【作者】 薛亚东；

【导师】 李迪强；

【作者基本信息】 中国林业科学研究院 ， 野生动植物保护与利用， 2014， 博士

【摘要】 野骆驼(Camelus ferus)是亚洲中部对极端干旱环境具有高度适应性的珍稀濒危动物。自19世纪中期以来，野骆驼分布范围不断缩小，种群数量日渐减少。由于野骆驼生性机警，且在远离人迹、自然条件极端恶劣的荒漠、半荒漠地区分布等特点，野骆驼种群生态学、行为生态学基础研究较为缺乏，对其分类地位和进化关系仍有争论。野骆驼分布区面临着道路建设、矿产开发以及畜牧业的发展等威胁，生境丧失和破碎化严重。人为干扰下野骆驼异质种群动态及其对破碎化生境的响应是急需研究的科学问题。本研究在库姆塔格地区，通过野外样线、样方调查和相机陷阱调查，同时结合物种分布概率模型和生境适宜性评价模型，对影响野骆驼分布和迁徙的三个重要环境因子，水源、植被、干扰动态进行了研究。通过水源点红外相机监测，结合野外调查数据，研究野骆驼的种群分布和群体结构变化，分析野骆驼水源利用规律，探讨荒漠地区野生动物水源利用生态位竞争和分化。通过研究野骆驼的生境选择，分析影响其生境选择的关键因素，以确定野骆驼的生境选择特征，从而进行生境适宜性分析。在此基础上将野骆驼适宜生境和人为干扰因素叠加，构建野骆驼生境连通性模型，探讨景观要素和格局对其迁移的影响，辨识影响野骆驼生境连通及阻碍其迁徙的关键威胁，为野骆驼种群及其生境保护管理提供对策。主要结论如下：（1）在阿尔金山北麓地区，红外相机共记录到26个物种，包括兽类11种，鸟类15种。兽类主要包括野骆驼、藏野驴、鹅喉羚、岩羊、盘羊、狼、赤狐、雪豹、豺、猞猁等；鸟类主要包括胡兀鹫、高山兀鹫、金雕、秃鹫、大鵟，石鸡等。野骆驼在每个水源地都被拍到，是食草动物中分布最为广泛的物种。野骆驼是水源地占优势地位的物种，其在水源地出现的强度和时间，一定程度上决定了其他食草动物的水源利用时间分配。狼是该地区主要的食肉动物，也是包括野骆驼在内食草动物的主要捕食者。各主要物种的相对丰富度按从高到低排序依次为野骆驼（72.94）＞藏野驴（17.51）＞鹅喉羚（4.68）＞赤狐（4.37）＞盘羊（2.92）＞岩羊（2.82）＞狼（1.56）。在西湖湿地共记录到6种兽类，分别为野骆驼，鹅喉羚、狼、赤狐、猞猁、草兔。257张独立照片中有191张为野骆驼，食肉动物中，以猞猁的照片数较多。两个地区相比，阿尔金山北麓的物种多样性要高于西湖湿地，阿尔金山北麓（72.94）的野骆驼相对丰富度要远高于西湖湿地（6.15）。（2）2011-2013年，在库姆塔格沙漠地区进行了8次野外调查。阿尔金山北麓地区野骆驼分布最为集中，个体数最大的群体均分布在该地区。8次野外调查共记录野骆驼64群，个体430峰。5月-9月野骆驼群个体数平均值为2.94；而10月-4月野骆驼群个体数平均值为10.74，在此期间的记录到的野骆驼数量明显多于5月-9月。野骆驼活动以单独活动或者小群体活动为主，在5月-9月期间尤其明显。通过分析野外观测数据，群个体数平均值、个体数≤3峰的群体数和野骆驼集群规模三个指标表明，野骆驼群体数量存在季节性差异，倾向于在10月-4月间集成大规模的群体。物种分布概率模型表明干扰期/非干扰期因素和雨季/非雨季因素对包括野骆驼在内的7个主要物种的分布影响最大，人为干扰和水源是影响荒漠地区物种分布的主要因素。（3）野骆驼日间对水源的利用率要高于夜间，其水源利用以昼行性为主。野骆驼在水源点的相对活动强度呈明显的单峰型，活动的最高峰出现在10:00-11:00。其活动强度从日出开始逐渐增强，5:00-11:00呈上升趋势。然后活动强度渐渐降低，11:00-18:00呈下降趋势，至日落时分降至最低。夜间18:00-5:00活动趋于稳定。野骆驼各月份水源点活动天数存在显著差异。野骆驼7月份出现天数最少(=2.6，SE=1.54)。活动天数最多的月份出现在4月份(=17.57，SE=2.63)，其次为10月份的14.25天。野骆驼冬季水源利用强度和频率均高于夏季。野骆驼夏季，即6月-9月水源点相对活动强度最低。分别在4月和10月季节过渡时期出现了水源利用的高峰期。冬季11月-3月相对活动强度要高于夏季。野骆驼4个季节的日活动模式没有显著差异。（4）在300m和30m的尺度上，MAXENT模型的预测结果均达到优秀水平。两个尺度上距水源距离、距采矿点距离、距居民点距离、海拔是影响野骆驼生境分布的主要环境因子。4个环境变量的累积贡献率分别达87.5%和93%，表明这4个环境因子对野骆驼的生境选择具有重要影响，其中距水源距离因子贡献率最大。野骆驼偏好海拔800m-2000m左右的生境，并且偏好坡度小的地带；野骆驼对水源的利用表现了其对安全因素的权衡；对人为干扰因子的响应表明，人类活动已经深入野骆驼的适宜生境。随着人为活动的加剧，野骆驼的适宜生境被进一步挤压。野骆驼适宜生境面积为5068.7km2，野骆驼在库姆塔格沙漠地区的适宜生境主要分为三个独立的区块，分别是阿尔金山北麓、阿奇克谷地和西湖湿地。其中以阿尔金山北麓适宜生境分布最大，面积最大的斑块为3581.9km2；阿奇克谷地适宜生境中最大斑块的面积为481.8km2；西湖湿地适宜生境中最大斑块的面积为235.5km2。野骆驼生境的连通性受沙漠北部探险旅游活动干扰较大，因此应在阿奇克谷地加强保护区巡护执法，严格管控探险旅游等行为，尤其是在野骆驼产羔季节。（5）对适应性放归野骆驼进行了一年的的GPS颈圈监测，MCP家域表明适应性放归的野骆驼在6月份活动范围较小，从8月份开始，其活动范围基本扩张至围栏面积的一半，随后，野骆驼活动范围呈现逐渐扩大的趋势，直至覆盖整个围栏的范围，一年内野骆驼的活动面积为22.8km2（108444）和21.9km2（108445）。散养条件下野骆驼行为时间分配和活动节律表现出与大多数有蹄类类似的策略，包括时间分配上采食行为时间比例较高，存在晨昏活动高峰。随时间的推移，野骆驼对放归生境逐步适应，因此表现出休息、警戒的下降趋势，而随着冬季气温的不断降低，野骆驼为获取能量，表现出运动时间和采食时间的上升趋势。对于生态环境脆弱的荒漠、半荒漠地区，全球气候变化带来的极端高温、干旱事件，必然造成水资源的更为稀缺，基于水源管理和保护的野生动物保护措施将更为重要。近年来，阿尔金山地区采矿活动日益增加，道路扩建带来的生境破碎化和水源侵占现象十分严重，对野生动物带来的影响更值得进一步关注。水源地相机陷阱的使用，可以实现动物水源利用的动态监测，将有助于评估环境变化和人为干扰，尤其是家畜利用水源从而带来的疾病传播对野生动物活动的影响，并应与该地区水源管理和保护措施的制定相结合，为该地区野生动物的研究和保护提供具体的参考信息。野生动物的管理和保护对策往往依赖于对动物迁移机制的理解来制定，包括评估人类活动及干扰对种群的影响，物种廊道设计，提高景观连接度等。分布区内人为干扰严重影响了野骆驼的栖息和迁徙，所造成的栖息地生境丧失及破碎化已成为影响野骆驼种群生存的最直接的威胁。野骆驼种群间的迁徙，以及栖息地的连接度研究将有助于种群基因交流，对保持整个种群的遗传多样性有重要意义。通过野骆驼生境连接度研究，明确了其潜在迁徙活动路线，为避免关键威胁制定相应的管理策略提供了科学依据。更多还原

【Abstract】 Wild Bactrian camel (Camelus ferus） is a critically endangered species, which has highadaptability in extreme arid environments of Central Asia. Wild camel population anddistribution areas have decreased significantly since middle of19th century. Most researchesabout wild camel were difficult as the species was cryptic and distributed in arid and semi-aridareas. There are great knowledge gaps concerning the ecology of the wild camels. Even thegenetic status and the purity of wild camel populations have not yet been fully resolved andpreliminary results need to be treated with caution. In recent years, mineral activity, roadconstruction and livestock development have been increasing, which caused habitat loss andfragmentation of wild camel population. It’s urgent to focus on metapopulation dynamic and itsresponse to habitat fragmentation under human disturbance.In this paper, various methods and technics were used to study the ecology of wild camel,including field survey, camera traps, occupancy estimate, and niche models. Water, vegetation,and human disturbance were three important factors that influence the distribution of wildcamel in Kumtag Desert region. The water utilization and habitat selection of wild camel werestudied to identify the key environmental factors, and futher to analyse the suitable habitatdistribution. Habitat connectivity of wild Bactrian camel in Kumtag Desert region wereanalyzed based on suitable habitat and human disturbance. Meanwhile, we simulated thepotential migration path and identified the sensitive area of wild camel. Then the influence oflandscape pattern and key threat on wild camel migration were evaluated. The mainconclusions are as follows:(1) To examine the diversity and abundance of wildlife at water source, camera-trappingwas conducted at seven sites from2010to2012in the northern slope of Altun Mountains.Totally26species were captured in the survey. There are11species of mammal includingCamelus ferus, Equus kiang, Uncia uncia, Lynx lynx, and15species of birds including Gypaetus barbatus, Aquila chrysaetos, and Gyps himalayensis in photos. Wild camel (Camelusferus) was the only species captured at all observation sites and was photographed in almost66%of the total pictures. It suggested that wild camel is the dominant species of herbivore atwater source area, the time and intense of wild camel occurrence at water sources play animportant role in other herbivores’ time budget at water source. For carnivores, wolf (Canislupus) and red fox (Vulpes vulpes) were the dominant species of carnivores at water sourcesince they had been captured in123and268pictures respectively. In the Dunhuang Xihuwetland, six species were recorded with camera traps at four water sites,including Camelusferus, Gazella subgutturosa, Canis lupus, Vulpes vulpes, Lynx lynx, Lepus capensis.There were191independent pictures of wild camel in the total257pictures. For carnivores, Lynx lynx wasthe dominant species. The species diversity in the northern slope of Altun Mountains washigher than Dunhuang Xihu wetland. And the relative abundance of wild camel was muchhigher in the northern slope of Altun Mountains（72.94） than that in Dunhuang Xihu wetland（6.15）.(2) During2011-2013, eight field surveys were conducted in Kumtag desert, and observedindividual and population of wild camel were recorded to evaluate the group in differentseasons.We recorded430wild Bactrian camel individuals in a total of64groups, the largestgroup of71, the least group of1. The average group size is10.74during October-April, and2.94during May-September. Our data on group membership or stability of camel groups,including average group size, numbers of group less than three individuals, and maximumgroup size supported that wild camels live in open fission fusion groups, which tend toconcentrate during the rutting season in winter. Occupancy models showed that disturbanceand rain season were the most important factors influencing the occupancy of wild camel andother species.(3) In the northern slope of Altun Mountains, our camera monitoring efforts ranged from224to1,830camera trapping days for a total of7,098camera trapping days for the studyduration. The proportion of relative activity at watering sites differed by monitoring interval (F23,144=8.870, P<0.001). Wild camels used watering sites all through the day, but visitedmore frequently during the daytime (0700–1700,68.69%). The relative activity revealed anincreasing trend in the morning (0500–1100), and showed a decreasing trend during theafternoon (1100–1800) and evening (1800–0500). The proportion of the relative activity atwatering sites was greatest from1000to1100(=8.33, SE=0.76). The relative activity of wildcamels was highest during the transition seasons (April and October) and lowest in July. Thewild camels visited watering sites in differing ways in different months (F11,63=3.385,P=0.001). Wild camels visited watering sites rarely in July (=2.6days, SE=1.54), but maderegular visits to watering sites in April (=17.57days, SE=2.63). The present days of wildcamels at watering sites in winter was5.8±0.3higher than in summer (t=3.733, df=56,P<0.001).(4) The method of Receiver Operating Characteristic (ROC) curve analysis was used toassess the accuracy of MAXENT model. In the300m and30m scale, the area under ROCcurve (AUC) of the model was0.961, indicating that the result of assessment was excellent.The result of Jackknife test indicated that the distance to water, the distance to mining sites, thedistance to residence, and altitude were the main factors influencing habitat selection of wildcamel. The accumulating contribution of these four factors were87.5%and93%respectively.The suitable habitat of wild camel predicted by MAXENT is5068.7km2, and divided to threeseperated patches, the northern slope of Altun Mountains, the Aqike valley, and the Xihuwetland. The suitable habitat of the northern slope of Altun Mountains covers the most of thesuitable habitat with an area of3581.9km2.The habitat connectivityof wild camel wasinfluenced by exploring travel in the northern edge of the desert.(5) MCP home range analysis showed that at the beginning (June-July), the home range ofreleased wild camel covers a small area around the old enclosure. Till August, the individualhome range began to expand, from8.89km2to19.42km2gradually. During the monitoringperiod, the individual home range covers the whole enclosure, with95%Kernel home rangesof4.69km2(108444) and3.03km2(108445) respectively. The intensive occupancy also indicated the habitat selection of wild camels. As time passed, diurnal time spent on resting,vigilance and rumination showed decreasing trend, on the contrary, time spent on moving andforaging expressed increasing trend. Among different months, the resting, vigilance andforaging behavior of females were significant different (P <0.05). The behavioral time budgetand diurnal activity rhythm of wild camels during initial releasing period expressed the similarstrategy like most of ungulates. These results will be helpful to understand wild camels’behavioral ecology, and provide a guide for the conservation and management of wild camels.We recommend combining climate data in any future study to obtain a more accurateassessment of the effects of environmental factors on animal activity patterns. Wildlifeconservation is becoming more challenging and urgent because of extreme temperatures anddrought from persisting global climate change. Wildlife populations inhabiting desert or aridenvironments are most vulnerable under such conditions. With water becoming an ever-moresignificant limiting factor, the management of desert ungulates should consider waterdevelopments to be factors influencing various aspects of the ecology and behavior of desertungulates. In recent years, mining activities have been increasing in the northern piedmont ofthe Altun Mountains. The resultant construction of roads and use of watering sites has causedhabitat loss and fragmentation for the wild camels. Using camera traps, we could monitor theactivity of the species at water sources continually. It will be helpful to assess the effect ofclimate, environmental change and human disturbance on the frequency and seasonaldifferences of wild camels’ activities to provide the necessary information for wildlifemanagers making decisions concerning water development.更多还原