳定的， 虽然dECA 值在-0.5%和0.5%之间的区域仅占到了46%，dECA 值在-1%和1%之间的区域占到了约70%。然而，在不同的区域不同的县域之间，这种差异是比较明显的。随着距离情景的增加，连接性降低的区域和改善的区域绝对值均在减小。对于连接性下降的县域，在d=1km 时连接性下降的平均值最大，达到-1.09%。各县森林核心区面积平均下降了0.08%，无论在何种距离情景下，连接性下降的幅度均大于面积的下降幅度，表明了从区域总体上来看，森林核心生境面积的下降除了使生境的有效面积下降以外，还进一步造成了生境斑块间的隔离。但是，在大部分传播/迁移距离的情景下，出现连接性的负面效应的县域比例低于连接性的正面效应的县域比例。针对横断山中部dECA 下降较为明显的8 个县域，进一步在斑块尺度上分析了生境斑块的连接重要性作用。其中，1990-2000 年国家级保护区保护的重要生境面积最多，省级保护区其次，2000 后市县级保护区第三。保护区的类型分类上，1990-2000 年以及2000 年后建立的野生动物保护区保护效率最高。
英文摘要	Protected areas (PAs) are a cornerstone of biodiversity conservation efforts, as they provide various species with safe havens. PAs cover about 13% of the global land surface in 2012, an increase of 400% since the 1970s. Protected areas (PAs) have experienced explosive growth in the southwest China over the last three decades, but their effectiveness has not attracted considerable attention. The rapid increase in human land use surrounding PAs gradually diminishes these protected areas’ capacity to preserve species and maintain ecological processes. Habitat loss and habitat isolation are two direct consequences of threats both within and surrounding their PA boundaries. Therefore, PAs may not be effective when they fail to limit habitat loss and are unconnected via corridors to other wild areas. Evaluating the effectiveness of PAs is an important aspect of identifying conservation status and gaps, and can enable us to take reasonable actions. However, most of the studies conducted in the Mountains of Southwest China Hotspot have been based only on a single protected area. There is a shortage of comprehensive studies examining PA networks and the buffers surrounding them. Therefore, we analyzed and assessed the effectiveness of PA in this region from the following four parts.     (1) We analyzed land use changes with four period land cover data (1990, 2000, 2005, and 2010) and 159 PA spatial vector data. Outside each PA, we made three buffers, which were 5km, 10km, 20km, respectively. Analysis processes were primarily focused on degradation, restoration, and net effect of forests, grasslands,and shrubs. The same method was used in human disturbed land types. Inside the 10km buffer of PA, the area of forest loss and net loss increased in 2000-2010 compared to 1990-2000. Conversely, the area of forest loss and net loss decreased outside the 10km buffer of PA. Considering the forest loss and net loss per unit area,the highest forest loss was in the range of 10-20km buffer outside the PAs from 1990 to 2000. However, the highest forest loss tended to be in 0-5km buffer outside the PAs from 2000 to 2010. The potential threats around PAs were increasing. With the increase of buffer gradient, the percentages of PA numbers increased, which had forest loss in both periods of 2000-2005 and 2005-2010. The region of 0-5km buffer had the largest increase. The numbers of PAs with forest loss and total areas of forest loss were both large in 2005-2010 compared to 2000-2005. The percentage of PA with net forest, shrub and grassland loss was highest in the range of 0-5km buffer outside the PAs in the period of 2000-2010. Therefore, this region should be paid more attention for ecological protection.      (2) We used the method of covariate matching to analyze the effectiveness of PAs in reducing deforestation compared with unprotected areas. Paired PAs used to analyze were PAs established before 1990, PAs established between 1990 and 1999,and PAs established between 2000 and 2004. Gemmatch performed better than propensity scores matching (PSM) that forest loss reducing was 0.0617% after matching for PAs established before 1990, equal to 21.7% of conventional method.Forest loss reducing was also 0.0617% after matching for PAs established between 1990 to1999, equal to 47.5% of conventional method. Forest loss reducing was 0.0337% after matching for PAs established between 2000 and 2004, equal to 59.6% of conventional method. Considering the annual rate of forest loss reducing, PAs established before 1990 were lower than PAs established between 1990 and 1999, and PAs established between 2000 and 2004. The best effect of forest loss reducing was OPAs established before 1990, reached -0.01329%/year. The second was NPAs established between 1990 and 1999, reached -0.01182%/year. However, NPAs gained the largest areas in reducing deforestation. PPAs were relatively ineffective compared to NPAs and OPAs in both annual rate and total areas of forest loss reducing.      (3) Based on high resolution land cover data and detailed road dataset, we applied a threat framework as a means of quantitatively assessing local and surrounding threats and spatial patterns to different types of PAs with gradient buffers in the Hengduan Mountain Hotspot of southwest China. We used local threat index and integrated threat index to quantitatively assess the threats, which affected PAs with different protection levels, types and buffers. Our findings show that national protected areas (NPAs) have lower and significantly lower threat values (p<0.05) than provincial protected areas (PPAs) and other protected areas (OPAs), respectively,which indicates that NPAs are lands with a lower threat level and higher levels of protection and management. PAs have clear edge effects, as the proportion of areas with low threat levels decline dramatically in the 5-kilometer buffers just outside the PAs. However, NPAs suffered greater declines (58.3%) than PPAs (34.8%) and OPAs (33.4%) in the 5-kilometer buffers. Moreover, a significant positive correlation was found between the size of PAs and the proportion of areas with low threat levels that they contained in both PAs and PA buffers (p<0.01). More than 10% of key core area of main natural vegetation were protected by PAs in 2000. Key core areas of wetland, shrub and grassland were mainly protected by NPAs.      (4) We used dECA index to analyze landscape connectivity changes between 1990 and 2010 at county scale in the region. Landscape connectivity kept relatively stable from 1990 to 2010 in the region. Although dECA value between -0.5% and 0.5% accounted for only 46% in the region, dECA value between -1% and 1% accounted for nearly 70%. However, the differences were evident among different counties. With the increase of scenario distances, the absolute value of dECA increase or decrease gradually decreased. For the counties with negative dECA, the largest change of connectivity occurred under the scenario of d=1km, and the mean reached -1.09%. Forest core areas of each county dropped by an average of 0.08%, no matter what nnectivity decline was higher than the core area decline,which means forest loss have further caused habitat isolation besides the effective habitat area decline. However, under most of the disperse distance scenarios, county proportions of negative connectivity effect were lower than the county proportions of positive connectivity effect. For eight counties with relatively large connectivity declines in the center of Hengduan Mountains, we further analyzed the importance of habitat patches on maintain the overall landscape connectivity at patch level. NPAs established between 1990 and 2000 contained the most key connectivity patches,followed by PPAs established between 1990 and 2000, and OPAs established after 2000. APAs established between 1990 and 2000, and after 2000 had higher protection effectiveness than other types of PA.
内容类型	学位论文
源URL	[http://ir.rcees.ac.cn/handle/311016/37037]
专题	生态环境研究中心_城市与区域生态国家重点实验室
推荐引用方式 GB/T 7714	叶鑫. 横断山区自然保护区网络保护成效评价研究[D]. 北京. 中国科学院研究生院. 2016.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。