Conservation planning for retention, not just protection

Most protected area (PA) planning aims to improve biota representation within the PA system, but this does not necessarily achieve the best outcomes for biota retention across regions when we also consider habitat loss in areas outside the PA system. Here we assess the implications that different PA expansion strategies can have on the retention of species habitat across an entire region. Using retention of forest habitat for Colombia’s 550 forest-dependent bird species as our outcome variable, we found that when a minimum of 30% of each species’ habitat was included in the PA system, a pattern of PA expansion targeting areas at highest deforestation risk (risk-prevention) led to the retention, on average, of 7.2% more forest habitat per species by 2050 than did a pattern that targeted areas at lowest risk (risk-avoidance). The risk-prevention approach cost more per km² of land conserved, but i2Deforestation Probability Rastert was more cost-effective in retaining habitat in the landscape (50 – 69% lower cost per km2 of avoided deforestation). To have the same effectiveness preventing habitat loss in Colombia, the risk-avoidance approach would require more than twice as much protected area, costing three times more in the process. PA expansion should focus on the contributions of PAs to outcomes not only within PA systems themselves, but across entire regions. Although including these maps in the repository contravenes the terms of use under which they were shared by BirdLife International, we are including them to meet the requirements of Global Change Biology and to allow our analyses to be replicated. However, BirdLife International strongly suggest that users request the most up to date versions of the maps at https://datazone.birdlife.org/species/requestdis” Files in the repository- Shapefile of the 550 forest depenedent bird species in Colombia - File name: "Colombia_Forest_Birds.shp" - Raster file of the deforestation probability for the forest in Colombia used for the deforestation models and the Marxan runs - Folder name: "Deforestation Probability Raster" - Marxan Outputs based on the different approaches and targets - Folder name: "Marxan Outputs" - Raster layers of historical, 2015 and different projections for 2050 of forest cover - Folder Name: "Forest Cover Colombia" - Oportunity Cost of Conservation raster layers for a 5%,10% and 20% discount rate - Folder name: "OCC" - Table with the results of habitat los by species by 2050 under the different approaches and scenarios - Folder name: "Species Results by Scenario and Target"大多数保护地（Protected Area, PA）规划的目标是提升保护地体系内的生物群落代表性，但当同时考虑保护地体系外区域的栖息地流失时，这一目标未必能在跨区域层面实现生物群落留存的最优结果。本研究评估了不同保护地扩张策略对全域物种栖息地留存的影响。本研究以哥伦比亚550种依赖森林的鸟类的森林栖息地留存率作为因变量，结果显示：当每个物种至少30%的栖息地被纳入保护地体系时，以森林砍伐高风险区域为目标的保护地扩张策略（风险防控型），到2050年时，每个物种的平均森林栖息地留存量比以低风险区域为目标的策略（风险规避型）高出7.2%。每保护1平方公里土地，风险防控型策略的成本更高，但其在景观栖息地留存方面的成本效益更优（每避免1平方公里森林砍伐的成本降低50%~69%）。若要在哥伦比亚实现同等的栖息地流失防控效果，风险规避型策略所需的保护地面积需超过前者的两倍，且总成本将达到前者的三倍。保护地扩张不应仅关注保护地体系内部的成效，而应聚焦于保护地对全域范围内相关成果的贡献。 尽管将这些地图纳入本数据仓库违反了国际鸟类联盟（BirdLife International）分享它们时所遵循的使用条款，但为满足《全球变化生物学》（Global Change Biology）的刊发要求并确保本研究分析可被复现，我们仍将其纳入仓库。不过，国际鸟类联盟强烈建议用户通过https://datazone.birdlife.org/species/requestdis 申请获取这些地图的最新版本。 本数据仓库包含以下文件： - 哥伦比亚550种依赖森林鸟类的矢量形状文件（Shapefile）——文件名为："Colombia_Forest_Birds.shp" - 用于森林砍伐模型及Marxan分析的哥伦比亚森林砍伐概率栅格文件（Raster File）——文件夹名为："Deforestation Probability Raster" - 基于不同策略与目标的Marxan输出结果——文件夹名为："Marxan Outputs" - 包含历史、2015年及多种2050年森林覆盖预测情景的栅格图层——文件夹名为："Forest Cover Colombia" - 针对5%、10%及20%贴现率的保护机会成本栅格图层——文件夹名为："OCC" - 包含不同策略与情景下到2050年各物种栖息地流失结果的数据表——文件夹名为："Species Results by Scenario and Target"