家养动物入侵是全球生物多样性丧失的重要原因之一。家狗（Canis familiaris）作为陆地上数量最多的食肉动物，因其移动灵活、食性多样、繁殖力强，在人类的饲喂与庇护下已广泛入侵野生动物栖息地，甚至保护区。家狗对野生动物的影响包括捕杀、竞争、干扰、病毒传播和杂交等，但由于在人类生活中扮演着重要角色，他们的影响长期受到忽视。探究家狗的时空活动规律及其与野生动物的相互作用对于降低上述影响，减缓人与野生动物矛盾，为保护区家养动物饲养与管控提出科学建议具有重要意义。东北虎豹国家公园是濒危野生动物东北虎（Panthera tigris altaica）和东北豹（Panthera pardus orientalis）的重要栖息地，栖息地退化、过度放牧、猎物短缺等问题严重限制了野生动物种群恢复，但目前公园内对于家狗入侵的研究尚属空白。
本研究首先利用问卷调查、血清采集、粪便分析的方法，从社会统计、饲养管理、保护态度、健康防疫等方面评估家狗入侵现状。其次，利用项圈监测技术，分析家狗日活动节律与离家行为特征。最后，基于“东北虎豹生物多样性长期定位观测网络”，通过红外相机监测数据，运用动态占域模型（dynamic occupancy model）、双物种占域模型（two-species occupancy model）、多响应置换法（multi-response permutation procedure）、联合物种分布模型（joint species distribution model）、核密度函数（kernel density function）等多种方法分析家狗与公园内野生动物的时空关系，探究家狗对野生动物群落的潜在影响。本研究主要结果如下：

（1）家狗种群统计特征（dog demography）：本研究于2018年1月共调查东北虎豹国家公园东部71条沟系和121户分布于森林中的林蛙房，统计得到家狗508只，平均每户饲养4.2只，公：母比例为1.5：1，大部分进入森林中自由活动，仅28.5%的家犬全天栓养。

（2）家狗病毒携带、猎杀和管理现状：抽样调查的202只家狗中犬瘟热病毒携带者占33.5%，犬细小病毒携带者占87.7%，2岁以上的家狗更易受病毒感染，主人防疫意识较低。冬季粪便分析显示家狗捕杀了至少6种野生动物，野猪（Sus scrofa）、狍（Capreolus pygargus）、梅花鹿（Cervus nippon）和狗獾（Meles leucurus），相对出现频率最高，分别为37.50%、23.62%、11.08%和20.84%，对野猪表现明显的捕食偏好，表明低约束性管理导致自由活动家狗可对野生动物种群产生危害。
（3）家狗空间分布：占域模型结果显示家狗活动主要集中在海拔较低的山谷、距离林蛙房更近和人类活动强度较高的区域。夏季平均占域率0.46（SE:0.14）高于春季0.41（SE:0.13）、秋季0.44（SE:0.11）和冬季0.43（SE:0.20），保护区内春、夏、冬三季占域率显著高于保护区外，目前保护区并未对家狗活动形成有效限制。
（4）家狗活动规律：红外相机数据和GPS项圈数据显示家狗四季均呈明显昼行性活动，与人类活动节律高度重叠（重叠系数△＞0.75）。家狗运动呈现以家为聚集中心的活动模式，偏好利用道路。日离家频率为0.31-1.71次，单次离家时长1.47-11.17 h，单次离家轨迹长度达4.15-21.93 km，最远离家距离达2.76-10.00 km，活动范围达0.96-46.81 km2（95%自相关核密度估计值）。
（5）野生动物与家狗的空间相互作用：双物种占域模型结果表明，在不同季节家狗与不同体型大小和营养级的哺乳动物表现出不同的相互回避作用。在春季，东北豹的出现降低家狗的占域率，家狗出现降低野猪、狍、狗獾（Meles leucurus）、黄喉貂（Martes flavigula）、豹猫（Prionailurus bengalensis）的占域率；在夏季，家狗出现降低野猪、狍、狗獾的占域率；在秋季，东北豹的出现降低家狗的占域率，家狗出现降低野猪、狍、赤狐（Vulpes vulpes）、狗獾的占域率；在冬季，东北虎出现降低家狗的占域率，家狗出现降低梅花鹿、黄喉貂的占域率。
（6）野生动物与家狗的时间相互作用：昼行性活动的野生动物（东北豹、梅花鹿、狍、野猪、黄喉貂）与家狗日活动节律呈中高度重叠（重叠系数△＞0.50），偏夜行性活动动物与家狗日活动节律重叠较低，仅赤狐与狗獾时间节律对家狗活动表现出显著差异。梅花鹿、狍、野猪、豹猫与家狗在共同出现相机点表现出明显的时空生态位分化。
（7）家狗对野生动物群落的影响：在不同季节中，群落相似性分析显示哺乳动物群落构成在有狗出现区域与无狗出现区域表现出明显差异（p <0.05）。基于相对多度数据的联合物种分布模型结果表明，由于对环境因素的趋同性，家狗与东北虎、梅花鹿、赤狐、貉（Nyctereutes procyonoides）、豹猫呈现显著正相关分布；家狗与东北豹、狍、野猪、黄鼬（Mustela sibirica）、狗獾、黄喉貂等呈现显著负相关分布。群落阈值模型（community thresholds model）结果显示不同动物对家狗的活动呈现阈值效应，当狗的相对多度达到1.92时，东北虎、梅花鹿、赤狐、豹猫的多度显著增加，提高了家狗被猎捕或捕杀其它动物的风险；当狗的相对多度达到1.10时，狍、东北兔（Lepus mandshuricus）的相对多度显著下降。
本研究表明家狗已广泛入侵东北虎豹国家公园，人类因素对家狗日活动节律和空间分布具有关键调控作用。自由活动的家狗可通过捕杀、病毒传播、时空间关系等途径对野生动物产生影响，而且这些途径相互交织可进一步加剧狗对野生动物群落构成和多样性的影响。我们建议对公园内饲养的家狗定期登记，建档监控，追踪健康与死亡情况，定期免疫接种，加强饲养环境消杀，推广栓养，限制其进入野生动物栖息地，建议尽快开展公园内全范围的家狗猎杀行为调查，减缓其对野生动物的影响。

Invasive species are regarded one of the most serious threats to global biodiversity, and invasion mammalian predators have played a disproportionate role in the decline and extinction of native species in a variety of ecosystems. The domestic dog (Canis familiaris) is the most common and widespread carnivore on the planet, living in close proximity to human communities. They also are a generalist and opportunistic species with a high level of mobility, behavioral flexibility and reproductive rates. Together, these characteristics enable dogs to survive and thrive in a variety of habitats, making them one of the most successful invasive species and facilitating their interaction with wildlife. Dogs interact with wildlife on several levels, including as predators, prey, competitors, and disease reservoirs or vectors. Although this problem is similar to threats posed by other invasive animals, managing roaming dogs is more difficult due to their close association with humans. Understanding the movement patterns, habitat use and health status of these free-roaming dogs, particularly in relation to conservation areas and threatened species' habitat, is a critical gap in minimizing the risk to wildlife associated with interactions with dogs.
China’s Northeast Tiger and Leopard National Park is conservation priority area for the recovery of endangered Amur tiger (Panthera tigris altaica) and Amur leopard (Panthera pardus orientalis) in Northeast Asia. Introduced species now have become a significant barrier to the survival and dispersal of wildlife populations in the area. However, as domestic dog has become an important human disturbance in the park, an analysis of the effect of dogs as a disturbance agent was not considered. In this study,
we used interviews, questionnaires and descriptive statistics to document patterns of dog ownership and care and to measure public attitudes toward management strategies to control free-roaming dogs. Blood samples from 202 dogs were collected to determine the degree of canine distemper (CDV) and Canine parvovirus exposure in dogs in the park. Using fecal analysis methods, we assessed the killing of wild animals by dogs. We fitted GPS collars to 16 dogs and studied the spatial and temporal dynamics of space use by domestic dog. We calculated 95% home ranges using auto-correlated kernel density estimates (95% AKDE) determined the degree to which their movements were predictable, and identified correlates of movement patterns. Finally, we analyzed the temporal and spatial relationship between dogs and native wildlife in different seasons using large-scale camera trapping data, an occupancy model, activity pattern overlaps, a multi-response permutation procedure, and a joint species distribution model.
We conducted 121 interviews in the household scattered the forest and found an average of 4.2 dogs per household. Only 28.5% of dogs (N=508) were always leashed (< 100 m from the household). The majority of the dogs have strong hunting abilities. For a total of 202 blood samples, serological tests showed that 33.5% of domestic dogs were exposed to CDV, and 87.7% of domestic dogs were exposed to CPV. Domestic dogs over 2 years old are more vulnerable to virus infection, and their owners have low awareness of epidemic prevention. Diet analysis revealed that six prey species were identified in dog scats. Badger (Meles leucurus) and three ungulates, including wild boar (Sus scrofa), roe deer (Capreolus pygargus) and sika deer (Cervus Nippon), were mainly hunted by domestic dogs, and their percent occurrence were 20.29%, 21.01% and 11.59%, respectively. Dogs showed a notable preference for wild boar.
Space use by the domestic dogs is higher in summer (occupancy probability = 0.46) than in other seasons. Our findings show that dogs were more likely to use sites with lower elevations, closer proximity to forest houses and higher human encounters. GPS tracking data revealed dogs were mostly located near forest houses but exhibited a diurnal pattern of directed excursions away from their home locations. Foraging dogs rarely used forest habitat, when entered, trails and roads were selected for movement.
The daily frequency of leaving home ranged from 0.31 to 1.71 times, the time length of a single departure ranged from 1.47 to 11.17 h, and the track length of a single departure ranged from 4.15 to 21.93 km. The maximum line distance from home ranged from 2.76 to 10.00 km, and 95% AKDE ranged from 0.96-46.81 km2.
The overlap in dogs and predator (tiger, leopard), and their potential prey revealed species-specific spatial associations and activity overlaps. Dogs had higher activity overlap with leopard, sika deer, roe deer, wild boar, yellow-throat marten. Red fox (Vulpes Vulpes) and badger noticeably decreased diurnal activities at sites with dog presence than with dog absence, respectively. Sika deer, roe deer, wild boar, leopard cat ((Prionailurus bengalensis) showed obvious spatial and temporal niche separation with domestic dogs at the camera sites of co-occurrence.
Based on a two-species occupancy model, the interaction between domestic dogs and wildlife was found to be dependent on mammal body size and season. The presence of the Amur leopard decreased the occupancy of the domestic dog in spring and autumn, while the presence of the Amur tiger decreased the occupancy of the domestic dog in winter. Domestic dog reduced the occupancy probability of wild boar, roe deer, badger (Meles leucurus), yellow-throated martens (Martes flavigula), and leopard cats in spring. In the summer, the presence of domestic dog reduced the likelihood of wild boar, roe deer and badger occupancy. In autumn, the presence of domestic dog tends to reduce the occupancy probability of wild boar, roe deer, red fox and badger. Domestic dogs tend to reduce the occupancy probability of wild boar, roe deer, red fox, and badger in autumn.
Joint species distribution model showed that, spatially, domestic dog had positive co-occurrence with the Amur tiger, sika deer, red fox, raccoon dog (Nyctereutes procyonoides), and leopard cats, and had negative co-occurrence with the Amur leopard, roe deer, wild boar, Siberian weasel, badger and yellow-throated marten. When species were assessed individually, a significant threshold relationship to the dog distribution was found to occur in 6 of the 20 examined species, with two (roe deer and manchurian hare) having a negative drastic change when relative abundance index of dog is 1.10.
In conclusion, our finding demonstrated that domestic dogs in the Northeast Tiger and Leopard National Park can have a negative impact on wild animals through hunting, virus transmission, and changes temporal and spatial use of the animal. Human factors play an important role in regulating the daily activity rhythm and spatial distribution of domestic dogs. We propose that each year, local managers create files for each dog raised by the owners every year, register detailed information about their dogs, track health and deaths, implement regular immunization. Government managers should raise awareness about the important of controlling dog free movement in the forest, and encourage dog owners to develop responsible restrictions on dog’s daily activity.