【作者】 张常智；

【导师】 张明海；

【作者基本信息】 东北林业大学 ， 野生动植物保护与利用， 2013， 博士

【摘要】 2002-2013年应用大样方法、样带法、传统的俄罗斯有蹄类调查方（FMP法）、网络信息收集法、样线法、样线片段足迹链占有法等野外抽样调查技术,结合Bootstrap重抽样、蒙特卡洛模拟、多元统计分析和地理信息系统技术等,对东北虎及其猎物种群数量监测方法评估、东北虎及其猎物种群现状及变动趋势、猎物种群监测方案设计、虎及其猎物间关系、东北虎猎物种群恢复型和人虎冲突开展研究,获得如下主要结果：1)东北虎猎物种群数量监测方法评估。传统的样带法（样带宽度为100米）,猎物密度为：D（马鹿）=0.054±0.009只/km2,D（狍）=2.81±0.72只/km2,D（野猪）=0.99±0.16只/km2,D（梅花鹿）=0.26±0.06只/km2；改进后的样带法（样带宽度为足迹链宽带,即为昼夜活动直径）,猎物密度为：D（马鹿）=0.004±0.0017只/km2,D（狍）=0.32±0.07只/km2,D（野猪）=0.33±0.05只/km2,D（梅花鹿）=0.0094±0.002只/km2。传统的俄罗斯调查法（FMP法）,猎物密度为：D（马鹿）=0.0096±0.005只/km2；D（狍）=0.93±0.24只/km2；D（野猪）=0.058±0.013只/kn2；D（梅花鹿）=0.0194±0.009只/km2。大样方调查法,猎物密度为：D（马鹿）=0.006±0.007只/km2,D（狍）=0.654±0.148只/km2,D（野猪）=0.311±0.154只/km2,D（梅花鹿）=0.074±0.042只/km2。样带法调查误差主要来源于样带的宽度的确定和换算系数的准确性。传统样带宽度以透视度（远小于足迹链宽度）来计算密度,导致密度远高于实际密度。采用足迹链宽度作为样带宽度,用一条样带调查猎物足迹链宽度内的猎物密度在理论上会产生足迹链遗漏。换算系数在不同的时空,不同猎物个体间差异较大,有待深入研究。传统的俄罗斯调查方法（FMP）需要知道动物的每日运动距离。国内对东北虎猎物日运动距离需进一步研究。大样方法通过在足迹链宽度范围内（猎物家域内）布设多条样线,将足迹链宽度范围内产生足迹链遗漏的几率减低。同时在一个大样方内通过多条样线排除重复的足迹链,并以一条足迹链相对于一只动物（即换算系数为1）统计样方出现的动物数,避免了繁杂的换算系数。2)东北虎猎物足迹链密度和猎物密度关系探讨。猎物密度和其足迹链密度呈显著的线性正比例关系,马鹿、狍子和野猪密度和其足迹链密度关系式分别为：Y=0.6835x+0.0736（n=53,R2=0.5732,p<0.05）,Y=0.4847x+0.1746（n=48,R2=0.8143,p<0.05）,Y=0.3802x+0.1864（n=53,R2=0.5296,p<0.05）,（Y为猎物密度,x为猎物足迹链密度）。相对于野猪而言,马鹿和狍子的足迹链密度能更好的指示真实密度,这可能和不同物种的行为有关。3)东北虎猎物种群数量监测。调查期间的方差（within survey time variance）为0.98。整体样方间方差（overall within-sample variance）为0.15（范围0.04-0.22）。2011年2-3月,共调查40条样线,每条样线长度8-12km,对狍子而言,如每2年监测1次,a=0.2,B=0.2,r=10%开展监测则有0.916的能力监测到狍子种群足迹链≥10%的下降趋势或者0.926的能力监测到狍子种群足迹链足迹链≥10%的增长；对野猪而言,有0.813的能力监测到野猪种群足迹链≥10%的下降趋势或者1的能力监测到足迹链≥10%的增长。其他监测因素不变,随着样线数量的增加,探测到猎物种群足迹链固定变化率的能力也随之增强。每2年监测1次,每条样线为8-12km,当样线达到24-32条时有80%的概率监测到研究地区狍子种群足迹链减少≥10%或增加≥10%的波动。每2年监测1次,每条长度为8-12km,样线为16-24条时,有80%的概率野猪种群足迹链增加≥10%的变动,当样线达到32-40条,有80%的概率监测到整个地区野猪足迹链减少≥10%的变动。a从0.05,0.1,0.15,0.20变化时探测到东北虎猎物种群足迹链固定变化率的能力也随之增强。研究地区梅花鹿和马鹿的足迹链密度为0的样线比例特高,无法用MONITOR软件设置监测方案。4)东北虎猎物足迹链调查法准确性和精度探讨。当调查18条样线（或抽样距离达到171.71km）,30条样线（或抽样距离达到297.47km）时能分别满足狍子,野猪的种群数量调查准确性的最低需求。在研究地区,对狍子和野猪的调查,1km2调查面积内至少需分别抽样0.1648km,0.2855km的样线。其他东北虎猎物（梅花鹿,马鹿）由于其调查遇见率极低,无法被评估。基于准确性指标,不同研究地区,随着狍子或野猪密度的增加,其最佳取样努力减少。同一地区,猎物密度大的物种（狍子）,其最佳取样努力小于密度小的东北虎猎物物种（野猪）的最佳取样努力。5)东北虎猎物种群变动趋势：预实验15条样线,不同的猎物物种的探测概率都很高。狍子在阔叶林中p=0.96,其他林型中p=0.94；野猪在阔叶林中p=1.0,其他林型中p=0.92（p为探测概率）。阔叶林中的调查条件记录值（值域1.62-2.15）高于其他林型调查条件记录值（值域1.56-2.54）。从2010-2013年,研究地区狍子种群指数增加了2.4倍。年平均种群指数回归斜率显著的大于O。野猪、梅花鹿、马鹿、斑羚和原麝等物种的种群密度也呈上升趋势,但其增长都没有显著大于0。人为干扰减轻的情况下,狍子种群能快速野外恢复,但是种群年增长率远低于其内禀增长率。在保护干涉前,狍子和野猪的密度就相对较高,显示狍子和野猪抗人为干扰能力强。而梅花鹿、马鹿、斑羚和原麝种群基数太低,除了外部影响外,可能存在着物种内对种群恢复的限制。6)中国野生东北虎数量监测方法有效性评估。（1）用虎网络信息收集法研究2006年完达山东部地区东北虎的种群现状,结果显示东完达山地区2006年东北虎数量为6—9只,由1只成年雄虎,2—3只成年雌虎,2--4只亚成体虎和1只小于1岁的幼体虎组成；（2）用猎物生物量和捕食者关系法得到东完达山地区2002年东北虎的密度为0.356只/100km2,能容纳22—27只东北虎；（3）用样线法在黑龙江的老爷岭南部和吉林省大龙岭北部面积1735.99km2的区域内设置样线64条,总长609kmm,没有发现东北虎足迹链。样线调查的结果表明,在2011年2月—3月该调查区域东北虎的数量为0只。监测结果表明,用猎物生物量和捕食者关系得到东北虎数量远远超过现实数量,人们对东北虎猎物的盗猎和猎套对虎的伤害可能是其主要原因；样线法调查得出的结果低于现实种群,主要原因是虎数量极低和调查者对野生虎行为学了解甚少,较难在野外有效的发现虎信息；且样线法监测仅应用于当东北虎以一定的密度（即有定居虎）存在的情况下（多数监测样线能发现虎信息）。虽然和样线法一样存在着诸如专家估计密度和真实密度之间的关系、虎足迹数量和虎真实密度间关系不确定、保守估计等内在缺点,在目前中国东北地区野生东北虎种群密度极低、且多是穿越于中俄边境地区的游荡个体的现状下,网络信息收集法是一种高效、可行的东北虎监测方法。此外,其他监测方法,如占有法、基于标志重捕远红外照相法、粪便DNA法、足迹数码信息法、警犬法等,应根据各种方法的理论前提、误差来源、适用范围和虎是否定居及密度等具体情况有选择地加以应用,且有些方法可能成为未来中国野生东北虎种群的有效监测工具。7)中国东北虎种群现状及其对野生东北虎保护应用研究。（1）2004年11月到2010年4月,东完达山地区共监测到138条可靠信息,研究地区由2只定居雌虎,一只定居雄虎,4-5只游荡虎和5只幼崽组成,其中2只幼崽生长为亚成体,但都没能在研究地区定居。虎的平均种群保持相对稳定,为4.83±1.47只到6.33±1.63只,虎最高密度为1.15/1000km2,成年虎种群密度最高为0.72/1000km2。（2）2002-2007年于中国东北地区监测到343条可靠信息。国内东北虎共为5个分布区：东完达山分布区,南部老爷岭-大龙岭虎分布区,北老爷岭分布区,张广才岭分布区（吉林省张广才岭和黑龙江省张广才岭南部山区）,哈尔巴岭分布区。其中东完达山和南部老爷岭-大龙岭虎分布区每年都多次发现虎信息,且多次发现小虎信息,为虎的繁殖栖息地。而张广才岭分布区、哈尔巴岭分布区和北老爷岭分布区虎活动频次不高,每2-3年才监测到虎信息。表明此三处分布区可能能被游荡虎到达,却不适应定居虎生存。除2002年外,年东北虎种群数量保持相对稳定,为12-19只。东完达山分布区有2-3只长期定居个体,除此之外,东北境内其他东北虎均属来回于中俄边界的游荡虎。8)虎和猎物关系：猎物是否限制中国东北虎种群野外恢复?2002年,研究地区东北虎猎物的生物量密度为87.9kg/km2。基于Karanth et al.（2004）的模型,东完达山虎密度为0.205955076个/100km2：基于Carbone and Gittleman （2002）虎的密度为0.356个/100km2；基于Miquelle et al.（1999）的公式,虎的密度为0.722个/100km2。三个模型得到虎数量都远高于监测到的虎数量,暗示着研究地区可能存在着对虎的盗猎。基于15个地区虎猎物生物量和虎密度,获得关系式：T=4.1012（LnPb）-21.626[F=19.17454,n=15,p=0.000898,R2=0.6312],其中T为虎密度（individuals/100km2）,Pb为虎猎物生物量（,kg/km2）,最低可维持定居虎长期生存的猎物生物量为195kg/km2,东完达山猎物种群不足以维持可长期生存的东北虎种群需要。从2002年到2008年,东完达山地区东北虎的主要猎物（3种）种群数量急剧下降。7年间,东北虎猎物密度下降达39.53%-44.80%。显示猎物贫乏是影响东完达山虎种群恢复的重要因子。9)吉林省珲春地区人虎冲突研究。从2001年12月到2010年7月,珲春地区内共发生191起人虎冲突事件,年均发生23.125次（x=23.125±16.5567,范围：11-54）。导致1人死亡,3人受伤；16匹马,188只牛,11只狗,32只羊,1只猪死亡；1匹马,43头牛,1只狗受伤。平均单次虎攻击牛的数量显著大于攻击马的数量（T=2.036,p=0.043）。平均单次攻击牛的数量和攻击狗的数量（T=-0.147,p=0.883）及平均单次攻击狗的数量和马的数量（T=1.731,p=0.096）都没有显著差异。假设无人为干扰情况下,虎可取食完猎物全部可利用部分。珲春地区至少有84.03%的虎取食行为受到人为干扰。虎对马和牛不存在偏好性,但相对于牛犊而言,虎偏好成年牛（χ2=22.843,df=l,p<0.05）；对于母牛而言,虎偏好公牛（χ2=82.569,df=1,p<0.05）。下雪期月平均被虎攻击的牛数量（x=0.916±0.26565）显著低于无雪期月平均数（x=2.4043±0.25485）,（T=-5.293,p<0.01）。2001年12月到2010年7月份,家畜损失共达77035美元,年平均损失9062.94美元。拖迹平均长度为40.3455米（40.3455±₃7.46947,n=55）。拖迹的长度和被捕杀的家畜体重无关（p>0.05,rs=-0.7,n=40）。捕杀现场的隐蔽度（x=0.25±0.07）显著低于取食场所的隐蔽度（x=0.54±0.14）,（r=0.19,p<0.05,n=25）。71.31%的冲突发生在树林中,88.52%发生在坡度为0-30间,84.43%发生在海拔50-317米间,更多发生在东或南坡。更多还原

【Abstract】 A study was conducted to evaluate the accuracy and practicability of several ungulate and tiger sample survey methods, design suitable ungulate monitoring plan for nature reserves for tiger conservation, explore the relationship between amur tiger and their prey and alleviate the conflict between tiger and human in Northeat China from2002to2013by using field population sample survey techniques, multiple statistical analysis, Bootstrap resample techniques et al., The main results were revealed as followes:1) Assessment of monitoring methods for ungulate population abundance in Northeast China:The accuracy and feasibility of three methods (belt transect method, Sample Plot Method recommend by experts from America, Russia and China, FMP method.) for monitoring ungulate population abundance were assessed in the Dalongling Nature Reserve, Heilongjiang Province and Wangqing Nature Reserve, Jilin Province from2010to2011.The results showed that:(1) the ungulate density by using the traditional transect belt (the width of transect belt is100m) are0.054±0.009individuals/km2,2.81±0.72individuals/km2,0.99±0.16individuals/km2,0.26±0.06individuals/km2for red deer, roe deer, wild boar, and sika deer respectively, Using the improved transect belt method (the width of transect belt equal to daily movement length of unglate), the density of ungulate are0.0042±0.0017individuals/km2,0.32±0.07individuals/km2,0.33±0.05individuals/km2,0.0094±0.002individuals/km2for red deer, roe deer, wild boar, and sika deer respectively.(2) the ungulate density by using FMP method are0.0096±0.005individuals/km2,0.93±0.24individuals/km2,0.058±0.013individuals/km2,0.0194±0.009individuals/km2for red deer, roe deer, wild boar, and sika deer respectively.(3) the ungulate density by using Sample Plot Method recommend by experts from America, Russia and China are0.006±0.007individuals/km2,0.654±0.148individuals/km2,0.311±0.154individuals/km2,0.074±0.042individuals/km2for red deer, roe deer, wild boar, and sika deer respectively. In the Russian Far East and elsewhere in Russia, the Formozov-Malyshev-Pereleshin (FMP) formula has been used to estimate the density of large mammals. However, the FMP requires knowledge of the mean daily distances moved by individuals and the factors affecting daily travel distances are not known for ungualte species in northeastern China. The precision of the traditional Chinese belt transect method is diminished by uncertainty about the width of the strip and the accuracy of the conversion coefficient that relates the number of tracks to the number of individual ungulates. Contray to above two methods, sample Plot Method recommend by experts from America, Russia and China avoids the conversion coefficient and has more accurate estime for ungulate density.2) Study on the relationship between spoor density and true density of ungulate:The feasibility of ungulate spoor for monitoring ungulate population abundance was assessed in Eastern Wanda Mountains, Heilongjiang Province in2002. The result showed that there are significantly linear correlation between the ungulate population density and ungulate spoor density. The fomula are Y=0.6835x+0.0736(n=53, R2=0.5732) for red deer, Y=0.4847x+0.1746(n=48, R2=0.8143) for roe deer, Y=0.3802x+0.1864(n=53, R2=0.5296) for wild boar respectively.3) Ungulate population monitoring design for nature reserve:We examined the efficacy of employing a spoor-count index to monitor trends in abundance of the ungulate, the main prey of amur tiger in the Northeast China. Results of simulations examining power suggested that spoor counts could be employed as part of a system to monitor unglate abundance given the critical assumption that changes in spoor counts reflected changes in ungualte population size. The mean coefficient of variance in sample unit track rate, estimated by ten sample units for3times, was0.15(range:0.04-0.22). Monte Carlo simulation suggests a monitoring system employing24to32routes8to12km long, sampled twice each year, could provide over80%power to detect a10%annual decline and increase in rod deer tracks with a20%chance of type I errors (a=0.20); A monitoring system employing32to40routes8to12km long, sampled twice each year, could provide over80%power to detect a10%annual decline in wild boar tracks with a20%chance of type I errors (a=0.20); and16to24routes8to12km long, sampled twice each year, could provide over80%power to detect a10%annual increasing in wild boar tracks with a20%chance of type I errors (a=0.20).4) Study on accurey of transect line method based on ungulate spoor:Estimations approached the real situation with increasing sampling effort. With increasing sampling effort, the accuracy of abundance estimation followed an exponential form. Bootstrap analysis of surveyed ungulate indicated that population sizes could reasonably be established from30line transects for wild boar (sampling distance:297.47km),150line transects for red deer (sampling distance:750km) and18line transects for roe deer (sampling distance:171.71km). A power law relationship between ungualte density and the optimal sampling effort was determined.5) Population trend and recovery patterns of ungulateUsing a sign-based index of abundance, we measured4-year trends in abundance of six ungulate species in Wangqing Nature Reserve in Northeast China. Regression slopes of annual indices against time indicated that population growth rates (r) is0.107for roe deer, which is significant bigger than zero. Thus, roe deer can recover relatively rapidly from low population levels. Although population growth rates (r) of other ungulate species all are positive value, they are not significant bigger than zero. Wild pigs were already relatively abundant when monitoring started, illustrating their resilience to hunting The unexpected failure of red deer, sika deer, musk deer, goral to recover suggests that reproductive behavior may override seemingly positive interventions (i.e., stopping poaching) that reduce mortality.6) Assessment of monitoring methods for tiger population abundance in Northeast China:The accuracy and feasibility of three methods (Information collecting networks, traditional transect survey and tiger-prey biomass relationships) for monitoring Amur tiger population abundance were assessed in the eastern Wanda Mountains, Heilongjiang Province and8bordering forest area between the southern Laoye Mountains in Heilongjiang Province and the northern Dalong Mountains in Jilin province during2002-2011.The results showed that:(1) there were at least6-9wild Amur tiger in the eastern Wanda Mountains (1male,2-3adult females,2-4sub-adults and1cub), in2006by using an information network for tigers;(2) there were22-27wild tigers in the eastern Wanda Mountains in2002-2004based on the prey biomass relationship method, which obviously exaggerated the tiger population size; and (3)0tigers in8bordering forestry areas between the southern Laoye mountains in Heilongjiang Province and the northern Dalong Mountains in Jilin province, in2011by using traditional transect method, which underestimated the true tiger population size. The reasons for exaggeration of the tiger population using the biomass method could be previous losses of tigers from the area due to snares and competition with another carnivore, especially with people for ungulates. The transect method may have underestimated tiger densities in the survey areas because it was based on little prior knowledge of tiger behavior. It could only be usefully applied when tigers exist in at least moderate densities (i.e., when there is a high probability of encountering tiger tracks along a suite of routes). Although there is inherent potential error and bias, such as the unclear relationships of "expert estimates" and the true density, and between tiger track densities and actual tiger densities, same as the transect method, the monitoring of tiger populations using information networks provides a record of minimum tiger presence, and may be an appropriate approach when tiger presence is extremely rare, transitory and unstable, such as in northeast China. This approach is economically efficient and should be further improved by established a wider network across the landscape to encompass all potential tiger habitat using better trained monitoring staff.7) Population status of Amur tiger and implication for its conservation in Northeast China:From November2004to April2010, tiger information colleting system was used to monitor tigers in Eastern Wandashan Montains, Northeastern China, to quantify changes in abundance of demographic groups and to identify underlying causes. Mean abundance in Eastern Wandashan Mountains were5or6tigers, the tiger population remained relatively stable. Transients and the number of tiger offspring were generally recorded at low levels. The number of breeding animals in the study also remained stable, with about2breeding females and1breeding male, which highlight the region as a potential source pool for Amur tiger population recovery in Northeast China. We also researched population status of Amur tiger in Northeast China,2002-2007, using tiger information colleting system. The results show that:there were6isolated Amur tiger distribution regions in Northeast China. They were the Eastern Wanda Mountains, the Northern Laoye Mountains, the Southern Laoye Mountains and the Dalong Mountains, the Zhangguangcai Mountains, the Haerba Mountains tiger distribution regions respectively, among which the Zhangguangcai Mountains and the Haerba Mountains distribution regions, central Northeast China distribution areas, and Northern Laoye Mountains apparently represented regions still accessible to dispersing tigers, but were not suitable for retaining a resident population. The amur tiger was on the verge of extinction in northeast China and average less than20tigers occurred in Northeast China from year2002-2007, which was presently sustained by emigration of individuals from Russia. Only with the exception of the eastern Wandanshan Mountains, there no longer existed a resident, stable tiger population elsewhere in Northeast China, and that, no reproduction of young was occurring. The Amur tiger in Northeast China existed largely as nomadic, isolated individuals. Based on the results of our research, we suggest identify tiger ecological corridor between China and Russian, alleviate the threats hindering the fluent immigration, shift our tiger conservation strategy from nature reserve or region-focused to landscape-focused management so that entire tiger populations are treated as a single management unit, establish a management zone less strict than nature reserve in Eastern Wanda Mountain and take effective activities to increase the prey density and alleviate the human-tiger conflict.8) The relationship between tiger and their prey:does prey density limit tiger recovery in northeast China?A residual population of Amur tigers probably survives in the Eastern Wanda Mountains (EWM) in China, where the main prey species are red deer (Cervus elaphus), eastern roe deer (Capreolus pygargus) and wild boar (Sus scrofa ussuricus). We used53snow sample plots each containing about29km of transects to detect ungulate presence, and determined their total density in EWM in2002to be87.9±8.9kg·km-2. We then applied these data to three published models that predict the relationship between tiger density and prey biomass density to obtain three estimates of tiger carrying capacity in EWM. Existing estimates of tiger density suggest that tigers were below carrying capacity estimates. Relationships between prey density and tiger density from15studies indicated a threshold prey biomass of195(CI,33-433)kg·km-2below which a tiger population cannot be sustained.Therefore, we concluded that the EWM population of tigers is in peril. We compared densities between the years2002and2008using comparable data and found that the EWM populations of the three ungulate prey species all experienced decreases of c.40-45%, apparently due to intense poaching. This rapid decline in prey density and pervasive threats to tigers and their prey in the EWM demands immediate and effective protection of ungulate and tiger populations from poaching if tigers are to persist and recover.9) Conflict between tiger and human (HTC) in Hunchun National Nature Reserve, Northeast China:We examined human-tiger conflict in Hunchun, Northeast China using data gathered from December2001to July2010.191cases of human-tiger conflict were documented, caused to3people injured and1people killed,16horses,188cattle,14dogs,27sheep,1pig killed and1horse,43cattle and1dog injured. Human disturbed the tiger feeding for cattle in at less100case of119total cattle killed cases (84.03%). Tigers attacked more cattle than horses per attack There were no significant difference between numbers of cattle attacked per attack than dog and between dogs and horse per attack. Tigers killed cattle and horse according to their availability. Examination of cattle kills showed that tiger killed a significantly greater proportion larger prey (e.g., bulls for cattle sex and adult for cattle classes). Overall, cattle predation was greatest in non-snowfall period which corresponded with lax livestock management. Cattle freely roam in forest and less well guarded during that time. The total of economic loss to tiger was valued at US$77035, of which the majority (88.21%, US$67950) was cattle loss. Horse and sheep loss account for8.44%and2.29%of total monetary loss respectively. Average annual livestock loss to tiger was US$9062.94. We measured the distance of drag for55times, the average of drag distance was about40.3455m (40.3455±37.46947, n=55). The mean hiding cover in killing sites (x=0.25±0.07) was lower than hiding cover in feeding sites (x=0.54±0.14),(T=0.19,p<0.05, n=25). The mean nearest distance to road was0.832056314km (0.832056314km±1.002989), and and the mean nearest distance to river was0.2633203km (0.2633203±0.24379), mean nearest distance to settlement was3.596962188km (3.596962188±2.199674308). Implications of our findings for mitigating livestock losses and for conserving large carnivores in Northeast China are discussed.更多还原