Local ecological knowledge (LEK) has been increasingly invoked in biodiversity monitoring and conservation efforts. Although methods involving LEK have become more widespread in ecology, it remains an undervalued source of information in understanding the ecology of wildlife in the context of human-wildlife conflict. People who regularly interact with wildlife, and often with notable consequences, as is the case with human-wildlife conflict, will likely build up ecological knowledge of that species. We gathered LEK on the landscape use of the African elephant (Loxodonta Africana) in a region where its range overlaps with human land use and results in conflict, the western Okavango Panhandle of Botswana. We interviewed community-defined local experts and used participatory ranking activities to gather information on landscape use of elephants. The scores from the rankings were then incorporated with environmental data following resource selection function methods common in ecology. The resulting LEK-based model had high predictive ability for elephant locations when modeled at a local scale (25 km, Spearman's rho = 0.98, P < 0.0001). We also calculated resource selection models using elephant telemetry data combined with the same environmental data as the LEK models. These models showed a complementary pattern, with better predictive ability at the regional scale (Spearman's rho = 0.98, P < 0.0001) than at the local scale (rho = 0.92, P < 0.0031). In addition to being used for the resource selection functions, each method provided different kinds of information on elephant landscape use. Our results support the use of LEK as a tool for understanding local patterns of wildlife landscape use in the context of human-wildlife conflict, where the knowledge can be used to complement other data across scales and the use of which can itself contribute to better conservation outcomes.

中文翻译：

---

专家与大象：当地生态知识可预测参与人类与野生动物冲突的物种的景观用途

当地生态知识 (LEK) 越来越多地被用于生物多样性监测和保护工作。尽管涉及 LEK 的方法在生态学中变得越来越普遍，但在人类与野生动物冲突的背景下，它仍然是理解野生动物生态的一种被低估的信息来源。经常与野生动物互动并经常产生显着后果（例如人类与野生动物冲突的情况）的人可能会建立对该物种的生态知识。我们就非洲象 (Loxodonta Africana) 在博茨瓦纳西部 Okavango Panhandle 的活动范围与人类土地使用重叠并导致冲突的地区的景观使用情况收集了 LEK。我们采访了社区定义的当地专家，并使用参与式排名活动来收集有关大象景观利用的信息。然后按照生态学中常见的资源选择函数方法将排名的分数与环境数据相结合。在局部范围内建模时，基于 LEK 的模型对大象位置具有很高的预测能力（25 公里，Spearman's rho = 0.98，P < 0.0001）。我们还使用大象遥测数据结合与 LEK 模型相同的环境数据计算资源选择模型。这些模型显示出互补模式，在区域尺度（Spearman's rho = 0.98，P < 0.0001）比在局部尺度（rho = 0.92，P < 0.0031）具有更好的预测能力。除了用于资源选择功能外，每种方法还提供了有关大象景观使用的不同类型的信息。