Abstract: Ursus arctos pruinosus is a unique subspecies of brown bear found in the Qinghai-Tibet Plateau, which serves as an umbrella and flagship species for biodiversity conservation. To manage and protect this species, it is very important to understand the factors that determine its suitable habitat distribution, as well as its daily activity patterns. From July 2021 to July 2022, a monitoring program was conducted in Dulan County, Qinghai Province, using 150 infrared camera traps to survey the Ursus arctos pruinosus. The MaxEnt model was used to predict its historical and future habitat distributions by combining three topographic variables (elevation, slope, and aspect) as well as 19 bioclimatic variables using data from the base period of 1970 - 2000 and the future climate of 2040 - 2060 projected by the Coupled Model Intercomparison Project Phase 6 (CMIP6) under the medium development path. Based on the data captured by the infrared cameras, the daily activity patterns of Ursus arctos pruinosus were estimated using kernel density estimation. The results showed that the MaxEnt model has a high predictive accuracy for the suitable habitat distribution of Ursus arctos pruinosus in both the historical and projected future time periods, with the area under the receiver operating characteristic curve exceeding 0. 9, indicating reliable predictions. The main environmental factors affecting the habitat distribution of Ursus arctos pruinosus were elevation, minimum temperature of coldest month, and annual temperature range. Under projected future climate conditions, there will be a slight increase in the unsuitable habitat area for Ursus arctos pruinosus and the highly suitable habitat in Gouli Township will shift eastward. We found that the daily activity of Ursus arctos pruinosus mainly occurs at night following a crepuscular pattern, with peak activity occurring around sunset. This study identified the spatial distribution of Ursus arctos pruinosus in Dulan County and the factors influencing it, and then predicted future trends in this distribution. Given the unique daily activity patterns of Ursus arctos pruinosus compared to other brown bear subspecies, we believe that the region may face more severe human-bear conflict in the future and should be given more attention in conservation work.

Key words: Climate change, MaxEnt model, Kernel density curve, Human-wildlife conflict, Brown bear

摘要： 棕熊青藏亚种 (Ursus arctos pruinosus) 是青藏高原特有，生物多样性保护的伞护种和旗舰种，明确其适宜栖息地分布及影响因子和日活动节律，对管理和保护工作具重要意义。2021 年 7 月至 2022 年 7 月，在青海省都兰县布设 150 台红外相机对棕熊开展调查监测工作。结合海拔、坡度、坡向 3 个地形数据及基准期 (1970—2000 年) 和CMIP6 中等发展路径下未来气候 (2040—2060 年) 的 19 个生物气候数据，构建 MaxEnt 模型对棕熊基准期及未来的栖息地分布进行预测；根据红外相机拍摄时间数据，利用核密度估计法分析棕熊的日活动节律。结果表明：(1) MaxEnt 模型对基准期和未来两个时期的棕熊适宜栖息地预测精度较高，受试者操作特征曲线下面积值均大于0. 9，预测结果可信；(2) 影响基准期棕熊栖息地分布的主要环境因子为海拔、最冷月最低气温和年均气温变化范围；(3) 未来气候条件下棕熊不适宜栖息地面积小幅度增加，沟里乡的高适宜栖息地有向东移动趋势；(4) 棕熊的活动模式表现为以夜间为主的晨昏型，在日落前后活动强度最高。本研究探明了研究区域内棕熊空间分布格局及影响因子，并预测了其未来变化趋势。结合棕熊青藏亚种不同于其他亚种种群的日活动节律特征，我们认为未来该区域内可能面临更加严峻的人-熊冲突状况，应在后续的保护工作中予以重视。

关键词: 气候变化, MaxEnt模型, 核密度曲线, 人-兽冲突, 棕熊