Abstract One of the challenges of working with rare or elusive species is knowing their current distribution. Species distribution modeling (SDM) is a robust approach to estimate the distribution of a species where this information is unknown. The Andean cat (Leopardus jacobita) is an extremely rare and endangered carnivore living in the central Andes of South America. This study sought to determine priority areas for the conservation of the species in the dry puna of Chile using fine-scale SDM approaches. The potential distribution of the Andean cat was estimated through Maxent and random forest modeling algorithms. The predictive variables with the greatest contribution to the distribution models included three related to temperature, one to precipitation, one to the distance to wetlands and finally the topographic position index. The total suitable area predicted for the Andean cat was 923.4 km2, which showed a highly fragmented pattern. Based on the information generated by the distribution models, its threats and formal protection, four priority areas were defined. This information will be useful for guiding and prioritizing future actions towards the conservation of the Andean cat in northern Chile.

中文翻译：

---

智利干普纳濒危安第斯猫（Leopardus jacobita）保护区的精细规模方法

摘要 处理稀有或难以捉摸的物种的挑战之一是了解它们的当前分布。物种分布建模 (SDM) 是一种可靠的方法，用于估计此信息未知的物种的分布。安第斯猫 (Leopardus jacobita) 是一种极为罕见和濒临灭绝的食肉动物，生活在南美洲安第斯山脉中部。本研究旨在使用精细规模的 SDM 方法确定智利干普纳物种保护的优先领域。安第斯猫的潜在分布是通过 Maxent 和随机森林建模算法估计的。对分布模型贡献最大的预测变量包括三个与温度相关的变量，一个与降水相关的预测变量，一个与到湿地的距离相关的预测变量，最后是地形位置指数。预测的安第斯猫的总适宜面积为 923.4 平方公里，呈现高度分散的格局。根据分发模型生成的信息、其威胁和正式保护，定义了四个优先领域。这些信息将有助于指导和确定智利北部安第斯猫保护未来行动的优先顺序。