Division of Forest and Biomaterials Sciences, Graduate School of Agriculture, Kyoto University, Kyoto, 606-8502, Japan.
Department of Systems Innovation, Graduate School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-8656, Japan.
J Environ Manage. 2021 Feb 15;280:111711. doi: 10.1016/j.jenvman.2020.111711. Epub 2020 Nov 30.
The establishment of protected areas has been considered a tool to battle deforestation in densely populated countries like Bangladesh. Since 2000, Bangladesh's Forest Department has declared about 41 protected areas. However, before more land is designated as a protected area, it is important to know how effective existing protected areas are in achieving conservation goals. Unfortunately, the determination of the conservation effectiveness of protected areas, such as their capacity to reduce deforestation and forest fragmentation over a considerable period at high temporal frequency (e.g., yearly), is still unavailable despite some known methods being available, for example, the System for the Integrated Assessment of Protected Areas. In this study, we processed and analyzed the Hansen dataset from 2000 to 2018 to produce yearly forest/non-forest maps of four protected areas in Bangladesh and used these maps, with a matching method, to estimate the effectiveness of protected area in reducing deforestation after controlling for potential hidden bias. We also analyzed the forest fragmentation scenario over the same time frame. The forest cover change results from 2010 to 2018 revealed a large-scale deforestation pattern in areas adjacent to the protected area boundary of Chunati Wildlife Sanctuary (CWS) and Baroiyadhala National Park-Hazarikhil Wildlife Sanctuary (together B-HWS). Using a propensity score matching (PSM) approach with a caliper of 0.25, we found that B-HWS was the best performing of the studied protected areas, and that 37% of forest pixels in B-HWS would have been deforested in 2018 if they had not been brought under protection in 2010. Similarly, the estimated avoided deforestation rates were approximately 21% and 4% for CWS and Dudpukuria-Dhopachari Wildlife Sanctuary, respectively. Despite an improvement in deforestation scenarios, during the period 2010-2018, for all protected areas, the forest fragmentation scenarios were exacerbated both inside their boundaries and in adjacent unprotected areas. Therefore, it remains questionable whether protected areas can ultimately maintain the integrity of conservation.
建立保护区被认为是对抗人口密集的国家如孟加拉国森林砍伐的一种手段。自 2000 年以来，孟加拉国林业部已宣布设立约 41 个保护区。然而，在指定更多土地作为保护区之前，了解现有保护区在实现保护目标方面的效果是很重要的。不幸的是，尽管有一些已知的方法，如综合保护区评估系统（System for the Integrated Assessment of Protected Areas），但仍然无法确定保护区的保护效果，例如它们在相当长的时间内（例如每年）减少森林砍伐和森林破碎化的能力。在这项研究中，我们处理和分析了 2000 年至 2018 年的 Hansen 数据集，生成了孟加拉国四个保护区的每年森林/非森林地图，并使用这些地图和匹配方法，在控制潜在隐藏偏差的情况下，估计保护区减少森林砍伐的效果。我们还分析了同一时间段的森林破碎化情况。2010 年至 2018 年的森林覆盖变化结果显示，在 Chunati 野生动物保护区（CWS）和 Baroiyadhala 国家公园- Hazarikhil 野生动物保护区（B-HWS）保护区边界附近的区域存在大规模的森林砍伐模式。使用倾向得分匹配（PSM）方法和 0.25 的卡尺，我们发现 B-HWS 是研究保护区中表现最好的，并且如果 2010 年不将 B-HWS 纳入保护区，那么保护区内的 37%的森林像素将在 2018 年被砍伐。同样，CWS 和 Dudpukuria-Dhopachari 野生动物保护区的估计避免砍伐率分别约为 21%和 4%。尽管在 2010-2018 年期间，所有保护区的森林砍伐情况都有所改善，但保护区内及其周边未受保护地区的森林破碎化情况都有所加剧。因此，保护区是否最终能够保持保护的完整性仍然值得怀疑。
