Abstract Anthropogenic land use is a major contributing factor to global environmental change, including forest cover loss and fragmentation. Land managers need up to date spatio-temporal land use information to plan for future land uses and mitigate adverse impacts. Here, we use spatially-explicit logistic regression analysis and methods from remote sensing and geographic information systems (GIS) to model the relationship between land use factors and forest patches using information from satellite images from the Tano-Offin Forest Reserve in Ghana. The study finds that it is 3,350 times more likely for forest patches that are not within 1 km from barren grounds to be greater than the mean patch size of forest (800 m2) (CI = 297-37,785, P = 0.000). Also, the study finds that it is 2,365 times more likely for the forest patches that are not within 1 km from access roads to be greater than the mean patch size of forest (CI=245-22,776; P = 0.000). These findings suggest that to maintain larger and contiguous patches of forest cover, there is a need for a policy that prevents the creation and expansion of logging-induced barren grounds. Also, the creation of buffers along access roads will be necessary to prevent further anthropogenic influences that are likely to fragment forest cover.

中文翻译：

---

确定与加纳森林保护区森林斑块大小相关的空间明确的土地利用因素

摘要 人为土地利用是导致全球环境变化的主要因素，包括森林覆盖丧失和破碎化。土地管理者需要最新的时空土地利用信息来规划未来的土地利用并减轻不利影响。在这里，我们使用来自遥感和地理信息系统 (GIS) 的空间显式逻辑回归分析和方法，使用来自加纳 Tano-Offin 森林保护区的卫星图像信息对土地利用因素与森林斑块之间的关系进行建模。研究发现，距离贫瘠土地 1 公里以内的森林斑块大于森林的平均斑块大小 (800 平方米) 的可能性要高 3,350 倍 (CI = 297-37,785，P = 0.000)。此外，研究发现它是 2，距离通路 1 公里以内的森林斑块大于森林平均斑块大小的可能性是森林斑块的 365 倍（CI = 245-22,776；P = 0.000）。这些发现表明，为了保持更大和连续的森林覆盖面积，需要制定一项政策来防止造成和扩大因砍伐导致的荒地。此外，有必要沿通路建立缓冲区，以防止可能破坏森林覆盖的进一步人为影响。