Wildlife Institute, School of Nature Conservation, Beijing Forestry University, Beijing 100083, China; E-mail:
Everest Snow Leopard Conservation Center, Rikaze Tibet 857000, China.
Zool Res. 2018 Nov 18;39(6):373-386. doi: 10.24272/j.issn.2095-8137.2018.057. Epub 2018 May 24.
Habitat evaluation constitutes an important and fundamental step in the management of wildlife populations and conservation policy planning. Geographic information system (GIS) and species presence data provide the means by which such evaluation can be done. Maximum Entropy (MaxEnt) is widely used in habitat suitability modeling due to its power of accuracy and additional descriptive properties. To survey snow leopard populations in Qomolangma (Mt. Everest, QNNR) National Nature Reserve, Tibet, China, we pooled 127 pugmarks, 415 scrape marks, and 127 non-invasive identifications of the animal along line transects and recorded 87 occurrences through camera traps from 2014-2017. We adopted the MaxEnt model to generate a map highlighting the extent of suitable snow leopard habitat in QNNR. Results showed that the accuracy of the MaxEnt model was excellent (mean AUC=0.921). Precipitation in the driest quarter, ruggedness, elevation, maximum temperature of the warmest month, and annual mean temperature were the main environmental factors influencing habitat suitability for snow leopards, with contribution rates of 20.0%, 14.4%, 13.3%, 8.7%, and 8.2% respectively. The suitable habitat area extended for 7001.93 km2, representing 22.72% of the whole reserve. The regions bordering Nepal were the main suitable snow leopard habitats and consisted of three separate habitat patches. Our findings revealed that precipitation, temperature conditions, ruggedness, and elevations of around 4000 m influenced snow leopard preferences at the landscape level in QNNR. We advocate further research and cooperation with Nepal to evaluate habitat connectivity and to explore possible proxies of population isolation among these patches. Furthermore, evaluation of subdivisions within the protection zones of QNNR is necessary to improve conservation strategies and enhance protection.
生境评价是野生动物种群管理和保护政策规划的重要基础步骤。地理信息系统（GIS）和物种存在数据为进行这种评价提供了手段。最大熵（MaxEnt）由于其准确性和附加描述性属性而被广泛用于栖息地适宜性建模。为了调查中国西藏珠穆朗玛峰（珠穆朗玛峰，QNNR）国家自然保护区的雪豹种群，我们在 2014 年至 2017 年期间沿线路收集了 127 个爪印、415 个刮痕和 127 个动物的非侵入性识别，并通过相机陷阱记录了 87 次出现。我们采用 MaxEnt 模型生成了一张突出显示 QNNR 中雪豹适宜栖息地范围的地图。结果表明，MaxEnt 模型的准确性非常好（平均 AUC=0.921）。最干燥季度的降水量、崎岖度、海拔、最温暖月份的最高温度和年平均温度是影响雪豹栖息地适宜性的主要环境因素，贡献率分别为 20.0%、14.4%、13.3%、8.7%和 8.2%。适宜栖息地面积扩展到 7001.93 平方公里，占保护区总面积的 22.72%。与尼泊尔接壤的地区是雪豹的主要适宜栖息地，由三个独立的栖息地斑块组成。我们的研究结果表明，降水、温度条件、崎岖度和海拔约 4000 米影响了 QNNR 景观水平的雪豹偏好。我们主张与尼泊尔进一步合作开展研究，评估栖息地连通性，并探索这些斑块之间种群隔离的可能替代指标。此外，有必要评估 QNNR 保护区内的细分区域，以改进保护策略并加强保护。
