东北虎 (Panthera tigris altaica) 和东北豹 (P. pardus orientalis) 作为亚洲温带阔叶混交林的顶级食肉动物，对维持该地区生态系统的完整性和稳定性起到重要作用。然而由于栖息地的丧失和破碎化、猎物资源不足、疾病传播和人类干扰等威胁，种群正处于濒危状态。由于这2种大型猫科动物需要较大的栖息地，以及较高的能量需求，生活在低质量的破碎生境斑块中，对这些物种来说是一种挑战。当前种群规模和分布范围已萎缩到历史最低水平。当大型食肉动物种群规模下降到某个阈值以下时，物种可能会迅速灭绝，并产生级联效应。生活在中俄边境的孤立的虎、豹种群是中国东北虎、豹恢复的主要希望，尽管由于近些年中俄双方保护措施的加强和国家公园建设，虎豹种群已呈现增长趋势，但由于该跨境种群规模小 (2015年红外相机调查数据显示东北虎38个个体和东北豹87个个体)，以及犬瘟热和放牧干扰，种群处于中等近交衰退水平，其延续正面临着严重的挑战。基于已有的种群基础生态学研究和调查数据，针对各种威胁，构建不同模型，对中俄边境的东北虎和东北豹进行全面和系统的种群生存力分析 (P VA)，预测在不同保护情景下种群的发展趋势，对制定保护措施和中国的虎豹种群恢复至关重要。
为了探究犬瘟热、栖息地丧失、近交衰退的威胁，本研究采用由基于个体的种群统计学模型和流行病学模型组成的P VA元模型 (metamodel)，在3.14、6.29和12.26致死当量 (LEs) 三种近交衰退情况下评估了保护区内家犬管控、增加与邻近大种群的连通 (针对东北虎种群) 、低覆盖的疫苗接种 (针对东北豹种群) 和扩大栖息地的保护方案。此外，为了评估过度放牧对包括顶级食肉动物在内的食物链的影响，本研究以东北虎为例，构建了一个包含年龄结构的梅花鹿-东北虎食物链时间动态模型。模型将牛的放牧作为一种干扰，从理论上探讨了放牧强度对野生动物种群稳态的影响以及放牧的强度和持续时长对种群时间轨迹的动态影响。为更全面地了解复杂景观中的虎、豹与主要猎物间的相互作用，本研究以东北虎为例，对中国东北虎豹国家公园与俄罗斯豹地国家公园组成的东北虎分布区进行了参数化，搭建了一个基于个体的空间显式仿真模型。该模型具有解决与东北虎、豹研究相关的一系列问题的灵活性。本研究以探究放牧及人类干扰景观对野生动物种群扩散的影响为例，初步展示其在理论和实际应用中的表现，评估了禁牧和建立扩散廊道的有效性。本研究的主要研究结果如下：
P VA元模型结果：(1) 在三种近交衰退情况下，若不采取任何措施中俄边境东北虎种群100年内的灭绝率分别为64.4%、90.6%和99.8%；(2) 单独的家犬管控或栖息地扩大并不能有效保证虎种群在未来100年内的存活，与邻近种群的连通使种群在中弱近交衰退下实现了正增长，灭绝概率分别下降到1.4%和4.2%，但在强近交衰退情况下，种群灭绝概率依旧较高为14.4%；(3) 当上述三种保护方案相结合时，即使在最高水平的12.26 LEs的近交衰退情况下，东北虎种群的灭绝概率也小于5.8%；(4) 在三种近交衰退水平下，若不采取任何保护措施，东北豹种群100年内的灭绝概率分别为20.2%、61.8%和99.9%；(5) 对豹进行低覆盖率的疫苗接种和家犬管控来可以有效地提高东北豹种群的生存概率，在这些管理措施的基础上加上栖息地的扩大，种群将有机会进一步扩大，在3.14LEs、6.29LEs近交水平下，种群灭绝概率下降至0.9%和9.7%。
食物链模型结果：(1) 牛的放牧密度会导致野生动物种群的突然稳态转变。当放牧密度在临界点7.763头/km²以下，野生动物可以与放牧共存，但当牛的密度超过临界点时，梅花鹿和东北虎种群会在100年内迅速走向灭绝；(2) 在过度放牧的情况下，梅花鹿种群受放牧强度的影响更大，而东北虎种群受放牧持续时间的影响更大。
基于个体空间显式模型结果：(1) 影响梅花鹿和东北虎向东北虎豹国家公园内部扩散的主要因素是放牧和人类干扰景观的阻碍，核心分布区的梅花鹿只能绕过干扰景观利用研究区东北部有限的廊道扩散到公园内部；(2) 禁止放牧和建立廊道为促进野生动物种群扩散提供了可能。
本研究通过三种模型，探究了犬瘟热、近交衰退、栖息地丧失、过度放牧等威胁对东北虎和东北豹种群的影响，并分析不同保护措施的相对效益。对种群的主要管理建议强调了保护东北虎、豹种群需要多方面的协同努力，如果不缓解疾病等因素造成的高死亡率，仅扩大栖息地对种群的恢复作用十分有限。对于东北虎种群来说，再引入或重新建立与邻近种群的生境连通是一个重要的长期目标。食物链模型在理论上证明，野生动物保护和放牧可以共存，但放牧密度需保持在临界点以下，若超过临界点，野生动物种群可能发生严重的稳态转变，该结果为制定与平衡畜牧业和野生动物保护相关的政策提供了新的见解。此外放牧和人类干扰景观是阻碍野生动物种群扩散的主要因素，通过禁牧或建立廊道促进野生动物种群向虎豹公园内部扩散，是目前亟需解决的问题，需要立即关注并采取行动。本研究的建模框架可以为保护规划提供建议和科学支持，也可以为世界其他地区的濒危物种管理措施的制定提供参考。

The Amur tiger (Panthera tigris altaica) and Amur leopard (P. pardus orientalis), as the top predators of temperate broad-leaved mixed forests in Asia, play an important role in maintaining the integrity and stability of the ecosystem. However, they are struggling on the brink of extinction due to a variety of threats, such as habitat loss and fragmentation, scarcity or lack of prey, disease transmission and human interference. Living in low-quality fragmented habitat patches presents a challenge for these two big cat species because of their large habitat and high energy requirements. As a result, their population size and distribution range have recently shrunk to a historically low level. However, when the population size drops below a certain threshold, the species might likely go extinct quickly. The loss of the top carnivores might also initiate unexpected cascade effects throughout the entire food chain. The isolated tiger and leopard populations living along the China-Russia border are the main hope for the recovery of the Amur tiger and leopard in China. Although the populations have been growing in recent years due to the strengthening of conservation measures and the construction of national parks in both Russia and China, their continuation is facing serious challenges due to the populations are in moderate inbreeding depression given the small size of thses two cross-border populations (camera survey data from 2015 revealed 38 Amur tigers and 87 Amur leopards), as well as canine distemper and grazing disturbances. Based on the available basic ecological information and survey data of populations, a comprehensive and systematic population viability analysis (P VA) of Amur tiger and leopard populations along the China-Russia border, with different models constructed to address various threats and predict population trends under different conservation scenarios, which is essential for the development of conservation measures and the recovery of tiger and leopard populations in China. by constructing different models for various threats is essential for the conservation and recovery of the Amur tiger and leopard in China. Predicting population trends under different conservation scenarios provides a reference for the development of conservation measures.
In order to explore the threat of canine distemper, habitat loss, and inbreeding depression, a P VA metamodel that combines a traditional individual-based demographic model with an epidemiological model was developed. The P VA metamodel was used to assess options for controlling the impact of negative factors through domestic dog management in protected areas, increasing connectivity to the neighboring large population (only for Amur tiger population), low-coverage vaccination (only for Amur leopard population) and habitat expansion under three inbreeding depression scenarios at 3.14, 6.29, and 12.26 lethal equivalents (LEs). To evaluate the effects of overgrazing on food chains that involve carnivorous species, a temporally dynamic, age-structured trophic dynamics model for the sika deer-Amur tiger trophic chain was developed by using the Amur tiger as a case study, which incorporated cattle grazing as a disturbance. The model, was used to theoretically explore the impact of cattle grazing intensity on the regime behavior of wildlife populations, and the temporally dynamic impact of periods of heavy grazing on the population trajectory. To better understand the interaction between top predators and main prey in complex landscapes, the distribution area of Amur tiger composed of Northeast Tiger and Leopard National Park, China and Leopard National Park, Russia was parameterized . Using the Amur tiger as an example, an individual-based spatially explicit simulation model was built. The model has the flexibility to address a range of issues relevant to the study of the Amur tiger and leopard. In this study, the effects of grazing and human disturbance of the landscape on the dispersal of wildlife populations were explored as an example to initially demonstrate its performance in theoretical and practical applications. Also the model was used to evaluate the effectiveness of grazing bans and the creation of dispersal corridors.
P VA metamodel results: (1) Without intervention, under three cases of inbreeding depression, our metamodel predicted the extinction within 100 years is 64.4%, 90.6% and 99.8%, respectively; (2) The dog management or habitat expansion independently did not ensure tiger population viability for 100 years, and connectivity to the neighboring population resulted in positive population growth under 3.14 LEs and 6.29 LEs inbreeding depression, reducing the probability of extinction to 1.4% and 4.2% respectively, while under 12.26 LEs inbreeding depression the probability of population extinction remained high at 14.4%; (3) When the above three conservation scenarios were combined, even at the highest level of 12.26 LEs inbreeding depression, the probability of extinction was kept below 5.8%; (4) In the absence of intervention, the probability of extinction of the Amur leopard population within 100 years was 20.2%, 61.8% and 99.9%  at the three inbreeding depression levels, respectively; (5) Low-coverage vaccination of leopards and domestic dog management effectively improved the survival probability of the leopard population, and coupled with habitat expansion, the population had the opportunity to expand even more. Under 3.14 LEs, 6.29 LEs inbreeding depression, the probability of population extinction decreased to 0.9% and 9.7%.
Food chain dynamics model results: (1) cattle density led to an abrupt regime shift in wildlife populations. When grazing density was below the tipping point of 7.763 individuals/km², wildlife could coexist with grazing livestock. But when cattle density exceeded the tipping point, sika deer and Amur tiger populations moved towards extinction rapidly within 100 years; (2) In the case of heavy grazing, the sika deer population were more affected by grazing intensity, and the Amur tiger populations were more affected by grazing duration.
The spatially explicit agent-based model results: (1) the main factors that affect the dispersal of the sika deer and Amur tiger to the interior of the Northeastern Tiger and Leopard National Park were the obstruction of grazing and human disturbance landscape. There is a limited corridor for the dispersal of sika deer into the interior of the park in the north-eastern part of the study area; (2) grazing ban and establishment of corridors provided the possibility to enhance dispersal of wildlife.
In this study, three models were constructed to explore the impacts of threats such as canine distemper, inbreeding depression, habitat loss, and overgrazing on Amur tiger and Amur leopard populations, and analyzed the relative benefits of different conservation measures. The key management recommendations for the populations emphasize the importance of a multifaceted synergistic effort. Simply expanding the habitat alone has little impact on population recovery without mitigating the high mortality rate caused by factors such as disease. For Amur tiger, reintroduction and re-establishing habitat connectivity to the neighboring population is an important long-term objective. The food chain model theoretically demonstrates that wildlife conservation and cattle production can coexist when grazing densities is below the tipping point, but that serious regime shifts in wildlife populations may occur if grazing intensity exceeds the tipping point. These findings provide new insights that are useful in developing policies related to balancing livestock and wildlife conservation. Furthermore, grazing and human disturbances are the primary impediments to the expansion of wildlife populations. It is a pressing issue that necessitates prompt attention and remedial action to facilitate the unimpeded dispersal of wildlife populations into the park's interior via the implementation of grazing prohibitions or the establishment of corridors. More broadly, our modeling framework could provide critical perspectives and scientific support for conservation planning, as well as specific adaptive management actions for endangered species worldwide.