School of Forestry and Merriam-Powell Center for Environmental Research, Northern Arizona University, Flagstaff, AZ 96011-5018, USA.
Conserv Biol. 2010 Jun;24(3):701-10. doi: 10.1111/j.1523-1739.2009.01422.x. Epub 2010 Jan 11.
Even under the most optimistic scenarios, during the next century human-caused climate change will threaten many wild populations and species. The most useful conservation response is to enlarge and link protected areas to support range shifts by plants and animals. To prioritize land for reserves and linkages, some scientists attempt to chain together four highly uncertain models (emission scenarios, global air-ocean circulation, regional circulation, and biotic response). This approach has high risk of error propagation and compounding and produces outputs at a coarser scale than conservation decisions. Instead, we advocate identifying land facets-recurring landscape units with uniform topographic and soil attributes-and designing reserves and linkages for diversity and interspersion of these units. This coarse-filter approach would conserve the arenas of biological activity, rather than the temporary occupants of those arenas. Integrative, context-sensitive variables, such as insolation and topographic wetness, are useful for defining land facets. Classification procedures such as k-means or fuzzy clustering are a good way to define land facets because they can analyze millions of pixels and are insensitive to case order. In regions lacking useful soil maps, river systems or riparian plants can indicate important facets. Conservation planners should set higher representation targets for rare and distinctive facets. High interspersion of land facets can promote ecological processes, evolutionary interaction, and range shift. Relevant studies suggest land-facet diversity is a good surrogate for today's biodiversity, but fails to conserve some species. To minimize such failures, a reserve design based on land facets should complement, rather than replace, other approaches. Designs based on land facets are not biased toward data-rich areas and can be applied where no maps of land cover exist.
即使在最乐观的情况下，下个世纪人为引起的气候变化也将威胁到许多野生动物种群和物种。最有效的保护对策是扩大和连接保护区，以支持动植物的分布转移。为了优先为保护区和连接区选择土地，一些科学家试图将四个高度不确定的模型（排放情景、全球海洋-大气环流、区域环流和生物响应）串联起来。这种方法存在错误传播和累积的高风险，并且输出的结果比保护决策的规模更粗糙。相反，我们提倡确定土地特征——具有统一地形和土壤属性的反复出现的景观单元，并为这些单元的多样性和散布设计保护区和连接区。这种粗滤方法将保护生物活动的场所，而不是这些场所的临时占据者。综合性、敏感于上下文的变量，如太阳辐射和地形湿润度，对于定义土地特征很有用。分类程序，如 k-均值或模糊聚类，是定义土地特征的一种很好的方法，因为它们可以分析数百万个像素，并且对案例顺序不敏感。在缺乏有用的土壤图的地区，河流系统或河岸植物可以指示重要的特征。保护规划者应该为稀有和独特的特征设定更高的代表性目标。土地特征的高度散布可以促进生态过程、进化相互作用和分布转移。相关研究表明，土地特征多样性是当今生物多样性的良好替代物，但不能保护某些物种。为了最大限度地减少这种失败，基于土地特征的保护区设计应该是其他方法的补充，而不是替代。基于土地特征的设计不会偏向于数据丰富的地区，并且可以应用于没有土地覆盖地图的地区。
