Caribbean Habitat Patch Size (Small Islands) (Southeast Blueprint Indicator)

Large areas of intact natural habitat are favorable for conservation of numerous species, including reptiles and amphibians, birds, and large mammals. Historically, many Caribbean islands, especially smaller ones, experienced dramatic habitat alteration across much of their land mass from intensive colonial agriculture (such as sugar cane cultivation) and later by urbanization and other anthropogenic modifications (Fitzpatrick et al. 2007, CEPF 2023). As a result, contiguous natural habitat patches are particularly important for supporting native species, restoring native vegetation, and maintaining ecosystem services. This indicator roughly follows the same approach for evaluating patch size used by the intact habitat cores indicator in the continental part of the Southeast Blueprint (Esri 2017), but lowers the highest patch size threshold, recognizing that the maximum patch size on small Caribbean islands is lower than in the continental Southeast and on large islands. Input Data Southeast Blueprint 2023 subregions: Caribbean Caribbean island extent and size Southeast Blueprint 2023 extent 2020 LANDFIRE Existing Vegetation Type (EVT) and Operational Roads (Roads) (v2.2.0) for Puerto Rico and the U.S. Virgin Islands; access the data for U.S. Insular Areas Mapping Steps Reclassify the LANDFIRE EVT data into natural and unnatural classes. All classes in “EVT_NAME” that start with “Quarries”, “Developed”, or “Agriculture” are considered unnatural. “Caribbean bush fruit and berries”, which captures sun coffee plantations, is also considered unnatural, while “tropical agroforestry plantation”, which captures shade coffee, is considered natural. Water is classified as “natural”. Reclassify the Caribbean island extent layer to assign all islands a value of 1 and assign the ocean a value of NoData. Multiply the reclassified LANDFIRE natural/unnatural layer and island extent data. This makes the ocean NoData, but retains freshwater, smaller salt ponds, and enclosed brackish water areas, as well as terrestrial areas on the islands. Reclassify primary, secondary, and tertiary roads in the LANDFIRE roads layer as barriers. Combine the natural/unnatural raster and roads raster to identify natural areas without roads. Use region group to find the size of patch that each pixel belongs to. Reclassify based on the final indicator values seen below. Make a small island mask using the island extent layer for “small” and “medium” islands. Use this mask to include only habitat patches within that mask, excluding large islands, which are captured in a different indicator. Clip to the Caribbean Blueprint 2023 subregion. As a final step, clip to the spatial extent of Southeast Blueprint 2023. Note: For more details on the mapping steps, code used to create this layer is available in the Southeast Blueprint 2025 Data Download or Caribbean-only Southeast Blueprint 2025 Data Download under > 6_Code. Final indicator values Indicator values are assigned as follows: 4 = Large patch (>1,000 acres) 3 = Medium patch (>100-1,000 acres) 2 = Small patch (>10-100 acres) 1 = Very small patch (≤10 acres) 0 = Developed or agriculture Known Issues Some small dirt roads serve as hard boundaries for habitat cores. While this makes sense for some species, this indicator likely underestimates the effective size of the patch for some more mobile animals. This indicator doesn’t account for variation in habitat condition within the patch. In Eastern St. Croix, where it’s particularly dry, this indicator can underestimate or overestimate patch size. LANDFIRE has trouble differentiating pasture from natural thornscrub in this area. Pasture isn’t considered natural in this indicator, so any areas of thornscrub misclassified as pasture would underestimate patch size, while areas of pasture classified as thornscrub would overestimate patch size. Some docks are overprioritized due landcover misclassification. LANDFIRE sometimes misclassifies docks as forests, so they can get included in habitat patches for this indicator. Disclaimer: Comparing with Older Indicator Versions There are numerous problems with using Southeast Blueprint indicators for change analysis. Please consult Blueprint staff if you would like to do this (email [e-mail]). Literature Cited Critical Ecosystem Partnership Fund (CEPF). 2023. Caribbean Islands – Threats. [https://www.cepf.net/our-work/biodiversity-hotspots/caribbean-islands/threats]. Esri Green Infrastructure Center, Inc. 2017. Esri Green Infrastructure Data Description. [https://www.arcgis.com/sharing/rest/content/items/5561da06127d48a982d6b91929f29666/data]. Fitzpatrick, Scott & Keegan, William. (2007). Human Impacts and Adaptations in the Caribbean Islands: an Historical Ecology Approach. Earth and Environmental Science Transactions of the Royal Society of Edinburgh. 98. 29 - 45. 10.1017/S1755691007000096. [https://www.cambridge.org/core/journals/earth-and-environmental-science-transactions-of-royal-society-of-edinburgh/article/abs/human-impacts-and-adaptations-in-the-caribbean-islands-an-historical-ecology-approach/68BFF9386CA9BDAD3FB44E5C937AC92F]. LANDFIRE, Earth Resources Observation and Science Center (EROS), U.S. Geological Survey. Published August 1, 2022. LANDFIRE 2020 Existing Vegetation Type (EVT) Puerto Rico US Virgin Islands. LF 2020, raster digital data. Sioux Falls, SD. [https://www.landfire.gov]. LANDFIRE, Earth Resources Observation and Science Center (EROS), U.S. Geological Survey. Published August 1, 2022. LANDFIRE 2020 Operational Roads Puerto Rico US Virgin Islands. LF 2020, raster digital data. Sioux Falls, SD. [https://www.landfire.gov].

大面积完整自然生境对众多物种的保护极为有利，涵盖爬行类、两栖类、鸟类以及大型哺乳类。历史上，诸多加勒比海岛（尤以小型岛屿为甚）的绝大多数陆地曾因集约化殖民农业（如甘蔗种植），以及后续的城市化与其他人为改造活动，发生了剧烈的生境改变（Fitzpatrick等人，2007；关键生态系统合作基金（Critical Ecosystem Partnership Fund，CEPF），2023）。因此，连片自然生境斑块对支撑本土物种、恢复本土植被以及维持生态系统服务而言尤为关键。本指标大体沿用了《东南蓝图》（Southeast Blueprint）大陆区域完整生境核心指标所采用的斑块大小评估方法（Esri，2017），但下调了最高斑块尺寸阈值，这是考虑到加勒比小型海岛的最大斑块规模低于美国东南部大陆区域以及大型加勒比海岛。

输入数据：
1. 《东南蓝图》2023年次区域：加勒比海
2. 加勒比海岛范围与面积
3. 《东南蓝图》2023年空间范围
4. 适用于波多黎各与美属维尔京群岛的2020年LANDFIRE现有植被类型（Existing Vegetation Type，EVT）与作业道路（Operational Roads，Roads）数据集（v2.2.0）；美国属地的相关绘图步骤数据可另行获取

绘图步骤：
1. 将LANDFIRE EVT数据重分类为自然与非自然两类。"EVT_NAME"字段中以"采石场（Quarries）"、"已开发（Developed）"或"农业（Agriculture）"开头的所有类别均视为非自然生境。其中，指代阳光咖啡种植园的"加勒比灌丛果蔬"亦被归为非自然生境，而指代遮阴咖啡种植的"热带农林业种植园"则属于自然生境。水体被归类为自然生境。
2. 对加勒比海岛范围图层进行重分类，为所有岛屿赋值为1，海洋区域赋值为无数据（NoData）。
3. 将重分类后的LANDFIRE自然/非自然图层与海岛范围数据相乘。该操作可将海洋区域设为无数据，但会保留淡水水域、小型盐塘与封闭咸淡水区域，以及岛屿上的陆地区域。
4. 将LANDFIRE道路图层中的一级、二级与三级道路重分类为障碍因子。
5. 合并自然/非自然栅格与道路栅格，以识别无道路分布的自然区域。
6. 使用区域分组工具（region group）计算每个像素所属斑块的面积。
7. 根据下述最终指标值进行重分类。
8. 利用海岛范围图层为"小型"与"中型"岛屿创建小型岛屿掩膜，仅保留该掩膜范围内的生境斑块，排除大型岛屿（大型岛屿的相关评估由其他指标完成）。
9. 将结果裁剪至《加勒比蓝图》（Caribbean Blueprint）2023年次区域范围。
10. 最后一步，将结果裁剪至《东南蓝图》2023年的空间范围。

注：如需了解绘图步骤的更多细节，可在《东南蓝图》2025数据下载或仅针对加勒比海区域的《东南蓝图》2025数据下载的>6_代码目录中获取生成该图层的代码。

最终指标值：
指标值赋值规则如下：
4 = 大型斑块（面积>1000英亩）
3 = 中型斑块（面积>100~1000英亩）
2 = 小型斑块（面积>10~100英亩）
1 = 超小型斑块（面积≤10英亩）
0 = 已开发区域或农业用地

已知问题：
部分小型土路可作为生境核心的硬边界。尽管这对部分物种而言合理，但本指标可能低估了部分移动性更强的动物的有效斑块面积。
本指标未考虑斑块内部生境质量的差异。
在圣克罗伊岛东部区域，由于气候格外干旱，本指标可能高估或低估斑块面积。LANDFIRE难以区分该区域的牧场与自然刺灌丛。本指标中牧场不属于自然生境，因此被误分类为牧场的刺灌丛区域会导致斑块面积被低估，而被误分类为刺灌丛的牧场区域则会导致斑块面积被高估。
部分码头因土地覆被分类错误而被过度优先纳入，LANDFIRE有时会将码头误分类为森林，因此它们会被纳入本指标的生境斑块中。

免责声明：与旧版指标版本的对比
使用《东南蓝图》指标进行变化分析存在诸多问题。如需开展此类分析，请咨询蓝图团队（邮箱：[e-mail]）。

参考文献：
关键生态系统合作基金（Critical Ecosystem Partnership Fund，CEPF）. 2023. 加勒比群岛——威胁因素. [https://www.cepf.net/our-work/biodiversity-hotspots/caribbean-islands/threats].
Esri绿色基础设施中心（Esri Green Infrastructure Center, Inc.）. 2017. Esri绿色基础设施数据说明. [https://www.arcgis.com/sharing/rest/content/items/5561da06127d48a982d6b91929f29666/data].
Fitzpatrick, Scott 与 Keegan, William. (2007). 加勒比群岛的人类影响与适应：历史生态学方法. 《爱丁堡皇家学会地球与环境科学汇刊》. 98: 29-45. 10.1017/S1755691007000096. [https://www.cambridge.org/core/journals/earth-and-environmental-science-transactions-of-royal-society-of-edinburgh/article/abs/human-impacts-and-adaptations-in-the-caribbean-islands-an-historical-ecology-approach/68BFF9386CA9BDAD3FB44E5C937AC92F].
LANDFIRE，美国地质调查局地球资源观测与科学中心（Earth Resources Observation and Science Center，EROS）. 2022年8月1日发布. LANDFIRE 2020年波多黎各与美属维尔京群岛现有植被类型（EVT）数据. LF 2020，栅格数字数据. 苏福尔斯，南达科他州. [https://www.landfire.gov].
LANDFIRE，美国地质调查局地球资源观测与科学中心（EROS）. 2022年8月1日发布. LANDFIRE 2020年波多黎各与美属维尔京群岛作业道路数据. LF 2020，栅格数字数据. 苏福尔斯，南达科他州. [https://www.landfire.gov].