Probability of Synanthropic Feral Dog Presence in the Western United States

This model is based on how dogs utilize wildlands near human habituation. These predators can have detrimental effects on wildlife populations (Alterio et al. 1998). We based our model on the data collected by Odell and Knight (2001) that investigated habitat utilization of these predators with regard to distance from housing and on the probability for a homeowner to possess a dog. We buffered the both the populated areas and the campground distance layers in ARC/INFO using probability functions [P = 0.548 - 1.4589 * Distance (km)]. Any cell with distance less than 0.36km received a probability based on the function (0.556 to 0.001572) and all distances greater than or equal to 0.36km from populated areas or campgrounds were assigned a probability of 0. We combined the two models into the dog model by selecting the maximum value at each pixel location from the 2 models using the MAX command in ARC/INFO. The resulting dataset was then resampled to 180m using the bilinear interpolation option.

本模型基于犬类对人类聚居区附近荒野地带的利用模式。这类捕食者会对野生动物种群产生不利影响（Alterio等，1998）。我们的模型构建依据Odell与Knight（2001）的调研数据，该研究围绕这类捕食者的栖息地利用模式（以距住宅的距离为量化指标）以及屋主饲养犬只的概率展开。我们使用概率函数[P = 0.548 - 1.4589 × 距离（km）]，在ARC/INFO中对聚居区与露营地的距离图层进行缓冲区分析。对于距离小于0.36km的栅格单元，依据该函数计算得到其概率值（取值范围为0.556至0.001572）；而距聚居区或露营地的距离大于或等于0.36km的所有栅格单元，则被赋值为0。我们通过ARC/INFO中的MAX命令，选取两个模型在每个像元位置的最大值，将二者合并为犬类模型。随后，利用双线性插值选项将生成的数据集重采样至180m分辨率。